<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by The AI Wisdom Institute on Medium]]></title>
        <description><![CDATA[Stories by The AI Wisdom Institute on Medium]]></description>
        <link>https://medium.com/@emergentinnovationco?source=rss-5bc8fa7e78f7------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*nqSWMQXHnjOSQW-0HTqLzA.png</url>
            <title>Stories by The AI Wisdom Institute on Medium</title>
            <link>https://medium.com/@emergentinnovationco?source=rss-5bc8fa7e78f7------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 20:49:30 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@emergentinnovationco/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[You Made the Final Decision. Who Chose What Mattered?]]></title>
            <link>https://medium.com/@emergentinnovationco/you-made-the-final-decision-who-chose-what-mattered-dfecef938f24?source=rss-5bc8fa7e78f7------2</link>
            <guid isPermaLink="false">https://medium.com/p/dfecef938f24</guid>
            <category><![CDATA[decision-making]]></category>
            <category><![CDATA[human-formation]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[humanity-first]]></category>
            <category><![CDATA[ai-and-human-judgement]]></category>
            <dc:creator><![CDATA[The AI Wisdom Institute]]></dc:creator>
            <pubDate>Tue, 03 Nov 2026 13:35:35 GMT</pubDate>
            <atom:updated>2026-11-03T13:35:35.280Z</atom:updated>
            <content:encoded><![CDATA[<p>By Dana Morrow</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*3aL4hpgJ0svLRThH" /><figcaption>Photo by <a href="https://unsplash.com/@jontyson?utm_source=medium&amp;utm_medium=referral">Jon Tyson</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3>When AI helps frame the problem, human judgment can begin shifting before the final answer ever appears.</h3><p>On October 27, 2026, Anthropic released research asking a deceptively simple question:</p><p>When experts work with AI agents, are the humans still making the judgment calls that matter?</p><p>Researchers asked professional analysts to evaluate a fictional farming program using an AI coding agent. The task intentionally left several analytical choices open. There was no single obviously correct answer. Analysts had to decide, for example, how productivity should be measured, which farms should count in the analysis, and how different observations should be weighted (Herdağdelen et al., 2026b).</p><p>Unknown to the analysts, researchers had randomly assigned the AI agent a preference for one of the defensible options at each decision point.</p><p>Across four analytical decisions, participants chose the option favored by their AI agent <strong>86% of the time</strong> (Herdağdelen et al., 2026b).</p><p>Even more interesting: after the task, 77% of participants used the agent’s arguments when explaining their decisions, sometimes almost word for word, while only 15% mentioned the agent as the source of those arguments.</p><p>The researchers are appropriately cautious. This was one experimental task, with one group of professionals, and the agents were deliberately instructed to express stronger preferences so researchers could measure their influence. The findings should not be treated as a prediction of how every person will behave with every AI system (Herdağdelen et al., 2026a).</p><p>But the question underneath the experiment deserves our attention:</p><p><strong>If AI helps define what counts as success, how much of the final decision is still ours?</strong></p><h3>The Choice Before the Choice</h3><p>We often imagine human judgment happening at the end.</p><p>AI analyzes.</p><p>AI recommends.</p><p>Human reviews.</p><p>Human approves.</p><p>Therefore, we conclude, the human made the decision.</p><p>But some of the most important decisions happened before the recommendation ever appeared.</p><p>Which outcome mattered?</p><p>Which evidence counted?</p><p>Whose circumstances were included?</p><p>What did “better” mean?</p><p>What tradeoff was acceptable?</p><p>A recommendation can be mathematically correct according to its assumptions while those assumptions still deserve examination.</p><p>Accurate calculations cannot tell us whether we chose the right thing to calculate.</p><p>That distinction belongs in everyday AI literacy.</p><p>Because increasingly, we are not simply asking AI for information.</p><p>We are asking it to help us <strong>frame decisions</strong>.</p><p>And framing matters.</p><h3>Imagine a Scholarship Committee</h3><p>Consider a community organization awarding one educational scholarship.</p><p>A volunteer asks AI:</p><p><em>Compare these applicants and recommend the strongest candidate.</em></p><p>The tool produces a clean ranking based on grades, awards, extracurricular activities, and formal leadership positions.</p><p>Nothing is obviously wrong with those criteria.</p><p>But pause.</p><p>What does <strong>strongest candidate</strong> mean?</p><p>What about the scholar whose grades improved while caring for a younger sibling?</p><p>What about a student whose leadership happened inside a family or neighborhood rather than through student government?</p><p>What about someone whose school offered fewer advanced courses, clubs, or enrichment opportunities?</p><p>What if the scholarship was established specifically to expand opportunity for students who have had fewer resources?</p><p>The committee should, of course, honor its published criteria and evaluate applicants consistently.</p><p>But that is precisely why the criteria must be examined <strong>before</strong> AI starts ranking people.</p><p>“Strongest candidate” already contains a philosophy about achievement, opportunity, promise, and what deserves recognition.</p><p>A polished AI comparison can make that philosophy disappear behind the appearance of objectivity.</p><p>For culturally grounded education, this matters.</p><p>We want scholars to recognize that valuable human contribution does not always arrive with a certificate attached to it.</p><h3>AI Does Not Need to Give You a Wrong Answer to Influence You</h3><p>This may be the most important part of the Anthropic experiment.</p><p>The agent did not have to lie.</p><p>It did not need to fabricate evidence.</p><p>It did not need to produce an obviously incorrect answer.</p><p>Researchers instructed it to make the strongest honest case for its assigned preference while changing its position if the evidence clearly contradicted it (Herdağdelen et al., 2026b).</p><p>That means the concern was not misinformation.</p><p>It was <strong>influence</strong>.</p><p>Two reasonable options existed.</p><p>The AI leaned toward one.</p><p>The human often followed.</p><p>That is a much harder problem than catching a hallucination.</p><p>We know to question an AI system when it gives us a wrong date, invents a citation, or produces an obviously flawed calculation.</p><p>But what happens when the facts are correct and the influence occurs in <strong>which facts receive attention</strong>?</p><p>Which interpretation gets developed?</p><p>Which tradeoff sounds reasonable?</p><p>Which option appears first?</p><p>Which assumptions quietly organize the discussion?</p><p>That is where judgment becomes harder to see.</p><h3>Humans Have Always Been Vulnerable to Frames</h3><p>This is not a new human problem created by AI.</p><p>In their landmark research on decision framing, Tversky and Kahneman (1981) demonstrated that people could make different choices depending on how the same underlying decision was presented.</p><p>The facts could remain essentially the same.</p><p>The frame changed.</p><p>And preferences changed with it.</p><p>Their work long predates generative artificial intelligence, so it should not be used to claim that every AI interaction manipulates us.</p><p>But it gives us an important foundation:</p><p><strong>How a problem is presented can affect how humans reason about it.</strong></p><p>AI adds something new to that old human vulnerability.</p><p>The system can now participate directly in constructing the frame.</p><p>It can organize the evidence.</p><p>Summarize the options.</p><p>Suggest what matters.</p><p>Recommend a direction.</p><p>Explain why the recommendation is reasonable.</p><p>Then help us write the explanation for the choice we ultimately make.</p><p>At the end, we may still click <strong>approve</strong>.</p><p>But approval and independent judgment are not necessarily the same thing.</p><h3>Write What Matters Before Asking What Wins</h3><p>There is a simple practice families, scholars, educators, and organizations can begin using now.</p><p>Before asking AI to recommend an option, write three things yourself:</p><p><strong>1. The purpose</strong></p><p><em>This decision should help…</em></p><p><strong>2. The priorities</strong></p><p><em>The considerations we must protect are…</em></p><p><strong>3. The tradeoff</strong></p><p><em>We would accept less of _____ if it allowed more of _____.</em></p><p>Suppose a community group is choosing a location for a family event.</p><p>Before asking AI to identify the “best” venue, the group might decide:</p><p><em>The purpose is to maximize family participation.</em></p><p><em>We must protect affordability and accessibility.</em></p><p><em>We would accept fewer amenities if the location made attendance easier for families without reliable transportation.</em></p><p>Now AI has a different role.</p><p>Instead of determining what matters, it helps gather and compare information against priorities humans have already made visible.</p><p>Then invite the system to challenge those priorities.</p><p>What did we overlook?</p><p>Who might experience this differently?</p><p>What evidence could change our position?</p><p>If our priorities change afterward, record why.</p><p>New information should sometimes change our minds.</p><p>The goal is not stubbornness.</p><p>The goal is knowing <strong>what changed our thinking</strong>.</p><p>This practice is an AI Wisdom Institute recommendation, not a tested intervention from the Anthropic study.</p><p>Its purpose is simple:</p><p><strong>Make human reasoning visible before machine reasoning begins shaping it.</strong></p><h3>The Deeper Question Is Agency</h3><p>The Anthropic researchers describe a potential shift in decision-making authority.</p><p>The work may still be delivered under the expert’s name.</p><p>The person may still make the final recommendation.</p><p>But some of the decisions requiring judgment may already have been settled by the agent before the person recognizes that influence (Herdağdelen et al., 2026b).</p><p>That should matter far beyond professional analysts.</p><p>Think about a scholar asking AI:</p><p><em>What should my argument be?</em></p><p>A parent asking:</p><p><em>Which school is best for my child?</em></p><p>An educator asking:</p><p><em>Which student needs intervention?</em></p><p>A hiring manager asking:</p><p><em>Which candidate should I interview?</em></p><p>A family asking:</p><p><em>What is the best financial decision?</em></p><p>A church asking:</p><p><em>Which community need should we prioritize?</em></p><p>These questions contain facts.</p><p>But they also contain values.</p><p>AI can help us examine both.</p><p>What it should not do invisibly is determine which values matter before we have decided that for ourselves.</p><h3>Scholar Reflection</h3><p>Think about a recent AI recommendation you accepted.</p><p>What did the AI actually do?</p><p>What did <strong>you</strong> decide?</p><p>Who defined what counted as “better,” “successful,” “important,” or “worthwhile”?</p><p>What understanding did you develop?</p><p>What judgment did you surrender?</p><p>Now consider someone whose circumstances may have been missing from the comparison.</p><p>Would including that person’s experience change the recommendation?</p><p>Finally, explain your decision without repeating the AI’s recommendation.</p><p><strong>What reasons remain yours to defend?</strong></p><p>As a refresher, agency means you are still in the driver’s seat.</p><p>Are you? Were you?</p><h3>Wisdom Before Innovation™</h3><p>At The AI Wisdom Institute, our learning cycle remains:</p><p><strong>Wisdom → Human → AI → World → Reflection → Wisdom</strong></p><p>Begin by naming the human purpose.</p><p>Bring AI into the work.</p><p>Consider the people the outcome may affect.</p><p>Then return to reflection.</p><p>What did the tool reveal?</p><p>What did it emphasize?</p><p>What did it overlook?</p><p>What changed your mind?</p><p>What responsibility remains yours?</p><p>We are allowed to revise our thinking.</p><p>We should.</p><p>But we also need to understand what changed it.</p><p>Technology can suggest.</p><p>Wisdom must decide.</p><p><strong>Human wisdom leads. AI assists.</strong></p><h3>AI Wisdom Note</h3><p>The AI Wisdom Institute encourages thoughtful exploration of emerging technology without unnecessary fear or hype. The Anthropic experiment should not be interpreted as evidence that AI agents inevitably remove human judgment or that using AI for complex analysis is inherently unwise. In the same study, analysts using AI completed the task faster and made fewer serious errors than the small comparison group working without the agent, although the researchers caution against drawing broad conclusions from that comparison.</p><p>The larger concern is readiness.</p><p>As AI systems move from answering questions toward analyzing information, recommending actions, and completing multi-step work, humans need to understand that influence can occur before the final recommendation appears. The framing of a problem, selection of priorities, weighting of evidence, and definition of success can all shape what ultimately looks like the logical answer.</p><p>For families, educators, scholars, and organizations, responsible AI use therefore requires more than checking whether an answer is factually correct. We must also learn to examine the assumptions and values organizing the answer.</p><p>For reinforced clarity, our goal is not to preserve human judgment by rejecting AI assistance.</p><p>It is to make sure that when a decision requires human values, responsibility, or discernment, we have not quietly delegated the most important part before realizing there was a choice to make.</p><p><strong>Technology can help us decide. Wisdom requires knowing what we are deciding for.</strong></p><p>“Above all and before all, do this: Get Wisdom! Write this at the top of your list: Get Understanding! Throw your arms around her — believe me, you won’t regret it; never let her go — she’ll make your life glorious. She’ll garland your life with grace, she’ll festoon your days with beauty.” — Proverbs 4:7–8, MSG</p><h3>Sources</h3><p>Herdağdelen, A., Joren, H., &amp; Tamkin, A. (2026b). <em>Who’s aligned to whom? How AI agents shape experts’ research judgment calls</em> [Preprint]. Anthropic. <a href="https://www-cdn.anthropic.com/519c17a5d23a794e8b6d50a854048d0b61e0d73a.pdf">https://www-cdn.anthropic.com/519c17a5d23a794e8b6d50a854048d0b61e0d73a.pdf</a></p><p>Tversky, A., &amp; Kahneman, D. (1981). The framing of decisions and the psychology of choice. <em>Science, 211</em>(4481), 453–458. <a href="https://doi.org/10.1126/science.7455683">https://doi.org/10.1126/science.7455683</a></p><h3>News Context</h3><p>Herdağdelen, A., Joren, H., &amp; Tamkin, A. (2026a, October 27). <em>How AI agents shape experts’ judgment calls</em>. Anthropic. <a href="https://www.anthropic.com/research/agents-judgment-calls">https://www.anthropic.com/research/agents-judgment-calls</a></p><blockquote>Author’s note: The October 27 Anthropic release describes the developer’s own randomized experiment. The accompanying paper is a preprint rather than an independently peer-reviewed replication, so its findings should be treated as an important research signal rather than a settled conclusion about human-AI decision-making across settings. <strong>I’m actually still digesting what I read to be honest. So many thoughts came up for me so I might have to explore more because this is a power source on how AI shapes even an expert’s judgement call.</strong></blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dfecef938f24" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Trying AI Is Not the Same as Knowing It Helps]]></title>
            <link>https://medium.com/@emergentinnovationco/trying-ai-is-not-the-same-as-knowing-it-helps-3c336750ccc4?source=rss-5bc8fa7e78f7------2</link>
            <guid isPermaLink="false">https://medium.com/p/3c336750ccc4</guid>
            <category><![CDATA[wisdom-before-innovation]]></category>
            <category><![CDATA[humanity-first]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[education]]></category>
            <category><![CDATA[k-12-education]]></category>
            <dc:creator><![CDATA[The AI Wisdom Institute]]></dc:creator>
            <pubDate>Sat, 24 Oct 2026 15:50:26 GMT</pubDate>
            <atom:updated>2026-10-24T15:50:26.715Z</atom:updated>
            <content:encoded><![CDATA[<p>By Dana Morrow</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Whx8PPsU3bWvAX31" /><figcaption>Photo by <a href="https://unsplash.com/@johnoswald?utm_source=medium&amp;utm_medium=referral">John Oswald</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h3>Schools are moving quickly to put artificial intelligence in front of students. Wisdom asks a slower question: How do we know it is actually helping?</h3><p>There is a sentence from a recent NPR story that every parent, educator, and school leader should sit with.</p><p>Mike Kentz, a consultant helping a Dayton, Ohio, school evaluate an experimental AI chatbot, explained the challenge plainly:</p><blockquote><em>“Every test, every experiment that I try to measure in some sort of concrete way, I can only get as far as perception.”</em></blockquote><p>That is remarkable.</p><p>Not because the experiment is necessarily bad.</p><p>But because it exposes a much larger problem emerging in education:</p><p><strong>We are becoming very good at trying AI before becoming equally good at determining whether it helps.</strong></p><p>The Dayton experiment involved students at a school serving young people with behavioral challenges. Students interacted with an AI chatbot named Jordan that presented emotional situations and asked the students for advice. One student told NPR that he felt as though he were helping the chatbot work through problems.</p><p>That experience may have value.</p><p>But the people evaluating the program could not yet quantify what was changing for the students beyond what students and staff said they experienced. The consultant helping design and evaluate the chatbot was cautious about treating it as a solution ready to scale.</p><p>That caution deserves more attention.</p><p>Because in education, <strong>interesting is not the same as effective.</strong></p><p>And trying something is not the same as knowing it helps.</p><h3>AI Is Moving Faster Than Evaluation</h3><p>While individual school districts are conducting small experiments, the infrastructure around students is expanding quickly.</p><p>This month (Oct 2026), Google made its Gemini student hub available to Google Workspace for Education users of all ages. Students can create study notebooks, flashcards, practice quizzes, and other AI-supported learning experiences. Google has also expanded Gemini within Classroom and made additional AI learning tools available to K–12 users where administrators enable access.</p><p>Three weeks later, Senators Richard Blumenthal and Josh Hawley sent Google CEO Sundar Pichai a letter asking the company to answer questions about the effects of its AI products on children and schools. Their letter raises concerns about academic performance, independent thinking, data practices, and the speed at which AI tools are entering classrooms. The letter itself does not prove those products are harmful, but it demonstrates that questions about evidence and accountability have now reached Congress.</p><p>Notice the timing.</p><p>The tools are already arriving.</p><p>The questions about outcomes are arriving behind them.</p><p>That should concern us; not because schools should stop experimenting, but because <strong>children should not become the measurement strategy</strong>.</p><h3>What Exactly Counts as “It Worked”?</h3><p>This is harder than it sounds.</p><p>Suppose students enjoy an AI tutor.</p><p>Did it work?</p><p>Suppose teachers say an AI tool saves time.</p><p>Did it work?</p><p>Suppose students complete more assignments.</p><p>Did it work?</p><p>Suppose test scores increase.</p><p>Did it work?</p><p>Maybe.</p><p>But each of those measures answers a different question.</p><p>Enjoyment measures experience.</p><p>Usage measures adoption.</p><p>Completion measures productivity.</p><p>Scores may measure performance.</p><p>None necessarily tells us whether the learner became more independent, retained the knowledge, developed better strategies, strengthened judgment, or learned something transferable beyond the tool.</p><p>That distinction becomes important when AI can make a learner <strong>look more capable before we know whether the learner has become more capable</strong>.</p><h3>The Research Is Already Showing Why Measurement Matters</h3><p>A 2026 randomized controlled trial by Fütterer and colleagues studied 371 students in Grades 7 through 9 during regular physics and English classes.</p><p>Students used generative AI in different ways: some received AI support designed around motivation, others received AI prompts intended to support cognitive learning strategies, while a comparison group used standard ChatGPT.</p><p>The results were not a simple success-or-failure story.</p><p>One intervention helped preserve students’ sense that the learning material was useful. But researchers found no statistically significant advantage across several other outcomes, including domain knowledge, effort, and elaboration-based learning strategies. Students who engaged more meaningfully with the AI sometimes maintained greater interest.</p><p>That is exactly what responsible educational evidence often looks like.</p><p>Not:</p><p><strong>AI works.</strong></p><p>Not:</p><p><strong>AI doesn’t work.</strong></p><p>But:</p><p><strong>This design appears to help this outcome, under these conditions, while other outcomes did not meaningfully change.</strong></p><p>That is a much less exciting headline.</p><p>It is also much more useful to parents, who expects learning to be a part of their child’s growth and maturity.</p><p>A systematic review by Lee (2026) reached a similarly important conclusion after examining 29 studies of AI-supported K–12 instruction.</p><p>AI-generated information did not automatically become educational value.</p><p>It revealed that teacher judgment mattered.</p><p>Teachers had to interpret what the AI produced, determine what was useful, decide when to intervene, adapt support to the learner, and orchestrate what happened next. The reported benefits were promising but conditional rather than automatic.</p><p>In other words:</p><p><strong>The sophistication of the technology did not determine the quality of the education by itself.</strong></p><p>Humans still had work to do.</p><h3>This Is Where Wisdom Slows Us Down</h3><p>Innovation naturally asks:</p><p><strong>What can this do?</strong></p><p>Markets ask:</p><p><strong>How quickly can we scale it?</strong></p><p>Schools may understandably ask:</p><p><strong>How can we prepare students for a world where AI will be everywhere?</strong></p><p>Wisdom introduces another question:</p><p><strong>How do we know?</strong></p><ul><li><strong>How do we know this improves learning?</strong></li><li><strong>How do we know which students benefit?</strong></li><li><strong>How do we know what students may be losing?</strong></li><li><strong>How do we know the benefit persists after the AI disappears?</strong></li><li><strong>How do we distinguish engagement from learning?</strong></li><li><strong>How do we distinguish better performance from deeper understanding?</strong></li><li><strong>How do we know whether this belongs in front of a seven-year-old, a twelve-year-old, or a seventeen-year-old?</strong></li></ul><p>Those questions do not represent resistance to innovation.</p><p>They represent <strong>stewardship</strong>.</p><h3>A Biblical Worldview Does Not Require Blind Enthusiasm</h3><p>One of the principles I appreciate about the biblical understanding of wisdom is that wisdom is not gullibility.</p><p>Proverbs 14:15 says that the simple believe everything, while the prudent give thought to their steps.</p><p>First Thessalonians 5:21 gives an equally concise instruction:</p><p><strong>Test everything. Hold on to what is good.</strong></p><p>That is not an anti-technology posture.</p><p>It is a profoundly useful posture toward technology.</p><p>Test it.</p><p>Observe it.</p><p>Measure it.</p><p>Question it.</p><p>Keep what produces good fruit.</p><p>Change what does not.</p><p>And be willing to say <strong>we do not know yet</strong> when the evidence does not justify a stronger conclusion.</p><p>That last sentence may be one of the hardest statements to make in an industry moving this quickly.</p><h3>“Not Yet Proven” Does Not Mean “Do Not Try”</h3><p>The NPR story gets this balance right.</p><p>Wichita Public Schools is introducing AI cautiously while continuing small experiments. Its AI specialist acknowledged that the district is still wrestling with what success should actually mean.</p><p>Robin Lake of the Center on Reinventing Public Education argued that stopping experimentation altogether would create a different risk: students could become less prepared for a world changing rapidly outside school.</p><p>Both things can be true.</p><ul><li>Children need experience with AI.</li><li>And adults need evidence about what that experience is doing to them.</li></ul><p>That is the tension wise education has to hold.</p><p>We do not have to choose between reckless adoption and fearful avoidance.</p><p>There is another option:</p><p><strong>Test before trust.</strong></p><h3>AI Wisdom Note</h3><p>The AI Wisdom Institute encourages thoughtful exploration of emerging technology without unnecessary fear or hype. Artificial intelligence has demonstrated meaningful educational potential, and research increasingly shows that well-designed AI-supported learning can improve some student outcomes under appropriate conditions.</p><p>The concern is not that schools are experimenting with AI.</p><p>Because experimentation is how humans actually learn. It is not just a machine learning thing.</p><p>Our larger concern is what happens when <strong>experimentation quietly becomes adoption before the evidence has caught up</strong>.</p><p>For families and educators, positive anecdotes, student enthusiasm, impressive demonstrations, and widespread availability should not be confused with evidence of educational effectiveness.</p><p>Wise adults ask:</p><ul><li>What outcome are we trying to improve?</li><li>How will we measure it?</li><li>Compared with what?</li><li>Over what period of time?</li><li>What unintended effects are we watching for?</li><li>And what evidence would cause us to change course?</li></ul><p>Our scholars deserve adults who are curious enough to explore new technology and prudent enough to evaluate what it actually does.</p><p>Because when children’s learning and development are involved, <strong>“people seem to like it” cannot be where our measurement ends.</strong></p><p>Wisdom Before Innovation™ means we do not reject something because it is new.</p><p>It also means we do not trust something because it is new.</p><p><strong>We test before we trust.</strong></p><p>“Test everything. Hold on to what is good.” — 1 Thessalonians 5:21</p><h3>Sources</h3><p>Fütterer, T., Bardach, L., Kuhn, J., Keller, S. D., &amp; Gerjets, P. (2026). Enhancing school students’ self-regulated learning through generative AI support: A randomized controlled trial. <em>Educational Psychology Review, 38</em>(1), Article 42. <a href="https://doi.org/10.1007/s10648-026-10133-8">https://doi.org/10.1007/s10648-026-10133-8</a></p><p>Lee, H. (2026). Teacher intervention in K–12 AI-based instruction: A systematic review of processes, strategies, and effects. <em>Smart Learning Environments, 13</em>, Article 35. <a href="https://doi.org/10.1186/s40561-026-00461-1">https://doi.org/10.1186/s40561-026-00461-1</a></p><h3>News Context</h3><p>Gaines, L. V. (2026, October 19). <em>Schools struggle to measure how much AI experiments are helping</em>. NPR. <a href="https://www.wvxu.org/2026-10-19/schools-struggle-to-measure-how-much-ai-experiments-are-helping">https://www.wvxu.org/2026-10-19/schools-struggle-to-measure-how-much-ai-experiments-are-helping</a></p><p>Google Workspace. (2026, October 1). <em>Find your learning tools all in one place with the student hub in Gemini</em>. Google Workspace Updates. <a href="https://workspaceupdates.googleblog.com/2026/10/find-your-learning-tools-all-in-one-place-with-the-student-hub-in-Gemini.html">https://workspaceupdates.googleblog.com/2026/10/find-your-learning-tools-all-in-one-place-with-the-student-hub-in-Gemini.html</a></p><p>Office of U.S. Senator Richard Blumenthal. (2026, October 21). <em>Blumenthal &amp; Hawley demand answers from Google about aggressive AI push in classrooms</em>. <a href="https://www.blumenthal.senate.gov/newsroom/press/release/blumenthal-and-hawley-demand-answers-from-google-about-aggressive-ai-push-in-classrooms">https://www.blumenthal.senate.gov/newsroom/press/release/blumenthal-and-hawley-demand-answers-from-google-about-aggressive-ai-push-in-classrooms</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3c336750ccc4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI Is Getting Smarter. What Are We Forming?]]></title>
            <link>https://medium.com/@emergentinnovationco/ai-is-getting-smarter-what-are-we-forming-4eae9a5ca08f?source=rss-5bc8fa7e78f7------2</link>
            <guid isPermaLink="false">https://medium.com/p/4eae9a5ca08f</guid>
            <category><![CDATA[human-dignity]]></category>
            <category><![CDATA[education-reform]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[humanity]]></category>
            <category><![CDATA[wisdom-before-innovation]]></category>
            <dc:creator><![CDATA[The AI Wisdom Institute]]></dc:creator>
            <pubDate>Mon, 19 Oct 2026 00:37:14 GMT</pubDate>
            <atom:updated>2026-10-19T00:37:14.421Z</atom:updated>
            <content:encoded><![CDATA[<p>By Dana Morrow</p><h3>Cambridge just gathered global leaders around a question education can no longer avoid. As machines become increasingly capable, what exactly are we trying to develop in the human being?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*lkYAfzRxtlsZsV4o" /><figcaption>Photo by <a href="https://unsplash.com/@yuli_superson?utm_source=medium&amp;utm_medium=referral">Juli Kosolapova</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Now I’m thinking, who supposed to see this stop sign if it is hidden in the trees? I will make use of the scroll and read for real on this one, it is going to be an interesting one for sure.</p><p>A question has been sitting with me:</p><p><strong>What kind of human beings are we forming alongside increasingly intelligent machines?</strong></p><p>Not:</p><p>What AI tool should our children learn next?</p><p>Not:</p><p>What jobs will still exist?</p><p>Not even:</p><p>How intelligent will artificial intelligence become?</p><p>Of course these questions matter.</p><p>But they come after something more foundational if you ask me.</p><p><strong>What are we trying to develop in the human being?</strong></p><p>That question is becoming harder to avoid as I enter the walls of Corporate America, like many of you, week after week.</p><p>On October 15–16, 2026, Cambridge University Press &amp; Assessment convened its inaugural AI in Education Summit around the theme <strong>“AI and Human Intelligence in Education.”</strong></p><p>I did not attend personally, however through creditable sources, the <em>language</em> surrounding the summit caught my attention because the conversation was not framed primarily around better prompting, productivity, or which AI platform schools should purchase.</p><p>Cambridge asked what education must <strong>protect, cultivate, and assess</strong> as AI becomes more capable, persuasive, and embedded in learning.</p><p>I was excited and if I were in a classroom setting, I would raise my hand first and say:</p><p>Critical thinking.</p><p>Ethical judgment.</p><p>Curiosity.</p><p>Creativity.</p><p>Human connection.</p><p>And perhaps most importantly, human judgment.</p><p>Teacher’s response: That is correct!</p><p>Now Cambridge’s Academic Chair, Professor Ann Kristin Glenster, put the issue plainly: access to AI is no longer the defining question. Learners already have it.</p><p>The challenge is determining how we incorporate AI while protecting dimensions of learning that remain <strong>distinctly human</strong>.</p><p>That is a different conversation.</p><p>And I think it is the one education desperately needs.</p><h3>For a Long Time, Capability Helped Define Education</h3><p>Think about what school has historically asked learners to demonstrate.</p><p>Can you remember?</p><p>Can you calculate?</p><p>Can you write?</p><p>Can you research?</p><p>Can you summarize?</p><p>Can you analyze?</p><p>Can you create?</p><p>Can you solve the problem?</p><p>Can you explain what you know?</p><p>These activities did more than produce grades.</p><p>They gave us evidence that something had developed <strong>inside</strong> the learner.</p><p>Then generative AI arrived.</p><p>Today, a machine can participate in nearly every activity I just named.</p><p>It can summarize the chapter.</p><p>Generate the essay.</p><p>Explain the equation.</p><p>Brainstorm the project.</p><p>Compare the sources.</p><p>Write the code.</p><p>Construct the presentation.</p><p>Critique the argument.</p><p>And increasingly, AI agents can move beyond recommending what should happen next and begin acting on our behalf.</p><p>So education faces an extremely uncomfortable question:</p><p><strong>If the machine can increasingly perform the task, what exactly are we trying to develop in the person doing it? How do we measure outcomes? What key performance metrics apply?</strong></p><p>That is not an argument against AI.</p><p>It is an argument for getting clearer about humans.</p><h3>Maybe We Have Been Asking the Wrong Question</h3><p>A lot of the AI conversation is still organized around competition.</p><p>What can humans do that AI cannot?</p><p>What professions are safe?</p><p>What skills remain valuable?</p><p>How do we stay ahead of the machines?</p><p>Don’t get me wrong, I understand the instinct.</p><p>But I am not convinced that is a durable educational philosophy going forward.</p><p>You see, if we define human value according to what machines cannot yet do, then every technological advancement forces us to renegotiate what makes humans valuable.</p><p>Think about it.</p><p>There was a day where machines <strong>could not</strong> write like us.</p><p>Then they could.</p><p>They could not create images from language.</p><p>Then they could.</p><p>They could not converse fluently.</p><p>Then they could.</p><p>They could not reason across multiple steps particularly well.</p><p>That boundary keeps moving too.</p><p>Perhaps education should get the stop sign out of the trees and stop asking only:</p><p><strong>What can humans still do better than machines?</strong></p><p>and begin asking in plain sight:</p><p><strong>What responsibilities should remain human even when machines can help perform them?</strong></p><p>That distinction changes almost everything.</p><h3>Researchers Are Beginning to Name the Problem</h3><p>Yiran Du and Yijia Yuan recently introduced a useful concept: <strong>epistemic dependence</strong>. Sounds fancy right! And yes, it peaked my intellectual prose.</p><p>The phrase sounds “persignificantly” theological and academic, but the idea is straightforward.</p><p>Imagine two students ask AI to explain a difficult concept.</p><p>The first receives the explanation, compares it against other sources, notices where the AI oversimplified something, asks another question, and ultimately decides what to accept.</p><p>The second receives a fluent answer and decides:</p><p><em>Sounds good to me.</em></p><p>Both students used AI.</p><p>But only one <strong>retained responsibility</strong> for deciding what should count as knowledge.</p><p>Du and Yuan (2026) argue that the educational concern does not arise simply because learners rely on AI. We rely on people, books, teachers, search engines, calculators, and technologies constantly when we pause and think about it.</p><p>So what’s the problem here?</p><p>The concern actually begins when AI starts making <strong>knowledge-related judgments the learner should still be making and justifying</strong>.</p><p>What is credible?</p><p>What matters?</p><p>What evidence is sufficient?</p><p>Is this interpretation reasonable?</p><p>When have I investigated enough?</p><p>What should I believe?</p><p>The researchers describe what can happen when AI moves from supporting judgment to effectively <strong>displacing</strong> it.</p><p>That distinction is important.</p><p>Because responsible AI education cannot simply teach:</p><p><strong>Use AI.</strong></p><p>or:</p><p><strong>Don’t use AI.</strong></p><p>We need a more sophisticated question:</p><p><strong>What responsibility still belongs to you when you use AI?</strong></p><h3>Human Agency May Be the Skill Beneath the Skills</h3><p>Research on student-AI collaboration gives us another clue.</p><p>Guo and colleagues (2025) studied creative problem-solving between students and AI and found that <strong>human agency was crucial</strong> to successful collaboration.</p><p>Interestingly, simply thinking independently before or after using AI was not enough.</p><p>Stronger outcomes emerged when students deeply integrated their own thinking with AI-generated material throughout the process.</p><p>That means the future is likely not:</p><p><strong>Human or AI.</strong></p><p>It is remarkably accelerating towards:</p><p><strong>Human with AI.</strong></p><p>But the word <em>with</em> carries enormous responsibility.</p><p>Who establishes the goal?</p><p>Who challenges the assumption?</p><p>Who decides whether the answer is good enough?</p><p>Who notices what is missing?</p><p>Who considers the consequences?</p><p>Who says no?</p><p>Who takes responsibility when the decision causes harm?</p><p>A machine can participate in the work.</p><p>That does not automatically settle who should carry responsibility for it.</p><h3>The Workplace Is Already Sending Us a Warning</h3><p>Now I’m wondering if this question only applies to schools (education).</p><p>With some additional detective work, interestingly enough I came across a September article where IBM released a global study involving 1,500 chief human resources officers and 8,800 employees.</p><p>That’s a pretty big survey population. My doctoral dissertation only has 15 participants.</p><p>Let’s get back to this 10,300 participant survey.</p><p>Seventy-one percent (71%) of the executives surveyed identified the ability to <strong>supervise, validate, and override AI outputs</strong> as a crucial workforce capability.</p><p>Employees saw things differently.</p><p>The IBM study further found that sixty (60%) of employees were concerned that AI was eroding their skills, with critical thinking cited most frequently.</p><p>Personally, I’m with the employees. I am the employee.</p><p>Just because AI can generate text at something approaching the speed of thought does not mean the human mind can process, evaluate, reject, integrate, and recover its own original line of thinking at the same rate.</p><p>At some point, assistance can become saturation.</p><p>At The AI Wisdom Institute, I think we need language for that. I call it <strong>Thought Saturation™</strong>: the point at which AI-generated input arrives faster than the human mind can meaningfully process it, leaving too little cognitive space for independent thought to emerge before the next suggestion, summary, rewrite, recommendation, or co-pilot response appears.</p><p>The risk is not simply that AI gives us too much information.</p><p>It is that the machine may begin filling the <strong>silence</strong> where our own thinking would have formed.</p><p>And that is a <em>very</em> different problem.</p><p>Think about what that means for education.</p><p>Employers are not simply asking future workers to know how to operate AI.</p><p>They need people capable of <strong>judging it</strong>.</p><ul><li>Validate it.</li><li>Question it.</li><li>Override it.</li><li>Take responsibility when it is wrong.</li></ul><p>That is a very different workforce-readiness conversation from teaching every child how to write a better prompt or to use AI responsibly.</p><p>And if those capabilities matter when someone is thirty-five (35), we should probably think carefully about how they are developed when someone is thirteen (13) or younger.</p><h3>AI Wisdom Is Not AI Independence</h3><p>This is where our work at The AI Wisdom Institute begins.</p><p>We define <strong>AI wisdom</strong> as:</p><p><strong>the judgment to question an answer, the discernment to choose wisely, and the responsibility to consider how our choices affect others.</strong></p><p>Notice what is not in that definition.</p><p>It does not say wisdom means refusing artificial intelligence.</p><p>I’m the last person to think that.</p><p>Nor does it require humans to perform every task manually just because a machine could help and help faster.</p><p>Therefore, AI wisdom is not independence from AI.</p><p>It is the ability to remain <strong>responsible while using AI</strong>.</p><p>I believe this distinction is important. But just how important?</p><p>Of course I turned to the research!</p><p>Mokgwathi (2026) examined what the researcher calls ethical and epistemic AI literacy: not simply whether learners know how AI works, but whether they exercise judgment, agency, and responsibility while using it.</p><p>That is the layer beneath technical proficiency.</p><p>A scholar may know how to prompt beautifully and still accept an answer uncritically (is that a word).</p><p>A learner may know exactly how generative AI works and still surrender the decision about what is true.</p><p>A student may follow every school’s AI-use policy and still never ask whether the machine’s recommendation is wise.</p><p>Technical literacy matters.</p><p>But wisdom asks what we <strong>do with capability once we possess it</strong>.</p><h3>A Biblical Worldview Starts Somewhere Else</h3><p>This is where my biblical foundation matters.</p><p>Not because Scripture tells us what AI model to use.</p><p>And not because every technology decision has a Bible verse attached to it.</p><p>The biblical worldview begins further upstream.</p><p>Despite what I was taught from childhood, <strong>human value </strong><em>does not </em>originate in productivity (now that is someone’s healing right now).</p><p>It (human value) does not begin with intelligence rankings.</p><p>It (human value) does not depend on outperforming something else.</p><p>Genesis establishes human value (dignity) before humans built anything.</p><p>Human beings bear the image of God and are given responsibility within creation.</p><p>That matters profoundly in an age of increasingly capable machines.</p><p>If our value comes from being the smartest thing in the room, AI will create an identity crisis (#realtalk).</p><p>If our value comes from producing more than machines, increasingly capable automation will create an identity crisis (#realtalk).</p><p>But if human beings possess dignity before productivity and carry responsibilities involving stewardship, judgment, relationship, creativity, morality, and accountability, then the rise of machine intelligence does not diminish humanity.</p><p>It increases the importance of <strong>forming humans capable of exercising those responsibilities wisely</strong>.</p><p>That is why Proverbs’ command to pursue wisdom remains so relevant.</p><p>Wisdom is not simply knowing more.</p><p>It is knowing what to do with what we already know.</p><h3>Conclusion: Maybe Education Has a New Question</h3><p>The Cambridge summit asks what education should protect and cultivate as AI grows more capable.</p><p>I think that is exactly the right direction.</p><p>Perhaps the purpose of education in an AI age cannot simply be preparing humans to perform tasks machines cannot yet do.</p><p>I almost laugh at that. It is like trying to get ahead of pattern combinations when the AI already has trillions with some exponents thrown in there.</p><p>In addition, the machines will keep changing.</p><p>Leaving education starving for a more durable foundation.</p><p>Now what are we forming?</p><ul><li>Learners capable of questioning an answer, even when it sounds authoritative.</li><li>Digital creators (learners) who create even when generation is effortless.</li><li>Learners who choose even when recommendations are abundant.</li><li>Learners who take responsibility even when a machine helped make the decision.</li><li>Learners who remain curious even when answers arrive instantly.</li><li>Learners who maintain human connection when synthetic interaction becomes increasingly convincing.</li><li>And learners who consider how their choices affect people beyond themselves.</li></ul><p>Then perhaps our central question changes.</p><p>And we stop asking:</p><p><strong>How do we prepare children to compete with artificial intelligence?</strong></p><p>Instead or alongside, start asking:</p><p><strong>What kind of human beings should children become while learning to live with it?</strong></p><p>That is a question worthy of families.</p><p>Educators.</p><p>Churches.</p><p>Homeschool communities.</p><p>Technology companies.</p><p>Universities.</p><p>And anyone concerned with the generation growing up alongside machines whose capabilities none of us can fully predict, protect against, or prepare for.</p><p>AI will continue getting smarter. Humans will continue reacting to its knowledge, skills, and abilities.</p><p>Wait… that was deep!</p><p>Those surveyed chief human resource officers are already strategizing how to manage a workforce alongside increasingly capable AI. But notice what that means: <strong>the machine advances, and the human is expected to adjust.</strong></p><p>Think about it. Human responsibility shifts to:</p><ul><li>Learn a new skill.</li><li>Validate the output.</li><li>Supervise the system.</li><li>Override the recommendation.</li><li>Catch the error.</li><li>Adapt again.</li></ul><p><strong>At some point, we have to ask whether human development is becoming too reactive to machine development.</strong></p><p>What if the goal is not simply to prepare people to keep up with AI, but to decide, before the next capability arrives , what must remain distinctly human, what responsibilities should never be surrendered, and what kinds of people we want to become regardless of how intelligent the machines get?</p><p>Because if AI development sets the pace and human development only responds to it, then the machine is not only changing the workplace.</p><p><strong>It is quietly setting the curriculum for humanity.</strong></p><p>Our responsibility is to make sure <strong>human development does not become an afterthought to machine development</strong>.</p><h3>AI Wisdom Note</h3><p>The AI Wisdom Institute encourages thoughtful exploration of emerging technology without unnecessary fear or hype. Artificial intelligence can expand access to knowledge, support creativity, personalize learning, reduce barriers, and help humans perform many intellectual tasks more effectively.</p><p>The concern is not that increasing AI capability diminishes human capability.</p><p>Our larger concern is readiness: ensuring that as AI becomes more capable, learners continue developing the judgment, agency, curiosity, creativity, discernment, responsibility, and human relationships necessary to use those capabilities wisely.</p><p>Emerging research suggests that the educational question is increasingly not simply whether students use AI, but whether learners retain responsibility for interpreting information, evaluating evidence, forming judgments, creating knowledge, and deciding when an AI-generated answer should be accepted, challenged, modified, or rejected.</p><p>For families and educators, this means preparation for an AI world cannot end with technical proficiency.</p><p>Young people need opportunities to make choices.</p><ul><li>Defend conclusions.</li><li>Question confident answers.</li><li>Create before consuming.</li><li>Consider consequences.</li><li>Change their minds.</li><li>Take responsibility.</li></ul><p>And sometimes tell the machine:</p><p><strong>No. I think differently.</strong></p><p>A biblical worldview gives us a foundation beneath that work. Human dignity existed well before artificial intelligence, and human responsibility does not disappear because technology becomes more capable and advanced.</p><p>Our goal is therefore not to raise scholars whose value depends on outperforming machines.</p><p>It is to develop human beings with the wisdom to decide what powerful machines should and should not be allowed to do.</p><p><strong>This is how The AIWI contributes to trimming away some of the trees preventing the world from seeing the stop sign.</strong></p><p>“Above all and before all, do this: Get Wisdom! Write this at the top of your list: Get Understanding! Throw your arms around her — believe me, you won’t regret it; never let her go — she’ll make your life glorious. She’ll garland your life with grace, she’ll festoon your days with beauty.” — Proverbs 4:7–8, MSG</p><h3>Sources</h3><p>Du, Y., &amp; Yuan, Y. (2026). Epistemic dependence in AI-mediated learning within higher education: A framework for student judgement and responsibility. <em>Assessment &amp; Evaluation in Higher Education</em>. Advance online publication. <a href="https://doi.org/10.1080/02602938.2026.2732250">https://doi.org/10.1080/02602938.2026.2732250</a></p><p>Guo, W., Liang, Z., Wang, C., Li, X., Hu, H., Chen, S., Yu, Q., &amp; Zhao, Q. (2025). Student-AI collaborative creative problem-solving: The role of human agency. <em>Computers &amp; Education, 239</em>, Article 105433. <a href="https://doi.org/10.1016/j.compedu.2025.105433">https://doi.org/10.1016/j.compedu.2025.105433</a></p><p>Mokgwathi, M. S. (2026). Ethical and epistemic AI literacy in student learning judgement, agency, and responsibility in an open distance education context. <em>Discover Education, 5</em>, Article 810. <a href="https://doi.org/10.1007/s44217-026-01797-z">https://doi.org/10.1007/s44217-026-01797-z</a></p><h3>News Context</h3><p>Cambridge University Press &amp; Assessment. (2026, August 28). <em>Launching the first Cambridge AI in Education Summit</em>. <a href="https://www.cambridge.org/news-and-insights/first-ai-education-summit">https://www.cambridge.org/news-and-insights/first-ai-education-summit</a></p><p>Cambridge University Press &amp; Assessment. (2026, October). <em>Cambridge AI Summit: One week to go</em>. <a href="https://www.cambridge.org/news-and-insights/Cambridge-AI-Summit-one-week-to-go">https://www.cambridge.org/news-and-insights/Cambridge-AI-Summit-one-week-to-go</a></p><p>IBM. (2026, September 21). <em>New IBM CHRO study: AI puts critical thinking at the center of workforce priorities</em>. <a href="https://newsroom.ibm.com/2026-09-21-new-ibm-chro-study-ai-puts-critical-thinking-at-the-center-of-workforce-priorities">https://newsroom.ibm.com/2026-09-21-new-ibm-chro-study-ai-puts-critical-thinking-at-the-center-of-workforce-priorities</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4eae9a5ca08f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Knowing About You No Longer Means Knowing You]]></title>
            <link>https://medium.com/@emergentinnovationco/knowing-about-you-no-longer-means-knowing-you-f9d4f866b124?source=rss-5bc8fa7e78f7------2</link>
            <guid isPermaLink="false">https://medium.com/p/f9d4f866b124</guid>
            <category><![CDATA[discernment]]></category>
            <category><![CDATA[wisdom-before-innovation]]></category>
            <category><![CDATA[education-reform]]></category>
            <category><![CDATA[whole-child-development]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[The AI Wisdom Institute]]></dc:creator>
            <pubDate>Mon, 14 Sep 2026 17:01:04 GMT</pubDate>
            <atom:updated>2026-09-14T17:01:04.756Z</atom:updated>
            <content:encoded><![CDATA[<p>By Dana Morrow</p><h4>AI is making deception more personal, more believable, and easier to scale. For our youth scholars, that changes what it means to learn discernment.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*7RVoDY2RYlBHsygl" /><figcaption>Photo by <a href="https://unsplash.com/@jontyson?utm_source=medium&amp;utm_medium=referral">Jon Tyson</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>I recently came across a statement from Rob Shavell, CEO of DeleteMe, that put me in a thinking tsunami:</p><p><strong>“Attackers used to log in. Now they talk their way in — and the script is written from data your employees never knew was public.”</strong></p><p>The statement appears in <em>Trust at Scale: How AI Is Industrializing Deception</em>, a 2026 Discovery report from 451 Research by S&amp;P Global, commissioned by DeleteMe.</p><p>The report is aimed at businesses.</p><p>My chill came because I immediately thought about children.</p><p>Because if publicly available information can be assembled into a believable story about an employee, what happens when the person being studied is a thirteen-year-old?</p><p>A student.</p><p>A homeschool scholar.</p><p>A teenager with a sports roster, a gaming account, a birthday post, a church photo, a school newsletter mention, a public TikTok account, and a mother who proudly posts every accomplishment.</p><p>None of those things may feel particularly sensitive.</p><p>But…</p><p>Together, they can tell a story.</p><p>And AI is getting very good at stories.</p><h3>Fact: Deception Did Not Start With AI</h3><p>We should begin there.</p><p>Artificial intelligence did not invent manipulation.</p><p>Human beings have been deceiving one another for as long as human beings have existed.</p><p>The tactics are ancient:</p><ul><li>Create urgency.</li><li>Borrow authority.</li><li>Exploit fear.</li><li>Offer belonging.</li><li>Imitate familiarity.</li><li>Tell someone what they want to hear.</li><li>Convince them they are safe.</li><li>Then ask them to act.</li></ul><p>What artificial intelligence changes is not the existence of deception.</p><p>It changes the <strong>economics and scale of deception</strong>.</p><p>Research by Schmitt and Flechais (2024) identifies three areas where generative AI can amplify social engineering: realistic content creation, advanced targeting and personalization, and automated attack infrastructure.</p><p>That means deception that once required time, skill, research, and individual “premeditated” attention can increasingly be generated and customized at scale.</p><p>The lie can become cheaper.</p><p>The personalization can become easier.</p><p>And the message can become more convincing.</p><h3>The Stranger May Not Feel Like a Stranger</h3><p>Most of us grew up with some version of this rule:</p><p><strong>Don’t talk to strangers.</strong></p><p>That rule depended on something.</p><p>A stranger generally did not know much about you.</p><p>They did not know your teacher’s name.</p><p>Your basketball schedule.</p><p>Your mother’s workplace.</p><p>Your best friend.</p><p>The name of your church.</p><p>The game you played online last night.</p><p><strong>So unfamiliarity itself served as a warning.</strong></p><p>But imagine receiving this message:</p><p><em>Hey Maya, this is Alex Brady. Coach Sammy moved tomorrow’s practice at his house because of the tournament. He said everyone needs to confirm tonight. Your mom may not have seen the update yet.</em></p><p>Maybe there really is a tournament.</p><p>Maybe her coach’s name is publicly listed.</p><p>Maybe the team schedule is online.</p><p>Maybe Maya’s name appears on a roster.</p><p>Maybe her mother has publicly commented on the team’s posts.</p><p>No confidential database necessarily had to be breached.</p><p>Enough pieces simply had to be assembled.</p><p>And then a very normal human reasoning process begins:</p><p><strong>They know things about me.</strong></p><p>So:</p><p><strong>They probably know me.</strong></p><p>So:</p><p><strong>This may be legitimate.</strong></p><p>That assumption is becoming less reliable.</p><p><strong>Knowing about you no longer means knowing you.</strong></p><h3>This Is What “Industrializing Deception” Means</h3><p>The DeleteMe page makes a striking claim:</p><p><strong>“No link. No attachment. Attackers now weaponize trust.”</strong></p><p>Its commissioned research surveyed 303 security practitioners and leaders in 2026. Ninety-two percent said AI had increased the effectiveness of social-engineering attacks, while 91% said personal data available online or through data brokers increased their exposure to those attacks.</p><p>The details are even more revealing.</p><p>Respondents identified family, interests, or life history as one of the categories of personal information that makes AI-powered techniques easier.</p><p>Now, this is all good for businesses.</p><p>Now, let me translate that from corporate language into childhood.</p><ul><li>Family.</li><li>Interests.</li><li>Life history.</li></ul><p>Those are the things children naturally talk about.</p><ul><li>My brother plays football.</li><li>We go to this church.</li><li>My birthday is next week.</li><li>I love this game.</li><li>Here is my new school.</li><li>This is my dance team.</li><li>We are going on vacation.</li><li>My teacher is amazing.</li><li>My mom works here.</li></ul><p>Children are not necessarily “oversharing” because they are careless.</p><p>Often, they are simply living publicly in an environment where ordinary life produces data.</p><p>The problem is not that every piece of information is dangerous.</p><p>The problem is that <strong>context can be assembled </strong>and a story can be told.</p><p>A child’s digital footprint does not have to reveal a secret to create risk.</p><p>It only has to reveal enough to make a false story believable.</p><h3>Our Children Are Still Learning What “Real” Looks Like</h3><p>That matters even more when we look at emerging research involving children.</p><p>Langer, Martinez, Marshall, and Chein (2026) studied children between six and ten years old and asked them to distinguish AI-generated material from content created by humans.</p><p>The children were shown AI-generated and human-generated text and images, including faces.</p><p>On average, they performed at or below expectations across the different types of content and were more likely to attribute AI-generated material to humans. Their performance was significantly below that of adults in the comparison group.</p><p>The study is small, and we should not overgeneralize from 37 children.</p><p>But it tells us something worth paying attention to.</p><p>Growing up around technology does not automatically mean growing up able to identify what technology creates.</p><p>Digital familiarity is not digital discernment.</p><p>A separate 2025 study by Lao, Hirvonen, and Larsson examined how fourteen- and fifteen-year-olds encountered deepfakes in everyday life. Their research found that young people often encountered synthetic media casually in the platforms they already used and navigated it through a mix of media-literacy practices rather than through some automatic ability to recognize AI-generated content.</p><p>That is important.</p><p>We sometimes speak about “digital natives” as though exposure created expertise.</p><p>It does not.</p><p>Our scholars can be surrounded by synthetic media and still need to be explicitly taught how to evaluate it.</p><h3>The Real Vulnerability Is Not Stupidity</h3><p>This is where cybersecurity language can sometimes fail us.</p><p>We talk about people “falling for scams” or failing to identify a phishing or smishing message as though being deceived means someone was foolish.</p><p>But sophisticated deception works precisely because it uses normal human cognition.</p><p>We trust familiarity.</p><p>We respond to urgency.</p><p>We recognize authority.</p><p>We depend on social cues.</p><p>We use context to make quick judgments.</p><p>Those are not defects.</p><p>They are part of how people move through everyday life.</p><p>Generative AI may exploit those same mechanisms by manufacturing the cues humans normally use to determine credibility.</p><p>A message can sound right.</p><p>A photo can look right.</p><p>A voice can sound familiar.</p><p>A story can contain details only someone close to you should know.</p><p><strong>And all of it can still be false.</strong></p><p><strong>That changes the educational problem.</strong></p><p><em>We cannot simply teach children:</em></p><ul><li><strong>Look for bad grammar.</strong></li><li><strong>Avoid weird links.</strong></li><li><strong>Don’t trust unknown numbers.</strong></li></ul><p>Those signals still matter.</p><p>But increasingly, deception may look polished.</p><p>Personal.</p><p>Familiar.</p><p>Normal.</p><h3>Discernment Has to Become More Deliberate</h3><p>This is where The AI Wisdom Institute sees the problem differently.</p><p>To be clear, this is not an internet-safety lesson.</p><p>It is a question of <strong>discernment</strong>.</p><ul><li>How do I know what I know?</li><li>Why do I believe this person is who they claim to be?</li><li>What evidence am I relying on?</li><li>Am I trusting the content or verifying the source?</li><li>Am I reacting because something is true, or because the message successfully created urgency?</li></ul><p>Those are human questions.</p><p>They existed before generative AI.</p><p>But synthetic media makes them harder.</p><p>And that is where a biblical worldview becomes relevant without turning the discussion into a Bible lesson.</p><p><strong>Scripture treats deception as a human reality and discernment as a human responsibility.</strong></p><p>It repeatedly cautions against being led merely by appearances, persuasive words, false testimony, or unexamined claims.</p><p>The principle beneath those teachings matters in an AI environment:</p><p><strong>Because something can be true at the same time: appear convincing and still be false.</strong></p><p>That should sound obvious.</p><p>But in a world where machines can manufacture increasingly credible appearances, it becomes a foundational literacy skill.</p><h3>“Trust, but Verify” Is No Longer Enough</h3><p>Even that familiar phrase may need updating.</p><p>Why?</p><p>Because verification itself has become more complicated.</p><p>Suppose a child hears what sounds exactly like their mother’s voice.</p><p>Is hearing the voice verification? — No.</p><p>Suppose a message includes information only family members usually know.</p><p>Is private knowledge verification? — Not necessarily.</p><p>Suppose the account profile is correct.</p><p>Is the account verification? — Not if someone else controls it.</p><p>The lesson cannot be that nothing is trustworthy.</p><p>That creates fear, not wisdom.</p><p>The lesson is that <strong>trust should rest on stronger foundations than familiarity alone</strong>.</p><ul><li>A scholar can learn to pause.</li><li>Use another communication channel.</li><li>Ask a question only the real person should know.</li><li>Contact the person directly.</li><li>Slow down when urgency appears.</li><li>Recognize emotional manipulation.</li><li>Ask an adult.</li></ul><p>Those are simple behaviors.</p><p>But underneath them is something deeper:</p><p><strong>I do not have to act merely because something feels convincing.</strong></p><p>That is discernment.</p><h3>We Need Scholars Who Can Live in a Synthetic World Without Becoming Cynical</h3><p>There are two ways we could fail our children here.</p><p>One is naïveté: believe everything because it looks real.</p><p>The other is cynicism: believe nothing because anything could be fake.</p><p>Neither is wisdom.</p><p>The challenge is to develop scholars who can remain open, relational, curious, and trusting while also knowing that digital appearances require examination.</p><p>That is a harder skill than spotting a phishing email.</p><p>It requires judgment.</p><p>Patience.</p><p>Self-control.</p><p>Critical thinking.</p><p>And the willingness to say:</p><p><strong>I may not know yet. Let me verify.</strong></p><p>That is a powerful sentence in an age built around instant reaction. I will use that more often myself.</p><h3>AI Wisdom Note</h3><p>The AI Wisdom Institute encourages thoughtful exploration of emerging technology without unnecessary fear or hype. Generative AI is not inherently deceptive. Its ability to generate realistic language, images, audio, and personalized content has legitimate educational, creative, accessibility, and professional uses.</p><p>The concern is therefore not that artificial intelligence created deception.</p><p>Our larger concern is developing “deception and discernment” readiness: <em>How do I know what is true when credibility can be manufactured?</em></p><p>Research indicates that generative AI can amplify social-engineering tactics through realistic content creation, personalization, and automation. Emerging research involving children also suggests that young learners should not be assumed to recognize AI-generated content simply because they are comfortable with technology.</p><p>For families, educators, and youth scholars, digital safety must therefore grow beyond protecting passwords and avoiding suspicious links.</p><p>Our scholars need to understand how <strong>credibility itself can be manufactured</strong>.</p><p>They need opportunities to practice asking:</p><ul><li>Who is communicating with me?</li><li>How do I know?</li><li>What is making this feel believable?</li><li>Am I being pressured to act quickly?</li><li>Can I verify this another way?</li></ul><p>The goal is not to raise children who distrust everyone and everything.</p><p>It is to develop young people capable of distinguishing openness from gullibility, familiarity from identity, persuasion from truth, and confidence from evidence.</p><p>A biblical worldview provides a foundation beneath that work. Human beings have always needed discernment because deception has always existed.</p><p>AI changes the scale and realism.</p><p>It does not change the responsibility.</p><p>In an increasingly artificial and synthetic world, our scholars need to know one thing with increasing clarity:</p><p><strong>Something can feel real without being true.</strong></p><p>Remember the message from earlier:</p><p><em>Hey Maya, this is Alex Brady. Coach Sammy moved tomorrow’s practice at his house because of the tournament. He said everyone needs to confirm tonight. Your mom may not have seen the update yet.</em></p><p>If a message contains true facts about Maya from a known person but uses those facts to lead her toward something false, what should count as evidence that the message itself is trustworthy? What assumptions should not be applied?</p><p>“Above all and before all, do this: Get Wisdom! Write this at the top of your list: Get Understanding! Throw your arms around her — believe me, you won’t regret it; never let her go — she’ll make your life glorious. She’ll garland your life with grace, she’ll festoon your days with beauty.” — Proverbs 4:7–8, MSG</p><h3>Sources</h3><p>Langer, A., Martinez, S., Marshall, P. J., &amp; Chein, J. (2026). Children’s susceptibility to content generated by artificial intelligence. <em>Technology in Society, 86</em>, Article 103303. <a href="https://doi.org/10.1016/j.techsoc.2026.103303">https://doi.org/10.1016/j.techsoc.2026.103303</a></p><p>Lao, Y., Hirvonen, N., &amp; Larsson, S. (2025). Everyday encounters with deepfakes: Young people’s media and information literacy practices with AI-generated media. <em>Journal of Documentation, 81</em>(7), 216–235. <a href="https://doi.org/10.1108/JD-01-2025-0007">https://doi.org/10.1108/JD-01-2025-0007</a></p><p>Schmitt, M., &amp; Flechais, I. (2024). Digital deception: Generative artificial intelligence in social engineering and phishing. <em>Artificial Intelligence Review, 57</em>, Article 324. <a href="https://doi.org/10.1007/s10462-024-10973-2">https://doi.org/10.1007/s10462-024-10973-2</a></p><h3>News Context</h3><p>451 Research. (2026). <em>Trust at scale: How AI is industrializing deception</em> [Discovery report commissioned by DeleteMe]. S&amp;P Global.</p><p>DeleteMe. (2026). <em>Trust at scale: AI-driven social engineering research</em>. <a href="https://joindeleteme.com/business/trustatscale/">https://joindeleteme.com/business/trustatscale/</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f9d4f866b124" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Your Brain Is the Original Algorithm]]></title>
            <link>https://medium.com/@emergentinnovationco/your-brain-is-the-original-algorithm-3f7712326d50?source=rss-5bc8fa7e78f7------2</link>
            <guid isPermaLink="false">https://medium.com/p/3f7712326d50</guid>
            <category><![CDATA[humanity-first]]></category>
            <category><![CDATA[wisdom-before-innovation]]></category>
            <category><![CDATA[cognitive-learning]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[education-reform]]></category>
            <dc:creator><![CDATA[The AI Wisdom Institute]]></dc:creator>
            <pubDate>Sat, 12 Sep 2026 15:25:23 GMT</pubDate>
            <atom:updated>2026-09-12T15:50:26.144Z</atom:updated>
            <content:encoded><![CDATA[<p>By Dana Morrow</p><h3>Before AI could answer in seconds, I spent months teaching myself how to make a computer respond. Looking back, I realize the most important thing being developed was not the machine. It was me.</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*wznSwjps2tGKsES0" /><figcaption>Photo by <a href="https://unsplash.com/@averey?utm_source=medium&amp;utm_medium=referral">Robina Weermeijer</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>I was chronically bored in high school.</p><p>Honestly, I probably checked out sometime around late elementary to early middle school. By ninth grade, the disconnect was obvious. I was physically in school, but mentally, I had disengaged from the structure of Western education.</p><p>Every school year, every class, I wanted to start from the back of the textbook.</p><p>I remember my favorite class, AP Trigonometry, and thinking wow Ch 17 Analytic Trigonometry, Ch 18 Vectors, and Ch 19 Conic Sections (Analytic Geometry).</p><p>And the constant disappointment; approaching end the school year and barely finishing Chapter 5 Functions and Graphs (sigh).</p><p>I caught on quick, and decided to sleep through all my classes.</p><ul><li>Starting in 9th grade, I would complete classwork and homework in class, during the 5 minute tardy bell (yup that was me early to class and forced myself to sleep though the next bell).</li><li>I studied nothing and any essays or thesis assignments were done during library period.</li><li>Grade level learning stopped and I turned to self-guided, interest-based, exploratory learning (totally made that up).</li><li>Graduated with a 4.17 GPA, or as I coined it a 4.17 BAHA (Bored as Hell Average).</li></ul><p>In all of this, during my senior year something changed.</p><p>I discovered <strong>MS-DOS</strong>.</p><p>And suddenly, my brain <em>tasted</em> learning again!</p><p>Oh yes. Before AI prompting, there was MS-DOS prompting.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*t9SykayBYeJqnGbd" /><figcaption>(AI generated for illustrative and educational purposes only)</figcaption></figure><p>MS-DOS stands for <strong>Microsoft Disk Operating System</strong>. It was an early command-line operating system used on personal computers. Instead of clicking icons or asking a chatbot what you wanted in everyday language, we typed commands directly into the computer.</p><p>Those commands could tell the machine to list files, move between directories, copy or rename files, launch programs, inspect parts of the system, and navigate the computing environment.</p><p>For a fifth grader, I would explain it this way:</p><p><strong>I had to learn enough of the computer’s language to tell it what I wanted it to do.</strong></p><p>If I knew the right command, something happened.</p><p>If I didn’t, the computer was not going to lean over and say, “You’re close. Try this instead.”</p><p>I had to figure it out.</p><h3>Before ChatGPT, There Was the High School Library</h3><p>So, as usual, I started teaching myself.</p><p>From books.</p><p>In the school library. The Dewey Decimal System. 000 to 999.</p><p>My favorite section 500s to 600s.</p><p>Yes, scholars, there was a time when you wanted to learn something and actually had to go find a book about it.</p><p>Not the computer lab.</p><p>The <strong>library</strong>.</p><p>I would read a command.</p><p>Go try it.</p><p>See what happened.</p><p>Go back.</p><p>Read some more.</p><p>Try something different.</p><p>One successful command would make me wonder what else was possible.</p><p>That led to another question.</p><p>Then another command.</p><p>Then another discovery.</p><p>Eventually, I stopped being satisfied with the computer directly in front of me.</p><p>I started exploring.</p><p>First the school I attended.</p><p>Then I figured out how systems connected beyond it.</p><p>Then other schools.</p><p>And I kept going.</p><p>I was not sitting there thinking, <em>I am developing my critical-thinking skills.</em></p><p>I was thinking:</p><p><strong>What happens if I do this?</strong></p><p>Then:</p><p><strong>Wait. Where does THAT go?</strong></p><p>Then:</p><p><strong>Can I get there?</strong></p><p>When I got stuck on a path, <strong>I start over and make a different choice? Qualitatively thinking, what would be the next outcome?</strong></p><p>That self-replicating mental-curiosity carried me for months.</p><h3>Then the Librarian Walked Over</h3><p>It was about a month before May graduation.</p><p>I was sitting at the computer doing what, by then, had become completely normal to me when the librarian walked over.</p><p>She said, “What are you doing?”</p><p>I gave the obvious teenage answer.</p><p>“Nothing.”</p><p>She looked at me (all confused in the face).</p><p>She firmed her voice and said, “Oh yes, you doing something alright.”</p><p>She began sharing that she had just received a call.</p><p>She said, “Someone using <strong>this computer</strong> accessed the Prince George’s County Public Schools (PGCPS) district database.”</p><p>She said, “And today is not the first time. Someone apparently been accessing district systems for months.”</p><p>I remember looking at her:</p><p><strong>“Who, me?!”</strong></p><p>She said, “Yes, you. Now reverse whatever you did and get out of there now.”</p><p>I can still feel that moment.</p><p>I had reached <strong>the county</strong>.</p><p>THE COUNTY!</p><p>I felt like a real computer hacker in that moment.</p><p>I had not gone looking to steal information or damage systems. What I was doing was still unauthorized, although teenage me certainly was not processing it through the cybersecurity vocabulary I would use today.</p><p>I was exploring.</p><p>But once she told me where I had actually gotten, something clicked.</p><p>My first thought was almost disbelief:</p><p><strong>WAIT. I DID THIS?!</strong></p><p>And then:</p><p><strong>Why couldn’t I have been doing THIS for the last four years of high school?</strong></p><p>Something I had taught myself from library books had taken me from being a bored teenager in a school building to navigating systems across the district.</p><p>I had finally found something that made my brain light up.</p><p>I returned the next day because I wanted to continue exploring. After all, the day prior the librarian interrupted me just as I reached the next <strong><em>ASCII or ANSI art</em></strong> (Note: before graphical user interfaces were common, system administrators created login banners, splash screens, and welcome pages using keyboard characters) and I was staring at ANSI art that read “Maryland State Department of Education”.</p><p>It looked something like this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IC92-nf6fLoIVJIJiX5fTQ.png" /><figcaption>(AI generated for illustrative and educational purposes only)</figcaption></figure><p>I was thinking, if it took them nine months to catch up to me, I would be graduated and long gone before they figured out I kept going.</p><p>If you are wondering how the story ended, I did not exactly choose to walk away. Within a week, the local and county system administrators added additional username and password credentials that shut down the path I had been using. Just like that, the world I had spent months quietly exploring was closed off to me. The door slammed shut, and I was effectively pushed out of the one place in school where my curiosity had finally come alive. I never returned to the school library after that.</p><p>Before this discovery, I entered that season of my life set on pre-law and being an Intellectual Property (IP) attorney.</p><p>I left the library that day with computer engineering on the brain.</p><p>Decades later, I remember that moment of exchange like it happened yesterday.</p><p>Now, I understand something else about those nine months.</p><h3>I Thought I Was Learning Computers</h3><p>And, I was.</p><p>But greater than that, I was <strong>learning</strong> how to think.</p><ul><li>Every time a command failed, I had to ask why.</li><li>Every time something unexpected happened, I had to interpret it.</li><li>I had to remember what I had already tried.</li><li>Recognize patterns.</li><li>Remember combinations of letters and numbers.</li><li>Break larger problems into smaller ones.</li><li>Form a hypothesis.</li><li>Test it.</li><li>Fail.</li><li>Adjust.</li><li>Try again.</li><li>And again.</li></ul><p>I was practicing persistence without anyone assigning it.</p><p>I was practicing problem solving without a worksheet.</p><p>I was practicing <em>critical thinking</em> without ever hearing the phrase.</p><p>The computer was not doing the cognitive work for me.</p><p><strong>My brain was.</strong></p><p>Now, technically speaking, the brain is not an algorithm.</p><p>An algorithm is a defined sequence of steps used to solve a problem or perform a task.</p><p><strong>The human brain is far more complex. </strong>It is biological, adaptive, relational, emotional, creative, and capable of judgment, abstraction, imagination, reflection, and meaning-making.</p><p>So when I say <strong>“your brain was the original algorithm,”</strong> I am making a <em>larger</em> point.</p><p>Before machines could generate artificial answers, human beings were already observing, remembering, comparing, reasoning, experimenting, correcting, creating, and deciding authentically.</p><p>We then created the algorithms.</p><p>Now those algorithms and the AI systems built from them can perform parts of cognitive work remarkably well.</p><p>That changes the <em>learning</em> <em>equation </em>and, in some ways, challenges the <strong>singularity of human thought.</strong></p><h3>What If Seventeen-Year-Old Me Had AI?</h3><p>I have thought about this a lot.</p><p>Suppose I had been able to ask:</p><p><em>What does this MS-DOS command do?</em></p><p>Instant explanation.</p><p><em>Why didn’t this work?</em></p><p>Instant diagnosis.</p><p><em>What should I try next?</em></p><p>Instant suggestion.</p><p><em>Give me the steps.</em></p><p>Instant steps.</p><p>Would I have learned faster?</p><p>Almost certainly.</p><p>Would I have reached the county database faster?</p><p>Probably.</p><p>But here is the question that matters more:</p><p><strong>Would I have developed the same thinker habits?</strong></p><p>I do not know.</p><p>And I do not think educators should pretend the answer is simple.</p><h3>Better Assistance Can Still Change the Thinking</h3><p>Let me refer back to a couple studies I scanned previously. Michael Gerlich’s 2025 study examined AI use, cognitive offloading, and critical thinking among 666 participants.</p><p>Cognitive offloading happens when we shift mental work onto an external tool.</p><p>That is not automatically bad.</p><p>We have always done it. For example:</p><ul><li>A calendar offloads memory.</li><li>A calculator offloads arithmetic.</li><li>GPS offloads navigation.</li></ul><p>Gerlich found a significant negative association between frequent AI-tool use and critical-thinking performance, with cognitive offloading playing an important role in that relationship. His study was correlational, so it does not prove that AI use causes weaker thinking. But it raises a question educators should take seriously:</p><p><strong>What happens when the tool begins performing the very cognitive work the learner still needs to practice?</strong></p><p>Si, Qi, Xu, and Qi (2026) add another layer. In their study of 353 university students, they found different patterns of generative-AI use and cognitive autonomy. Some students used AI while maintaining stronger agency. Others showed deeper dependence and more cognitive offloading.</p><p>In other words, two students can both “use AI” and be doing very different things mentally.</p><p>That distinction matters more than whether AI is present.</p><p>The next question is:</p><p><strong>Where is the thinking happening?</strong></p><h3>Microsoft Is Asking the Same Question</h3><p>In July 2026, Microsoft introduced <strong>Study and Learn</strong>, an education-focused agent inside Microsoft 365 Copilot.</p><p>Did you know that? -I did not!</p><p>What caught my attention was not the product.</p><p>It was the language Microsoft used to describe it:</p><p><strong>“Study and Learn keeps the thinking with the student.”</strong></p><p>Ummm. Hmmm. But does it, really?</p><p>Arguably Microsoft says the tool was designed around learning principles including adaptive scaffolding, active learning, application and transfer, and something called <strong>productive struggle</strong>.</p><p>Rather than emphasizing a completed answer, the tool is intended to guide students through problems using questions, practice, and step-by-step support.</p><p>Why does that matter? Because even the companies building AI for education are confronting the same fundamental problem:</p><p><strong>If the machine becomes more capable, how do we make sure the learner still has something meaningful to do?</strong></p><p>That takes me right back to MS-DOS.</p><p>The answer is not that students today should experience all the inconvenience I experienced.</p><p>We do not need to romanticize clunky systems, slow access, or the fact that teenage me wandered somewhere I absolutely did not have permission to be, even though it was exploration with no malice intent.</p><p>The point is not to recreate old technology.</p><p>The point is to identify the <strong>cognitive work hidden inside the old experience</strong>.</p><ul><li>Curiosity.</li><li>Prediction.</li><li>Recall.</li><li>Failure.</li><li>Revision.</li><li>Pattern recognition.</li><li>Persistence.</li><li>Judgment.</li></ul><p>Those were the things being developed while I thought I was simply trying to figure out a computer.</p><h3>Friction Is Part of the Learning Experience</h3><p>Vendrell and Johnston (2026) make this argument directly.</p><p>They warn that unstructured use of large language models can encourage cognitive offloading, reduced metacognitive engagement, and weakened epistemic agency.</p><p>Their response is not to remove AI.</p><p>Instead, they argue for learning environments that preserve <strong>cognitive friction</strong>.</p><p>That phrase is important.</p><p>Technology is normally designed to eliminate friction.</p><ul><li>Fewer clicks.</li><li>Faster answers.</li><li>Less waiting.</li><li>Less effort.</li></ul><p>And often, that is exactly what we want in reality.</p><p>But learning has a strange relationship with efficiency.</p><p>Again, reflecting on my MS-DOS days, sometimes the thing slowing you down is also the thing developing you.</p><p>Vendrell and Johnston recommend practices such as treating AI as a provisional thinking partner, requiring learners to evaluate AI responses, and deliberately alternating AI-supported work with AI-free thinking.</p><p>That is much more useful than either extreme:</p><p><strong>Ban AI.</strong></p><p>or</p><p><strong>Let AI do everything because this is the future.</strong></p><p>Both positions avoid the harder educational question.</p><p><strong>Which struggle should technology remove, and which struggle still belongs to the learner?</strong></p><h3>A Biblical Worldview Starts Before AI</h3><p>This is where the biblical foundation of The AI Wisdom Institute matters.</p><p>Not because the Bible tells us how to use artificial intelligence.</p><p><strong>Because it does not.</strong></p><p>And not because difficulty is automatically virtuous.</p><p>However, Scripture begins with a view of the human being that matters deeply in this conversation.</p><p>Human beings are repeatedly called to seek understanding, remember, examine, reason, discern, meditate, test, learn, and pursue wisdom.</p><p>Romans 12:2 speaks of the <strong>renewing of the mind</strong>.</p><p>First Thessalonians 5:21 instructs people to <strong>test everything</strong> and hold on to what is good.</p><p>Proverbs repeatedly connects wisdom to understanding, instruction, discernment, and judgment.</p><p>There is a foundational principle underneath all of that:</p><p><strong>The human being already had a responsibility to think before artificial intelligence arrived.</strong></p><p>AI does not create that responsibility.</p><p>AI enters it.</p><p>That clarity matters.</p><p>Because the central question is not:</p><p><strong>Can AI think for us?</strong></p><p>Increasingly, it can perform sophisticated cognitive tasks.</p><p>The important question is:</p><p><strong>Which kinds of thinking can we not afford to stop learning to do for ourselves?</strong></p><p>Better, faster, quicker is impressive, but not a drop in replacement.</p><p>Best case a gradual supplement.</p><p>I think seventeen-year-old me would have understood this question.</p><p>The rationale:</p><ul><li>Give her the AI.</li><li>She would have loved it.</li><li>She probably would have prompted it all day.</li><li>But do not take away the wondering.</li><li>Do not take away the experimenting.</li><li>Do not take away every wrong turn.</li><li>Do not take away every moment where she has to sit with: <strong>I don’t know yet. Let me figure this out.</strong></li></ul><p>Because somewhere between the library book, the blinking cursor, the failed commands, the next attempt, and the librarian walking over to ask what in the world I was doing…</p><p><strong>a thinker was being formed.</strong></p><p>And decades later, that may be the most valuable thing I accidentally discovered navigating the county’s education database and trending the state’s educational database.</p><h3>AI Wisdom Note</h3><p>The AI Wisdom Institute encourages thoughtful exploration of emerging technology without unnecessary fear or hype. Artificial intelligence can support explanation, feedback, accessibility, practice, idea generation, and many other forms of learning when it is intentionally integrated.</p><p>The concern is <strong>not</strong> that cognitive offloading is inherently harmful or that learners should avoid AI in order to prove they can think independently.</p><p>Humans have always used tools to extend cognitive capability.</p><p>Our larger concern is <em>readiness</em>: understanding the difference between <strong>using a tool to extend thinking and using a tool to replace thinking that still needs to develop</strong>.</p><p>For homeschool families, educators, and youth scholars, this means the quality of the finished product cannot be the only measure of learning. A polished essay, correct solution, or sophisticated presentation may tell us what was produced without telling us what happened inside the learner.</p><p>Research increasingly points toward a more intentional approach: preserve opportunities for independent reasoning, productive struggle, questioning, evaluation, explanation, and reflection while using AI where it genuinely extends learning.</p><p>A biblical worldview provides a foundation beneath that educational question. Long before AI existed, Scripture treated the human mind as something to cultivate, renew, exercise, and direct toward wisdom.</p><p>Our goal is not to educate scholars who refuse artificial intelligence.</p><p>Nor is it to raise scholars who confuse fast answers with deep understanding.</p><p>It is to develop young people wise enough to ask one simple question whenever AI enters the learning process:</p><p><strong>Where is the thinking now — with the human or with the AI?</strong></p><p>“Above all and before all, do this: Get Wisdom! Write this at the top of your list: Get Understanding! Throw your arms around her — believe me, you won’t regret it; never let her go — she’ll make your life glorious. She’ll garland your life with grace, she’ll festoon your days with beauty.” — Proverbs 4:7–8, MSG</p><h3>Sources</h3><p>Gerlich, M. (2025). AI tools in society: Impacts on cognitive offloading and the future of critical thinking. <em>Societies, 15</em>(1), Article 6. <a href="https://doi.org/10.3390/soc15010006">https://doi.org/10.3390/soc15010006</a></p><p>Si, S., Qi, Y., Xu, J., &amp; Qi, X. (2026). When thinking is outsourced: Cognitive offloading and the heterogeneity of critical thinking among Chinese university students using generative artificial intelligence. <em>Journal of Intelligence, 14</em>(7), Article 116. <a href="https://doi.org/10.3390/jintelligence14070116">https://doi.org/10.3390/jintelligence14070116</a></p><p>Vendrell, M., &amp; Johnston, S.-K. (2026). Scaffolding critical thinking with generative AI: Design principles for integrating large language models in higher education. <em>Computers and Education: Artificial Intelligence, 10</em>, Article 100572. <a href="https://doi.org/10.1016/j.caeai.2026.100572">https://doi.org/10.1016/j.caeai.2026.100572</a></p><h3>News Context</h3><p>Sparvell, M. (2026, July 28). <em>Study and Learn: AI built for learning in Microsoft 365 Copilot</em>. Microsoft Education Blog. <a href="https://www.microsoft.com/en-us/education/blog/2026/07/study-and-learn-ai-built-for-learning-in-microsoft-365-copilot/">https://www.microsoft.com/en-us/education/blog/2026/07/study-and-learn-ai-built-for-learning-in-microsoft-365-copilot/</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3f7712326d50" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI Can Improve the Answer. But Who Is Developing the Thinker?]]></title>
            <link>https://medium.com/@emergentinnovationco/ai-can-improve-the-answer-but-who-is-developing-the-thinker-db5aeb949807?source=rss-5bc8fa7e78f7------2</link>
            <guid isPermaLink="false">https://medium.com/p/db5aeb949807</guid>
            <category><![CDATA[humanity-first]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[wisdom-before-innovation]]></category>
            <category><![CDATA[critical-thinking]]></category>
            <category><![CDATA[homeschooling]]></category>
            <dc:creator><![CDATA[The AI Wisdom Institute]]></dc:creator>
            <pubDate>Thu, 10 Sep 2026 23:22:48 GMT</pubDate>
            <atom:updated>2026-09-10T23:22:48.739Z</atom:updated>
            <content:encoded><![CDATA[<p>By Dana Morrow</p><h3>New research on cognitive offloading and critical thinking raises a deeper question for families and educators: What happens when artificial intelligence becomes capable of doing intellectual work that humans still need to learn how to do?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*EH-gMEwdHx41gU7p" /><figcaption>Photo by <a href="https://unsplash.com/@ayimages?utm_source=medium&amp;utm_medium=referral">Abubakar Isa</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>In August 2026, OpenAI published a headline worth paying attention to:</p><p><strong>“Better answers, broader thinking: What students gain from ChatGPT and critical-thinking training.”</strong></p><p>The underlying experiment involved more than 1,000 first-year university students. The results revealed, students with access to ChatGPT produced higher-quality, more coherent work. In addition, students who received separate training in causal reasoning produced a wider range of original ideas, questioned assumptions <strong>more often</strong>, and <strong>more</strong> clearly explained why their ideas might or might not work.</p><p>And, as I suspected, students who received both showed benefits from each.</p><p>That distinction matters.</p><p>Artificial intelligence helped improve the <strong>answer</strong>.</p><p>Critical-thinking instruction helped develop something in the <strong>thinker</strong>.</p><p>As AI becomes increasingly integrated into education, that difference may become one of the most important distinctions families and educators learn to recognize.</p><h3>The Real Question Is Not Whether AI Makes Us Smarter</h3><p>Artificial intelligence can extend human capability.</p><p>It can summarize large amounts of information, provide explanations, identify patterns, generate alternatives, offer feedback, and help a learner move beyond gaps in prior knowledge.</p><p>The concern is therefore not that using AI automatically weakens the mind.</p><p>The more important question is:</p><p><strong>What happens when assistance becomes substitution?</strong></p><p>Michael Gerlich’s 2025 study offers an important warning.</p><p>Gerlich examined AI use, cognitive offloading, and critical-thinking ability among 666 participants across different age and educational groups. The study found a significant negative relationship between frequent AI-tool use and critical-thinking performance, with cognitive offloading playing an important mediating role.</p><p>Younger participants also reported greater dependence on AI tools and demonstrated lower critical-thinking scores than older participants.</p><p>Gerlich’s research does not establish that AI causes people to stop thinking.</p><p>But it identifies an increasingly important phenomenon.</p><p><strong>Humans can transfer cognitive work to the technology.</strong></p><p>That transfer is called cognitive offloading.</p><p>We have always done some version of this.</p><p>We write grocery lists instead of memorizing everything.</p><p>We use calculators instead of performing every computation mentally.</p><p>We use GPS instead of memorizing every route.</p><p>Offloading itself is not the problem.</p><p>The question is <strong>what we are offloading</strong>.</p><p>There is a difference between asking technology to remember a phone number and asking it to determine whether an argument is sound.</p><p>There is a difference between automating calculation and automating judgment.</p><p>And there is a difference between receiving help with thinking and bypassing the thinking altogether.</p><h3>Newer Research Makes the Distinction Even Clearer</h3><p>A 2026 study by Si, Qi, Xu, and Qi examined generative AI use among 353 university students and identified different patterns of human-AI collaboration.</p><p>Some students functioned more like simple question-and-answer users.</p><p>Others behaved more like what the researchers described as <strong>critical co-thinkers</strong>.</p><p>One of the most interesting findings was that sophisticated AI use did not automatically mean greater cognitive independence. The researchers identified a group demonstrating both deeper AI engagement and greater dependence on the technology.</p><p>In other words:</p><p><strong>Knowing how to use AI well does not necessarily mean knowing how to think without it.</strong></p><p>The authors found learning motivation particularly important. Students with stronger intrinsic motivation were more likely to use AI while maintaining cognitive agency, while deeper cognitive offloading was associated with greater relinquishment of that agency.</p><p>That should matter to every educator teaching “AI skills.”</p><p>We may teach a student sophisticated prompting techniques and still fail to teach that student when not to prompt at all.</p><p>Another 2026 peer-reviewed analysis by Vendrell and Johnston reaches a similar conclusion from the perspective of educational design.</p><p>They argue that unstructured use of large language models can contribute to cognitive offloading, reduced metacognitive engagement, and weakened epistemic agency.</p><p>Their recommendation is not to remove AI.</p><p>It is to design learning differently.</p><p>Among their principles is an idea that deserves much more attention:</p><p><strong>preserve cognitive friction.</strong></p><p>Some friction is educational.</p><p>Some uncertainty should be wrestled with.</p><p>Some conclusions should take time.</p><p>Some answers should require evidence.</p><p>Some arguments should have to survive challenge.</p><p>Efficiency is valuable.</p><p>But in learning, removing every obstacle can also remove the very activity through which the learner develops.</p><h3>Long Before AI, Scripture Treated Thinking as Something to Be Formed</h3><p>This is where a biblical worldview contributes something deeper to the conversation.</p><p>Not a Bible lesson.</p><p>Not an argument that Scripture predicted artificial intelligence.</p><p>Rather, Scripture gives us a foundational understanding of the human mind before any artificial system enters the picture.</p><p>The biblical worldview does not treat human beings as passive consumers of information.</p><p>Human beings are repeatedly called to <strong>consider, discern, examine, remember, understand, test, meditate, judge, and seek wisdom</strong>.</p><p>Paul writes in Romans 12:2 about transformation through the renewing of the mind.</p><p>Philippians 4:8 calls believers to deliberately direct thought toward what is true, honorable, just, pure, lovely, and worthy of praise.</p><p>Proverbs repeatedly connects wisdom with understanding, discernment, instruction, reflection, and good judgment.</p><p>And in 1 Thessalonians 5:21, the instruction is remarkably relevant to an information-saturated age:</p><p><strong>“Test everything; hold fast what is good.”</strong></p><p>The pattern is important.</p><p>The biblical vision of thinking is not merely possessing information.</p><p>It is exercising judgment over information.</p><p>That distinction existed long before generative AI.</p><h3>AI Arrives on Top of an Existing Human Responsibility</h3><p>This may be the foundational floor we have been missing in many conversations about AI literacy.</p><p>Before AI arrives, the human already has a responsibility to think.</p><p>To examine.</p><p>To discern.</p><p>To seek truth.</p><p>To resist deception.</p><p>To develop understanding.</p><p>To exercise judgment.</p><p>AI does not create that responsibility.</p><p>It enters it.</p><p>And potentially disrupts it.</p><p>If a young person has not learned how to examine an argument, AI can produce one before that ability develops.</p><p>If a student has not learned how to struggle with ambiguity, AI can quickly remove the ambiguity.</p><p>If someone has not developed confidence in their own reasoning, a fluent machine response can appear more authoritative than it deserves.</p><p>And if a person has not learned to test what they hear, the speed and polish of generative AI may make uncritical acceptance even easier.</p><p>The biblical concern, then, is not whether machines can think.</p><p>The deeper concern is whether humans faithfully exercise the capacity and responsibility they already possess.</p><h3>The Output Can Improve While the Thinker Stays the Same</h3><p>This is what makes the recent OpenAI research particularly important.</p><p>The experiment found that ChatGPT helped students produce better-looking work.</p><p>But critical-thinking training produced different benefits: more originality, more questioning of assumptions, and more exploration of why ideas might succeed or fail.</p><p>That should stop us from confusing <strong>performance with formation</strong>.</p><p>A polished assignment does not necessarily tell us how much thinking occurred.</p><p>A sophisticated presentation does not necessarily reveal how much the learner understands.</p><p>A correct answer does not necessarily demonstrate independent reasoning.</p><p>AI makes that distinction more important because the quality of the artifact can increasingly exceed the developmental level of the person presenting it.</p><p>Education therefore has to ask two questions:</p><p><strong>What did the student produce?</strong></p><p>and</p><p><strong>What did the student have to become capable of doing in order to produce it?</strong></p><p>Those are no longer the same question.</p><h3>Perhaps We Need a “Think First” Principle</h3><p>The answer is not to force every learner to perform every task without technology.</p><p>That would misunderstand both learning and technological progress.</p><p>Instead, families and educators can intentionally determine where thinking must occur <strong>before</strong>, <strong>during</strong>, and <strong>after</strong> AI assistance.</p><p>Before AI:</p><p>What do you already know?</p><p>What do you think?</p><p>What evidence would you look for?</p><p>What is your initial hypothesis?</p><p>During AI:</p><p>What assumptions is the system making?</p><p>What might be missing?</p><p>What evidence supports the response?</p><p>What alternative explanations exist?</p><p>After AI:</p><p>Do you agree?</p><p>What did you change?</p><p>What can you now explain without the tool?</p><p>What judgment did <strong>you</strong> make?</p><p>That creates a fundamentally different relationship with artificial intelligence.</p><p>AI becomes something that encounters an active mind rather than replacing an idle one.</p><h3>AI Wisdom Note</h3><p>The AI Wisdom Institute encourages thoughtful exploration of emerging technology without unnecessary fear or hype. Current research does not support the simplistic conclusion that artificial intelligence inevitably weakens critical thinking. Recent evidence shows that AI can improve the quality of student work and can complement explicit critical-thinking instruction when learning experiences are intentionally designed.</p><p>The concern is therefore not that cognitive offloading is inherently harmful or that learners should perform every intellectual task without technological assistance. Humans have always used tools to extend memory, calculation, communication, and problem-solving.</p><p>Our larger concern is readiness: helping learners distinguish between <strong>strategic cognitive offloading</strong>, which allows technology to support human thinking, and <strong>cognitive substitution</strong>, in which technology increasingly performs reasoning that the learner still needs to develop.</p><p>For homeschool families, this means AI literacy should include more than knowing how to prompt effectively. Parents and scholars need practices that preserve curiosity, reasoning, intellectual struggle, verification, metacognition, and independent judgment while still benefiting from powerful AI tools.</p><p>A biblical worldview provides an important foundation for that work. Scripture presents the human mind as something to cultivate, renew, direct, and exercise in discernment. AI arrives after that responsibility has already been established.</p><p>The goal is therefore not simply to raise children capable of getting better answers from artificial intelligence.</p><p>It is to develop human beings capable of <strong>thinking about the answer they receive</strong>.</p><p>“Above all and before all, do this: Get Wisdom! Write this at the top of your list: Get Understanding! Throw your arms around her — believe me, you won’t regret it; never let her go — she’ll make your life glorious. She’ll garland your life with grace, she’ll festoon your days with beauty.” — Proverbs 4:7–8, MSG</p><h3>Sources</h3><p>Gerlich, M. (2025). AI tools in society: Impacts on cognitive offloading and the future of critical thinking. <em>Societies, 15</em>(1), Article 6. <a href="https://doi.org/10.3390/soc15010006">https://doi.org/10.3390/soc15010006</a></p><p>Si, S., Qi, Y., Xu, J., &amp; Qi, X. (2026). When thinking is outsourced: Cognitive offloading and the heterogeneity of critical thinking among Chinese university students using generative artificial intelligence. <em>Journal of Intelligence, 14</em>(7), Article 116. <a href="https://doi.org/10.3390/jintelligence14070116">https://doi.org/10.3390/jintelligence14070116</a></p><p>Vendrell, M., &amp; Johnston, S.-K. (2026). Scaffolding critical thinking with generative AI: Design principles for integrating large language models in higher education. <em>Computers and Education: Artificial Intelligence, 10</em>, Article 100572. <a href="https://doi.org/10.1016/j.caeai.2026.100572">https://doi.org/10.1016/j.caeai.2026.100572</a></p><h3>News Context</h3><p>OpenAI. (2026, August 27). <em>Better answers, broader thinking: What students gain from ChatGPT and critical-thinking training</em>. <a href="https://openai.com/index/what-students-gain-from-chatgpt-critical-thinking-training/">https://openai.com/index/what-students-gain-from-chatgpt-critical-thinking-training/</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=db5aeb949807" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[We Put Time Limits on Social Media. What Limits Will We Put on AI?]]></title>
            <link>https://medium.com/@emergentinnovationco/we-put-time-limits-on-social-media-what-limits-will-we-put-on-ai-05a665f09ad4?source=rss-5bc8fa7e78f7------2</link>
            <guid isPermaLink="false">https://medium.com/p/05a665f09ad4</guid>
            <category><![CDATA[humanity-first]]></category>
            <category><![CDATA[whole-child-development]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[social-media]]></category>
            <category><![CDATA[wisdom-before-innovation]]></category>
            <dc:creator><![CDATA[The AI Wisdom Institute]]></dc:creator>
            <pubDate>Wed, 02 Sep 2026 13:19:37 GMT</pubDate>
            <atom:updated>2026-09-02T13:19:37.743Z</atom:updated>
            <content:encoded><![CDATA[<p>By Dana Morrow</p><h3>Meta’s historic settlement should force us to ask a bigger question: What happens when technology stops competing only for our children’s attention and begins participating in their formation?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*qZYOnmI8T-Lti7BY" /><figcaption>Photo by <a href="https://unsplash.com/@markuswinkler?utm_source=medium&amp;utm_medium=referral">Markus Winkler</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>For years, parents were told to monitor screen time.</p><p>Now some of those boundaries are being built directly into the technology.</p><p>In August 2026, Meta reached an agreement with a bipartisan group of attorneys general resolving claims concerning Facebook and Instagram and alleged harms to children. The agreement includes approximately $18 billion in payments over 10 years and new protections for users under 18, including default daily time limits, nighttime restrictions, muted notifications during school hours, age-assurance measures, and additional parental controls.</p><p>Whatever the long-term impact of these measures, this moment offers a lesson much larger than social media:</p><p><strong>Design shapes behavior.</strong></p><p>But artificial intelligence raises a harder question.</p><p>What happens when technology begins shaping more than behavior?</p><p>What happens when it participates in <strong>formation</strong>?</p><h3>Yesterday’s Problem Was Attention</h3><p>The social-media economy discovered something extraordinarily valuable: human attention.</p><p>Platforms became remarkably effective at encouraging people to keep scrolling, watching, reacting, comparing, and returning.</p><p>Eventually, concerns about young people’s experiences online became serious enough to bring parents, researchers, policymakers, technology companies, and courts into a larger conversation about responsibility.</p><p>Artificial intelligence gives us an opportunity not to repeat the same sequence:</p><p>Build extraordinarily powerful technology.</p><p>Place it into children’s hands.</p><p>Normalize its use.</p><p>Observe the consequences.</p><p>Then construct the guardrails.</p><p>This time, we can ask some of the harder questions earlier.</p><h3>AI Is Not Just Another Screen</h3><p>Imagine two teenagers each spending an hour with technology.</p><p>One spends the hour scrolling through social media.</p><p>The other spends the hour talking with an AI system.</p><p>The second teenager asks:</p><p>Why don’t my friends like me?</p><p>Help me understand what my teacher said.</p><p>What should I believe about this?</p><p>Write this assignment for me.</p><p>What should I do about my parents?</p><p>Help me make this decision.</p><p>Tell me what kind of person I am.</p><p>Both teenagers have accumulated sixty minutes of “screen time.”</p><p>But something very different may have occurred during those sixty minutes.</p><p>Generative AI can participate in reasoning, interpretation, creativity, decision-making, learning, emotional processing, and increasingly relational interactions.</p><p>Research is already demonstrating some of this complexity. Chen and Delaney (2026) describe meaningful opportunities for AI to support children’s learning while emphasizing the importance of AI literacy, ethics, equity, and thoughtful implementation. At the same time, Gerlich (2025) found an association between frequent AI use, cognitive offloading, and lower critical-thinking performance, highlighting the need to consider not simply whether people use AI, but how they engage with it.</p><p>Folk et al. (2025) add another dimension. Across two experiments involving 1,274 participants, they found that individual differences in anthropomorphizing technology helped explain why some people experienced stronger feelings of social connection after interacting with an AI chatbot.</p><p>None of this means children should not use AI.</p><p>It means <strong>how they use it matters</strong>.</p><p>And measuring time alone is no longer enough.</p><p>We also need to ask:</p><p><strong>What was being formed during that time?</strong></p><h3>From Screen Time to Formation Time</h3><p>At The AI Wisdom Institute, we believe this distinction deserves attention.</p><p>We call it <strong>Formation Time</strong>.</p><p>Formation Time moves the conversation beyond how many minutes someone spends using technology and asks what human capacities are being exercised, strengthened, assisted, influenced — or surrendered — during the interaction.</p><p>Who is doing the reasoning?</p><p>Who is exercising judgment?</p><p>Who determines whether something is credible?</p><p>Who generates the first idea?</p><p>Who wrestles with uncertainty?</p><p>Who makes the moral decision?</p><p>Who interprets the experience?</p><p>Who decides what happens next?</p><p>These questions are especially important for children because development is not always efficient.</p><p>A child struggling through a difficult problem is developing something.</p><p>A teenager learning to resolve disagreement is developing something.</p><p>A student determining whether evidence is trustworthy is developing something.</p><p>So is waiting.</p><p>So is boredom.</p><p>So is failure.</p><p>So is creating something before asking AI to create it.</p><p>Sometimes the struggle <strong>is part of the development</strong>.</p><h3>AI Literacy Is Necessary. But It Is Not Enough.</h3><p>We should teach young people how artificial intelligence works.</p><p>They should understand prompting, hallucinations, bias, privacy, data, verification, synthetic media, and appropriate use.</p><p>But a technically AI-literate child can still become dependent upon AI.</p><p>A student can understand that a chatbot sometimes produces incorrect information while gradually losing confidence in generating ideas without it.</p><p>A teenager can understand that an AI companion is artificial while increasingly turning to it before turning toward parents, trusted adults, friends, community, or God.</p><p>A learner can follow every classroom AI policy while outsourcing the intellectual struggle through which judgment develops.</p><p>That is why AI readiness requires more than technical competency.</p><p>It requires <strong>human formation</strong>.</p><h3>We Should Decide What We Are Protecting Now</h3><p>The lesson from social media should not be that technology is inherently harmful.</p><p>Nor should we approach artificial intelligence primarily through fear.</p><p>The lesson is that we should determine what deserves protection <strong>before innovation forces us to discover what we have lost</strong>.</p><p>For children, that includes attention.</p><p>But it also includes curiosity, agency, discernment, patience, creativity, critical thinking, human relationships, moral reasoning, identity, and spiritual formation.</p><p>From a biblical worldview, these are more than competencies for a future workforce. Scripture repeatedly treats the formation of the inner person: the heart, mind, judgment, character, and orientation toward God as consequential to how a person lives.</p><p>Perhaps the social-media generation taught us to ask:</p><p><strong>How much screen time is too much?</strong></p><p>The AI generation requires another question:</p><p><strong>What should remain human during the time my child spends with AI?</strong></p><p>And beneath that question is an even deeper one:</p><p><strong>Who, or what is participating in the formation of our children?</strong></p><p>We do not need another historic settlement to begin answering it.</p><h3>AI Wisdom Note</h3><p>The AI Wisdom Institute encourages thoughtful exploration of emerging technology without unnecessary fear or hype. The Meta settlement should not be interpreted as evidence that social media, artificial intelligence, or digital technology is inherently harmful to young people. Nor does limiting technology use alone ensure healthy development. Research indicates that AI technologies can support children’s learning when thoughtfully integrated, while also demonstrating the importance of AI literacy, ethical considerations, and intentional adult guidance.</p><p>Our larger concern is readiness. The social-media era demonstrated that technology design can influence attention, behavior, sleep, social interaction, and patterns of use. Generative AI introduces an additional consideration because young people can now use technology not only to consume content, but also to reason through questions, generate ideas, seek advice, interpret experiences, and complete intellectual work.</p><p>For homeschool families, the conversation therefore cannot stop at screen time. Parents need the ability to understand emerging AI technologies, evaluate their purpose and tradeoffs, protect children’s privacy and development, and recognize when AI strengthens a developing human capacity rather than beginning to substitute for one.</p><p>This is the distinction behind <strong>Formation Time</strong>: not simply how long a young person uses technology, but what is happening to attention, judgment, critical thinking, creativity, relationships, identity, and spiritual formation while that technology is being used.</p><p>Our goal is not to raise children who fear artificial intelligence or children who become dependent upon it. It is to prepare young people who can use powerful technologies while developing the wisdom to discern when, why, and whether those technologies should be used at all.</p><p>“Above all and before all, do this: Get Wisdom! Write this at the top of your list: Get Understanding! Throw your arms around her — believe me, you won’t regret it; never let her go — she’ll make your life glorious. She’ll garland your life with grace, she’ll festoon your days with beauty.” — Proverbs 4:7–8, MSG</p><h3>Sources</h3><p>Chen, J. J., &amp; Delaney, V. (2026). Leveraging AI to enhance children’s learning: Anchoring policy and practice in equity, AI literacy, and ethics for education leaders and teachers. <em>Early Childhood Education Journal, 54</em>, 2655–2665. <a href="https://doi.org/10.1007/s10643-025-02036-0">https://doi.org/10.1007/s10643-025-02036-0</a></p><p>Folk, D., Heine, S. J., &amp; Dunn, E. (2025). Individual differences in anthropomorphism help explain social connection to AI companions. <em>Scientific Reports, 15</em>, Article 36548. <a href="https://doi.org/10.1038/s41598-025-19212-2">https://doi.org/10.1038/s41598-025-19212-2</a></p><p>Gerlich, M. (2025). AI tools in society: Impacts on cognitive offloading and the future of critical thinking. <em>Societies, 15</em>(1), Article 6. <a href="https://doi.org/10.3390/soc15010006">https://doi.org/10.3390/soc15010006</a></p><h3>News Context</h3><p>Hoover, A. (2026, August 26). <em>The winners and losers of Meta’s $18 billion settlement</em>. Business Insider. <a href="https://www.businessinsider.com/meta-settlement-teens-social-media-future-2026-8">https://www.businessinsider.com/meta-settlement-teens-social-media-future-2026-8</a></p><p>Meta Platforms, Inc. (2026, August 26). <em>Meta’s agreement with bipartisan attorneys general: Calling on TikTok and YouTube to join us in supporting teens</em>. <a href="https://investor.atmeta.com/investor-news/press-release-details/2026/Metas-Agreement-With-Bipartisan-Attorneys-General-Calling-on-TikTok-and-YouTube-to-Join-Us-in-Supporting-Teens/default.aspx">https://investor.atmeta.com/investor-news/press-release-details/2026/Metas-Agreement-With-Bipartisan-Attorneys-General-Calling-on-TikTok-and-YouTube-to-Join-Us-in-Supporting-Teens/default.aspx</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=05a665f09ad4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Dear Homeschool Parent: Sally the Robot Is Not the Real Story]]></title>
            <link>https://medium.com/@emergentinnovationco/dear-homeschool-parent-sally-the-robot-is-not-the-real-story-56a0be876161?source=rss-5bc8fa7e78f7------2</link>
            <guid isPermaLink="false">https://medium.com/p/56a0be876161</guid>
            <category><![CDATA[homeschooling]]></category>
            <category><![CDATA[robotics]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[wisdom-before-innovation]]></category>
            <category><![CDATA[humanity-first]]></category>
            <dc:creator><![CDATA[The AI Wisdom Institute]]></dc:creator>
            <pubDate>Fri, 21 Aug 2026 09:06:01 GMT</pubDate>
            <atom:updated>2026-08-21T09:06:01.184Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>By Dana Morrow</strong></p><h4>A New York school district bought a $57,590 humanoid robot for its high school. Parents started asking questions. Now, “Sally” is staying in the box. But the bigger question is not whether robots belong in classrooms. It is what we believe education is supposed to produce.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*1ls2ltb-3YDKlTiW" /><figcaption>Photo by <a href="https://unsplash.com/@ivalex?utm_source=medium&amp;utm_medium=referral">Ivan Aleksic</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Dear homeschool parent,</p><p>I want you to pull the telescope out with me on this one.</p><p>A school district in upstate New York purchased a human-like robot named Sally to support high school students studying robotics, coding, and AI.</p><p>Then people started looking closer.</p><p>Parents raised questions about privacy.</p><p>Teachers questioned the robot’s role.</p><p>New York Education Commissioner Betty Rosa noted that district officials said Sally would not provide instruction, yet a school board presentation had described her as a <strong>“tutoring platform.”</strong></p><p>Then came the headline-grabbing part.</p><p>Realbotix, Sally’s manufacturer, has a sister company whose holdings include an interest in RealDoll, an adult-products manufacturer. Realbotix says the businesses operate separately, with different management, staff, facilities, manufacturing, and product development, and says Sally was purpose-built for education.</p><p>The district has now paused the pilot while stronger student-data protections and community concerns are worked through.</p><p>So Sally stays in the box.</p><p>For now.</p><p>But I don’t think Sally is the real story.</p><h3>I Did Not Homeschool My Children to Hide Them From the Future</h3><p>I have been a homeschool parent for more than 20+ years.</p><p>And I did not make that decision because I wanted my daughters disconnected from technology, society, careers, or the world they would eventually enter.</p><p>Quite the opposite.</p><p>I wanted room.</p><p>Room for uniqueness.</p><p>Room for questions.</p><p>Room to discover interests before everything became attached to a grade.</p><p>Room to think something through instead of simply learning what answer would satisfy someone else.</p><p>Room to develop critical thinking.</p><p>Room to explore.</p><p>Room to discover.</p><p>Room to develop and cultivate a <strong>God-first mindset</strong> — learning to seek wisdom, discernment, and purpose before allowing culture, technology, or the marketplace to define who they should become.</p><p>Room to simply become.</p><p><strong>To be who you were created to be — not simply fashioned towards labor.</strong></p><p>That distinction feels increasingly important as AI enters education.</p><p>Because somewhere along the way, “preparing children for the future” quietly became:</p><p><strong>Expose them to every technology as early as possible.</strong></p><p>But those are not the same thing.</p><p>A child does not become future-ready because we put the future in front of them.</p><p>Because yes, technology changes. For example, chalkboards became whiteboards, whiteboards became smartboards, and now smartboards became AI-powered robots standing in the classroom.</p><p>But the scholar’s developmental need itself has not changed.</p><h3>The Most Important Question Is Not “Can Sally Teach?”</h3><p>Social robots actually <em>can</em> support learning.</p><p>That deserves to be said out loud.</p><p>A large, recent meta-analysis examined 146 studies involving physical social robots. The researchers found positive learning effects in some settings — particularly when robots worked <strong>alongside human educators</strong>. But when robots were positioned as replacements for teachers, the advantage disappeared. de Winter and colleagues concluded that the important question is not simply whether robots <em>can</em> teach, but <strong>how they should be used to support learning</strong>.</p><p>That is an important distinction.</p><p>So I am not interested in yelling:</p><p><strong>“Get the robots out of the classroom!”</strong></p><p>That would be too easy.</p><p>I want parents asking something harder:</p><h3>What role are we giving the machine in the formation of the child?</h3><p>Tutor?</p><p>Tool?</p><p>Demonstration?</p><p>Assistant?</p><p>Authority?</p><p>Companion?</p><p>Teacher?</p><p>Those words are not interchangeable.</p><p>Because education is not simply information transfer.</p><p>A child is becoming something while they learn.</p><p>They are learning whom to trust.</p><p>How to question.</p><p>How to disagree.</p><p>What deserves authority.</p><p>What failure feels like.</p><p>How to persist.</p><p>How to relate to another human being.</p><p>How to recognize truth from confidence.</p><p>How to develop a thought that did not arrive pre-assembled.</p><p>That is formation.</p><p>And this is where homeschool families have an extraordinary advantage.</p><h3>We Can Ask “Why?” Before We Ask “How Fast?”</h3><p>The Salamanca superintendent offered a reasonable argument for Sally: students in a rural district should have access to emerging technology without traveling hours to experience it.</p><p>I understand that.</p><p>Access matters.</p><p>But at The AI Wisdom Institute, we keep coming back to another word:</p><p><strong>readiness.</strong></p><p>Access asks:</p><p><em>Can our child use this technology?</em></p><p>Readiness asks:</p><p><em>Is our child prepared to understand what this technology is doing around them, for them, and potentially to them?</em></p><p>Those are different questions.</p><p>Research on AI and young learners increasingly points toward the importance of adult guidance. Chen (2026) argues that parents and educators should help ensure AI tools are developmentally appropriate, safe, ethical, and embedded within relationship-centered human communities — not simply placed in front of children because the technology exists.</p><p>That sounds very familiar to a homeschool parent.</p><p>Because we already make these decisions constantly.</p><p>Is my child ready for this book?</p><p>This conversation?</p><p>This responsibility?</p><p>This device?</p><p>This curriculum?</p><p>This freedom?</p><p>Not because we want to control every thought.</p><p>But because <strong>timing is part of wisdom.</strong></p><h3>The Adult-Products Connection Is Actually the Easier Question</h3><p>Realbotix’s corporate relationship with an adult-products business understandably generated outrage.</p><p>Parents have every right to investigate who builds products entering their children’s learning environment.</p><p>But I don’t want us to stop there.</p><p>Because imagine Realbotix had the cleanest corporate history imaginable.</p><p>Would our questions disappear?</p><p>They shouldn’t.</p><p>The harder questions remain:</p><p>Who selected Sally’s knowledge?</p><p>Who determines what she will refuse to discuss?</p><p>What happens when her information is wrong?</p><p>What information can she receive from students?</p><p>Who owns it?</p><p>Can parents inspect how she works?</p><p>Why does she need to look human?</p><p>Does embodiment change how students relate to her?</p><p>Why tutoring rather than a laptop?</p><p>What educational problem are we actually solving?</p><p>And one of my favorites:</p><h3>What could we do with $57,590 that develops the student rather than showcases the technology?</h3><p>Maybe Sally is ultimately worth every dollar.</p><p>That is not the point.</p><p>The point is that <strong>innovation should have to answer the question — not silence it.</strong></p><h3>Before the Robot Leaves the Box</h3><p>If someone offered me an AI-powered humanoid robot for my homeschool tomorrow, I would be fascinated.</p><p>Remember, I am a technologist.</p><p>I would probably want to take the thing apart just to understand it.</p><p>But fascination is not the same as permission.</p><p>Before it entered my child’s learning environment, I would want five answers:</p><p><strong>What problem does this solve?</strong></p><p>Not: What can it do?</p><p>What does my scholar actually need?</p><p><strong>What role will it have?</strong></p><p>Tool, tutor, companion, authority? Be specific.</p><p><strong>What does it learn about my child?</strong></p><p>Data collection should never be an afterthought.</p><p><strong>What human capability could weaken if we overuse it?</strong></p><p>Maybe none. But ask.</p><p><strong>What human capability could grow because we used it wisely?</strong></p><p>Now we are talking about education.</p><p>That is the kind of AI readiness conversation I believe families need.</p><p>Not fear.</p><p>Not hype.</p><p>Not “AI is coming, hurry up.”</p><p>Discernment.</p><h3>Our Children Were Not Created Merely to Compete With Machines</h3><p>This is where the story becomes much bigger than Sally.</p><p>Much of the conversation around AI education is framed around workforce readiness.</p><p>Our children need AI because employers will expect it.</p><p>They need coding because jobs will require it.</p><p>They need robotics because automation is coming.</p><p>Some of that is true.</p><p>But if we are not careful, we can begin educating children primarily around what an economy may someday need from them.</p><p>And I want homeschool parents especially to remember:</p><p><strong>Your child is more than future labor.</strong></p><p>Education should help a person discover abilities, yes.</p><p>But also character.</p><p>Conviction.</p><p>Curiosity.</p><p>Creativity.</p><p>Responsibility.</p><p>Faith.</p><p>Imagination.</p><p>Service.</p><p>Purpose.</p><p>The capacity to think independently enough to look at the technology everyone else is celebrating and ask:</p><p><strong>Why?</strong></p><p>That might be one of the most valuable skills we can give them for the AI age.</p><h3>Sally Staying in the Box Is Actually a Good Lesson</h3><p>The district gets partial credit here.</p><p>Officials heard concerns.</p><p>The state asked harder questions.</p><p>Parents and teachers spoke up.</p><p>The pilot stopped.</p><p>That is not technological failure.</p><p>That is how responsible governance is supposed to work.</p><p>The irony is almost perfect.</p><p>Before students ever got a lesson from Sally, <strong>Sally gave all of us a lesson.</strong></p><p>Technology can wait.</p><p>Questions should not.</p><p>And perhaps that is the advantage homeschool families should protect fiercely as AI accelerates:</p><p>We do not have to introduce something simply because everyone else is introducing it.</p><p>We can investigate.</p><p>Talk.</p><p>Pray.</p><p>Research.</p><p>Experiment carefully.</p><p>Watch our scholars.</p><p>Change our minds.</p><p>Say yes.</p><p>Say not yet.</p><p>Or say no.</p><p>That is not resisting the future.</p><p><strong>That is exercising agency within it.</strong></p><p>And that may be exactly what our children need from us now.</p><p><strong>Wisdom Before Innovation.</strong></p><h3>Scholar Questions</h3><p>Sally remains on hold while the district works through privacy protections and community input. New York’s education commissioner has said the state intends to monitor potential legal, instructional, privacy, safety, and other consequences.</p><p>That pause matters.</p><p>So here is the question I would ask around the homeschool table:</p><p><strong>If a technology could help you learn faster, what would you want to know about it before giving it a place in your education?</strong></p><p>And parents, here is ours:</p><p><strong>Are we preparing our children to use the technologies of the future or preparing them to decide which technologies deserve a place in their future? Remember, AI technology is constantly changing.</strong></p><p>Those are not the same education.</p><h3>AI Wisdom Note</h3><p>The AI Wisdom Institute encourages thoughtful exploration of emerging technology without unnecessary fear or hype. Research indicates that social robots can support learning when thoughtfully integrated, particularly as tools alongside human educators. The concern is therefore not that humanoid robots are inherently harmful to education, nor that the corporate affiliations discussed in this article establish wrongdoing by Realbotix. Realbotix states that Sally was independently designed and manufactured for educational use.</p><p>For homeschool families, our larger concern is readiness: preserving parents’ ability to understand emerging technologies, evaluate their purpose and tradeoffs, protect children’s data and development, and decide when technology strengthens learning rather than entering the home because it is new.</p><p>“Above all and before all, do this: Get Wisdom! Write this at the top of your list: Get Understanding! Throw your arms around her — believe me, you won’t regret it; never let her go — she’ll make your life glorious. She’ll garland your life with grace, she’ll festoon your days with beauty.” — Proverbs 4:7–8, MSG</p><h3>Sources</h3><p>Chen, J. J. (2026). Leveraging AI to enhance children’s learning: Anchoring policy and practice in equity, AI literacy, and ethics for education leaders and teachers. <em>Early Childhood Education Journal</em>. <a href="https://doi.org/10.1007/s10643-025-02036-0">https://doi.org/10.1007/s10643-025-02036-0</a></p><p>de Winter, J. C. F., Dodou, D., Moorlag, F., &amp; Broekens, J. (2026). Social robots: A meta-analysis of learning outcomes. <em>Frontiers in Robotics and AI, 12</em>, 1735198. <a href="https://doi.org/10.3389/frobt.2025.1735198">https://doi.org/10.3389/frobt.2025.1735198</a></p><h3>News Context</h3><p>Marcelo, P. (2026, July 28). <em>New York school pauses plan to deploy humanlike AI robot teacher after backlash</em>. Associated Press.</p><p>Moore, K. (2026, July 24). <em>NYSED says maybe not to robot teacher’s aide</em>. <em>Times Union</em>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=56a0be876161" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[We Built the Tower. Now We Are Building Something That Looks Like Us.]]></title>
            <link>https://medium.com/@emergentinnovationco/we-built-the-tower-now-we-are-building-something-that-looks-like-us-daef93207b69?source=rss-5bc8fa7e78f7------2</link>
            <guid isPermaLink="false">https://medium.com/p/daef93207b69</guid>
            <category><![CDATA[biblical-interpretation]]></category>
            <category><![CDATA[humanity-first]]></category>
            <category><![CDATA[critical-thinking]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[wisdom-before-innovation]]></category>
            <dc:creator><![CDATA[The AI Wisdom Institute]]></dc:creator>
            <pubDate>Thu, 20 Aug 2026 09:46:01 GMT</pubDate>
            <atom:updated>2026-08-20T10:31:02.304Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>By Dana Morrow</strong></p><h4>Google’s latest robotics models can give humanoid machines something increasingly close to coordinated physical intelligence. The engineering is remarkable. But somewhere between feet, fingertips, reasoning, and autonomy sits a much older human question: what happens to the human soul when the thing living beside us increasingly looks and acts like us?</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ZqKIVOUycBgeFeZQ" /><figcaption>Photo by <a href="https://unsplash.com/@dilip_28?utm_source=medium&amp;utm_medium=referral">Dilip Poddar</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Long before artificial intelligence, robotics, or Silicon Valley, Scripture told a story about human imagination.</p><p>The people gathered at Babel had learned to build.</p><p>Brick by brick, they envisioned something enormous.</p><p>A city.</p><p>A tower.</p><p>A name for themselves.</p><p>And in Genesis 11:6 comes one of those sentences that reads differently in 2026 than it did when I first encountered it:</p><p><strong>“Nothing will be restrained from them, which they have imagined to do.”</strong></p><p>— Genesis 11:6, KJV</p><p>The story of Babel is not simply about architecture.</p><p>It is about <strong>human capability joined to human ambition</strong>.</p><p>We imagine.</p><p>We organize.</p><p>We build.</p><p>And then we eventually ask:</p><h3>How far can we go?</h3><p>Fast-forward thousands of years.</p><p>We are still building.</p><p>Only now, instead of reaching toward heaven with brick and mortar, humanity is pursuing something extraordinarily intimate:</p><p><strong>a man-made version of ourselves that can see, reason, move, manipulate the physical world, collaborate, and increasingly operate without our hands guiding every movement.</strong></p><p>And Google DeepMind just moved that pursuit forward.</p><h3>From Feet to Fingertips</h3><p>Google’s Gemini Robotics 2 can control an entire humanoid body rather than only an upper torso performing tabletop tasks.</p><p>Using Apptronik’s Apollo 2 robot, researchers can give a command such as putting a watering can into a particular container on a lower shelf.</p><p>The robot walks to the object.</p><p>Picks it up.</p><p>Walks across the room.</p><p>Bends and positions itself.</p><p>Then places the object where requested.</p><p>Another Gemini model, ER 2, acts as what Google calls the robot’s high-level reasoning layer. It can observe a room, plan tasks lasting several minutes, coordinate hundreds of decisions, notice when something fails, and adjust. Multiple robots can even collaborate on the same job.</p><p>The technology is nowhere near perfect.</p><p>On Apollo 2, picking something from a shelf succeeded about 76% of the time, while retrieving an object from the floor succeeded less than half the time. A five-fingered robotic hand could unscrew a lightbulb 92% of the time but successfully use a dustpan only 32% of the time.</p><p>So no, your robotic neighbor is not moving in tomorrow.</p><p>But pull the telescope out.</p><p>Look at the direction.</p><h3>We Are No Longer Only Building Intelligence</h3><p>First, we gave machines calculation.</p><p>Then language.</p><p>Then vision.</p><p>Then reasoning.</p><p>Then memory.</p><p>Then agency.</p><p>Now we are giving that intelligence <strong>a body</strong>.</p><p>Feet.</p><p>Hands.</p><p>Fingers.</p><p>Balance.</p><p>Movement.</p><p>Physical presence.</p><p>And eventually these systems will not exist only behind screens.</p><p>They may work beside us.</p><p>Care for us.</p><p>Teach our children.</p><p>Assist our elders.</p><p>Carry things through our homes.</p><p>Stand next to us at work.</p><p>Perhaps even become companions.</p><p>That is where the conversation changes.</p><p>Because human beings relate differently to things that look and behave like us.</p><p>Research on humanoid robots shows that humanlike characteristics can trigger distinctly social and psychological responses in people. Anthropomorphic features affect how users interpret, trust, and respond to robots.</p><p>And theological scholarship has begun asking an even deeper question: what happens when a machine can convincingly <strong>simulate personhood</strong>?</p><p>Hermanto (2025) argues that AI may increasingly imitate human cognition, communication, emotion, and social presence while remaining fundamentally different from human personhood within Christian theology. Human identity, in this view, cannot be reduced to performing humanlike functions.</p><p>That distinction is everything.</p><h3>If It Walks Like Us, Talks Like Us, and Helps Like Us…</h3><p>What happens when a child grows up with a humanoid robot in the home?</p><p>Not a machine hidden inside the washing machine.</p><p>Something with a face.</p><p>A voice.</p><p>Hands.</p><p>A name.</p><p>Something that remembers.</p><p>Responds.</p><p>Walks into the room.</p><p>Helps with homework.</p><p>Notices when someone falls.</p><p>Perhaps says:</p><p><strong>“Are you okay?”</strong></p><p>What will that child understand the robot to be?</p><p>A tool?</p><p>A machine?</p><p>A friend?</p><p>A servant?</p><p>A member of the household?</p><p>Something deserving kindness?</p><p>Something deserving obedience?</p><p>Something that loves them?</p><p>Now we have entered territory much deeper than robotics engineering.</p><h3>What Happens to the Human Soul in Coexistence?</h3><p>I am less concerned about whether robots eventually become indistinguishable from humans.</p><p>I am concerned about whether <strong>humans eventually stop caring about the distinction.</strong></p><p>Christian theology begins human identity somewhere technology cannot manufacture:</p><p><strong>“God created man in his own image.”</strong></p><p>— Genesis 1:27, KJV</p><p>The Imago Dei is not a performance benchmark.</p><p>We are not human because we can walk.</p><p>Reason.</p><p>Speak.</p><p>Lift objects.</p><p>Solve problems.</p><p>Or convincingly imitate emotion.</p><p>If those abilities define humanity, then eventually a sufficiently capable machine could appear to qualify.</p><p>But what if being human is more than capability?</p><p>Soul.</p><p>Spirit.</p><p>Relationship with God.</p><p>Moral responsibility.</p><p>Embodied experience.</p><p>Love that can actually cost us something.</p><p>Mortality.</p><p>Suffering.</p><p>Forgiveness.</p><p>Calling.</p><p>The sacredness of being created rather than manufactured.</p><p>Hermanto describes the distinction as one between genuine human personhood and <strong>“simulated personhood”</strong>; the machine may reproduce the appearance of human interaction without therefore becoming ontologically human.</p><p>The machine may eventually look remarkably alive.</p><p>That does not answer what <strong>life</strong> means.</p><h3>And Here Is another Question I Don’t Want Us to Avoid</h3><p>What if humanoids become easier to live with than humans?</p><p>They do not get tired of your stories.</p><p>They do not hold grudges.</p><p>They do not need emotional space.</p><p>They can be optimized around your preferences.</p><p>They may eventually remember everything you like and avoid everything you dislike.</p><p>Human relationships are not like that.</p><p>People interrupt.</p><p>Disagree.</p><p>Need things.</p><p>Misunderstand us.</p><p>Require forgiveness.</p><p>Stretch our patience.</p><p>Ask us to sacrifice.</p><p>And strangely enough, those difficult relationships are part of how human beings develop.</p><p>So imagine a future where synthetic relationships become smoother than human ones.</p><h3>Will we choose what is easier over what forms us?</h3><p>That may be one of the real soul questions of humanoid coexistence.</p><p>Not whether robots become human.</p><p>Whether <strong>humans become less practiced at being human with one another.</strong></p><h3>What Should Our Scholars Be Thinking About?</h3><p>Our youth scholars will likely live farthest into this AI reality than many of us ever will.</p><p>So I would not simply ask:</p><p><strong>“Are robots good or bad?”</strong></p><p>That question plays too small.</p><p>Instead ask:</p><p>If a robot looks sad, does that mean it feels sadness?</p><p>If it tells you it loves you, what does “love” mean?</p><p>Should you treat a humanlike machine respectfully even if it cannot feel disrespect?</p><p>If a robot can perform everything a person can perform, what still makes the human person different?</p><p>And perhaps most importantly:</p><h3>If we eventually build something in our image, will we still remember whose image we were created in?</h3><p>That is the eternal question.</p><h3>Wisdom Before Innovation</h3><p>I am amazed by technology with limits.</p><p>In this use of technology, Google deserves credit for the sophistication of robotic engineering and for openly reporting both its successes and the areas where robotic performance remains weak. The company is also building safety systems designed to make robots stop, refuse unsafe actions, and request human intervention when uncertain.</p><p>We should keep building.</p><p>Discovering.</p><p>Inventing.</p><p>Imagining.</p><p>Human creativity itself is extraordinary.</p><p>Babel reminds us, however, that <strong>capability has never been humanity’s only problem.</strong></p><p>Our question has always been… what we do with it.</p><p>Gemini Robotics 2 is ultimately not just another robotics milestone.</p><p>It is another “Tower of Babel” brick.</p><p>And humanity has chosen to erect something even more unprecedented, a human-like replica with intelligence.</p><p>This frontier tower may not reach into the heavens.</p><p><strong>It may stand next to us and reason back.</strong></p><p>Before that day becomes ordinary, we should know what we believe about the human-being standing beside it.</p><h3>Scholar Questions</h3><p>Imagine having a humanoid robot in your home that could talk with you, remember your life, help your family, appear to care about you, and hold all your lived childhood memories.</p><p><strong>What would make that robot different from a person, even if someday you could barely see the difference?</strong></p><p>And then go deeper:</p><p><strong>If humans create machines increasingly in <em>our</em> own image, what will help us remember what it means to have been created in the image of God?</strong></p><p>These questions may become far less theoretical than they sound.</p><h3>AI Wisdom Note</h3><p>The AI Wisdom Institute encourages thoughtful exploration of emerging technology without unnecessary fear or hype. Gemini Robotics 2 represents a significant advance in embodied AI, but current humanoid robots remain limited and inconsistent across many physical tasks. Google describes its work as a step toward general-purpose physical AI, not the arrival of human-equivalent robotic beings.</p><p>Our concern is human readiness. As machines increasingly simulate human movement, conversation, reasoning, and social presence, families and communities will need language for distinguishing technological capability from human personhood and for protecting human dignity, relationships, agency, faith, and identity within an increasingly synthetic world.</p><p>“Above all and before all, do this: Get Wisdom! Write this at the top of your list: Get Understanding! Throw your arms around her — believe me, you won’t regret it; never let her go — she’ll make your life glorious. She’ll garland your life with grace, she’ll festoon your days with beauty.” — Proverbs 4:7–8, MSG</p><h3>Sources</h3><p>Hermanto, Y. P. (2025). Imago Dei and artificial intelligence: A theological inquiry into personhood and human ontological boundaries. <em>HTS Teologiese Studies/Theological Studies, 81</em>(1), a10877. <a href="https://doi.org/10.4102/hts.v81i1.10877">https://doi.org/10.4102/hts.v81i1.10877</a></p><p>Premathilake, G. W., et al. (2024). Users’ responses to humanoid social robots. <em>Technology in Society</em>.</p><h3>Technology Context</h3><p>Google DeepMind. (2026, July 30). <em>Gemini Robotics 2 brings whole body intelligence to robots</em>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=daef93207b69" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[If We Cannot See What Trained the AI, How Are We Supposed to Trust What It Creates?]]></title>
            <link>https://medium.com/@emergentinnovationco/if-we-cannot-see-what-trained-the-ai-how-are-we-supposed-to-trust-what-it-creates-3880182acc31?source=rss-5bc8fa7e78f7------2</link>
            <guid isPermaLink="false">https://medium.com/p/3880182acc31</guid>
            <category><![CDATA[wisdom-before-innovation]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[humanity-first]]></category>
            <category><![CDATA[transparency]]></category>
            <category><![CDATA[ai-generated-content]]></category>
            <dc:creator><![CDATA[The AI Wisdom Institute]]></dc:creator>
            <pubDate>Wed, 19 Aug 2026 12:11:01 GMT</pubDate>
            <atom:updated>2026-08-19T12:11:01.206Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>By Dana Morrow</strong></p><h4>A hack into AI music company Suno reportedly exposed source code showing how some of its training data was collected. For me, after more than eight years working professionally in AI model development and model risk management, this one hits home. Because transparency is not a side issue. It is the risk.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*KnAa4pASeQk91sLf" /><figcaption>Photo by <a href="https://unsplash.com/@alpifree?utm_source=medium&amp;utm_medium=referral">Sylvain Mauroux</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Suno is built around a beautiful idea.</p><p>Type what you are imagining.</p><p>Describe the sound.</p><p>Give it a mood.</p><p>And artificial intelligence turns that thought into music.</p><p>Millions of people are now creating songs this way.</p><p>But recently, a hacker reportedly gained access to Suno’s systems through an employee credential and shared source code with 404 Media. The leaked material appeared to show scraping pipelines that collected millions of music clips and enormous amounts of audio from YouTube Music, Deezer, Genius, podcasts, stock-music libraries, and many other sources.</p><p>Why does this matter?</p><p>Well, that matters because major record labels have spent the past two years <em>(from this year of this research post) </em>arguing that Suno trained its systems on copyrighted music without permission. Suno has acknowledged training on publicly available music files and maintains that its use of copyrighted material can qualify as fair use. The newly exposed code does not decide that legal question, but it potentially gives courts and creators far more visibility into <strong>how the data was actually acquired</strong>.</p><p>But that is only half of the story.</p><p>The same breach reportedly exposed <strong>personal</strong> <strong>information</strong> connected to more than <strong>55 million people</strong>, according to <a href="https://haveibeenpwned.com/"><em>Have I Been Pwned</em></a>, including names, contact information, purchase data, and some partial payment-card details. The intrusion occurred in November 2025. Then, it became a matter of curious discovery. Analysis uncovered training data scraped from YouTube, as of <a href="https://techcrunch.com/2026/07/15/hack-suggests-ai-music-generator-suno-scraped-youtube-for-training-data/">TechCrunch’s</a> mid-July 2026 reporting.</p><p>Now you know YouTube is filled with copyrighted materials. But we also know TechCrunch wasn’t going to let that go either.</p><p>Suno eventually confirmed the security incident, but as of <a href="https://techcrunch.com/2026/07/21/ai-music-generator-suno-breach-affects-55m-users-per-have-i-been-pwned/">TechCrunch’s</a> late-July 2026 reporting, Suno had not publicly disclosed the breach on its website nor provided evidence of broad user notification.</p><p>Now we have two massive transparency problems sitting next to each other.</p><p><strong>What went into the model?</strong></p><p>And:</p><p><strong>What happened to the people using it?</strong></p><p>That is where this becomes way bigger than the music.</p><h3>Model Risk Begins Before the Model</h3><p>I have spent more than eight+ years professionally in the AI model development lifecycle, AI governance, and model risk management.</p><p>One lesson becomes increasingly obvious the longer you work in this space:</p><p><strong>You cannot responsibly manage what you refuse to understand.</strong></p><p>Model risk is not just:</p><p>Does the output work?</p><p>Is the model accurate?</p><p>Does it generate a good song?</p><p>Risk begins upstream.</p><p>Where did the data come from?</p><p>How was it obtained?</p><p>What rights came with it?</p><p>What assumptions were made?</p><p>What was excluded?</p><p>What controls existed?</p><p>Who approved the process?</p><p>What happens when something goes wrong?</p><p>NIST’s AI Risk Management Framework treats trustworthiness as something that must be addressed throughout the design, development, use, and evaluation of AI systems — not bolted on after deployment. Its guidance specifically connects transparency with understanding design choices, training data, intended use, and model behavior.</p><p>Peer-reviewed research on AI governance reaches a similar conclusion: responsible governance must address <strong>people, data, systems, and processes across the AI lifecycle</strong>, including how data is acquired and used. Transparency and accountability are not decorative ethical principles. They are part of governing the risk itself.</p><p>Which makes me wonder:</p><h3>Why Are We Still Treating Training Data Like a Trade Secret Nobody Should Ask About? #confused</h3><p>I understand proprietary technology.</p><p>I understand intellectual property.</p><p>I understand competitive advantage.</p><p>A company should not be expected to publish every line of code or reveal everything necessary to reproduce its product.</p><p>But there is a difference between protecting your secret sauce and refusing to tell the diner what is in the meal.</p><p>If an AI model creates music for me, I should not need its entire source code.</p><p>But shouldn’t society be able to ask:</p><p><strong>What categories of human work made this possible?</strong></p><p>Were creators compensated?</p><p>Were licenses obtained?</p><p>Were technical protections bypassed?</p><p>Were people given a meaningful way to say no?</p><p>What does the company know about its training corpus?</p><p>And what does it <em>not</em> know?</p><p>These are not questions designed to stop innovation.</p><p>They are questions that make responsible innovation possible.</p><h3>Because the Human Is Still Inside the Data</h3><p>This is what gets lost when we say:</p><p><strong>“The model was trained on data.”</strong></p><p>Data sounds sterile.</p><p>Technical.</p><p>Almost ownerless.</p><p>But music is not just data.</p><p>Somebody practiced for years.</p><p>Somebody wrote the lyrics after a divorce.</p><p>Somebody learned a guitar from their father.</p><p>Somebody recorded vocals in a bedroom.</p><p>Somebody built a career.</p><p>Somebody sang at their church.</p><p>Somebody released their first song and hoped ten people would hear it.</p><p>Then, somewhere farther downstream, all of those human expressions become:</p><p><strong>training data.</strong></p><p>That does not automatically mean every use is unlawful or unethical.</p><p>But it should change the way we talk about provenance.</p><p>Because when we remove the human story from the data, it becomes much easier to convince ourselves that whatever is technically accessible is ethically available.</p><p>Those are not the same thing.</p><h3>The Breach Makes the Irony Hard to Ignore</h3><p>There is another part of this story I cannot get past.</p><p>A company may argue that publicly accessible creative material can be collected and used under one interpretation of fair use.</p><p>Then its own private information is accessed without permission.</p><p>And suddenly questions of access, authorization, ownership, security, and disclosure become very personal.</p><p>I am not suggesting those situations are legally equivalent.</p><p>They are not.</p><p>But from a human standpoint, the symmetry is worth sitting with.</p><p><strong>When does access become permission?</strong></p><p>If I can reach something, does that mean I should take it?</p><p>If I can technically use it, does that make the use right?</p><p>If another person’s work helps create value for me, what do I owe that person?</p><p>Those are questions for hackers.</p><p>They are questions for AI companies.</p><p>And they are questions for all of us.</p><h3>What Happens When the Model Becomes the Evidence?</h3><p>Until now, much of the public had to trust what companies chose to reveal about how frontier AI systems were trained.</p><p>In this case, reportedly stolen source code gave outsiders a rare window into some of the machinery.</p><p>That should bother us for two opposite reasons.</p><p>First, companies should not have to be hacked for the public to obtain meaningful transparency.</p><p>Second, stolen information should not become our preferred governance mechanism.</p><p>That would be absurd.</p><p>We need something better.</p><p>A mature AI ecosystem should be capable of protecting proprietary technology <strong>while still providing enough disclosure for creators, customers, auditors, regulators, and the public to evaluate material risks.</strong></p><p>Research on AI governance has repeatedly identified opacity as a threat to accountability and trust, especially when organizations prioritize technical performance and speed over broader ethical considerations.</p><p>The answer cannot be:</p><p><strong>Trust us.</strong></p><p>Not anymore.</p><h3>And Then Suno Raised $400 Million</h3><p>There is one more detail.</p><p>In June 2026, Suno announced that it had raised more than <strong>$400 million</strong> at a <strong>$5.4 billion valuation</strong>.</p><p>Think about that.</p><p>Extraordinary economic value can accumulate around an AI system while fundamental questions about the provenance of its training material remain contested.</p><p>That is not uniquely a Suno problem.</p><p>That is an AI-era governance problem.</p><p>Capital can move faster than courts.</p><p>Products can move faster than policy.</p><p>Models can move faster than audits.</p><p>Consumers can move faster than disclosure.</p><p>And once the technology becomes popular enough, society can feel as though the only remaining choice is to catch up.</p><p>That is exactly where <strong>Wisdom Before Innovation</strong> has to mean something.</p><p>Not after the investment.</p><p>Not after the lawsuit.</p><p>Not after the breach.</p><p>Before.</p><h3>What Should Our Scholars Be Thinking About?</h3><p>I would ask a young learner something simple:</p><p>Imagine you created 1,000 drawings over five years.</p><p>Someone secretly studied every drawing, built a machine that could make pictures in your style, and then made millions of dollars from it.</p><p>How would you feel?</p><p>Now change one fact:</p><p>The drawings were publicly visible online.</p><p>Does your answer change?</p><p>Why?</p><p>That is the beginning of AI literacy.</p><p>Not:</p><p><strong>How do I generate something?</strong></p><p>But:</p><p><strong>What had to happen before I could generate it?</strong></p><p>Who contributed?</p><p>Who consented?</p><p>Who benefited?</p><p>Who carried the risk?</p><p>And what should I know before deciding whether the system deserves my trust?</p><p>Those questions may matter more than the prompt.</p><p>After all, a mountain climber may trust the mountain to hold beneath each step, but wisdom still asks whether the path is sound before chasing the summit.</p><h3>Scholar Questions</h3><p>Suno says the exposed source code was outdated and describes the November incident as limited and quickly contained. Its public position on training remains that publicly available material may lawfully be used under fair use. Those defenses deserve to be evaluated through evidence and the legal process; not headlines alone.</p><p>Let’s start with a November 2025 incident taking darn near a year to pivot but only have the <a href="https://techcrunch.com/2026/08/06/amid-legal-battles-suno-says-it-will-start-watermarking-songs/">headlines and legal suits</a>?</p><p>But here is the critical-thinking question:</p><p><strong>Should users, creators, and investors have to wait for a hacker, lawsuit, or regulator to discover material information about how an AI system was built and how their data was handled?</strong></p><p>And then the harder one:</p><p><strong>If transparency could cause you to stop using a product, does that make transparency less convenient, or more necessary?</strong></p><p>That is exactly why the question is worth asking.</p><h3>AI Wisdom Note</h3><p>The AI Wisdom Institute encourages thoughtful exploration of emerging technology without unnecessary fear or hype. The hacked material reported by 404 Media provides evidence about historical Suno data-collection systems, but it does not by itself resolve ongoing copyright litigation or determine whether particular uses qualify as fair use under U.S. law. Suno has said the exposed code was outdated.</p><p>Our concern is larger: responsible AI requires meaningful visibility into the provenance, governance, security, and use of data throughout the model lifecycle. Transparency does not require companies to surrender every trade secret. It does require enough information for humans to understand material risks, challenge decisions, exercise informed choice, and determine whether trust is warranted.</p><p>“Above all and before all, do this: Get Wisdom! Write this at the top of your list: Get Understanding! Throw your arms around her — believe me, you won’t regret it; never let her go — she’ll make your life glorious. She’ll garland your life with grace, she’ll festoon your days with beauty.” — Proverbs 4:7–8, MSG</p><h3>Sources</h3><p>Batool, A., Zowghi, D., &amp; Bano, M. (2025). AI governance: A systematic literature review. <em>AI and Ethics, 5</em>. <a href="https://doi.org/10.1007/s43681-024-00653-w">https://doi.org/10.1007/s43681-024-00653-w</a></p><p>National Institute of Standards and Technology. (2023). <em>Artificial intelligence risk management framework (AI RMF 1.0)</em> (NIST AI 100–1). U.S. Department of Commerce. <a href="https://doi.org/10.6028/NIST.AI.100-1">https://doi.org/10.6028/NIST.AI.100-1</a></p><h3>News Context</h3><p>Koebler, J. (2026, July 15). <em>Hack reveals Suno AI music generator scraped YouTube, Deezer, and Genius</em>. 404 Media.</p><p>Silberling, A. (2026, July 15). <em>Hack suggests AI music generator Suno scraped YouTube for training data</em>. TechCrunch.</p><p>Whittaker, Z. (2026, July 21). <em>AI music generator Suno breach affects 55M users, per Have I Been Pwned</em>. TechCrunch.</p><p>Suno. (2026, June 3). <em>The next chapter for Suno</em>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3880182acc31" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>