<?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 David Hundley on Medium]]></title>
        <description><![CDATA[Stories by David Hundley on Medium]]></description>
        <link>https://medium.com/@dkhundley?source=rss-82498630db6------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*jhm8zA0VcRvYzRXIM4EJJg@2x.jpeg</url>
            <title>Stories by David Hundley on Medium</title>
            <link>https://medium.com/@dkhundley?source=rss-82498630db6------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 14:46:56 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@dkhundley/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[Five Reasons Why AI Will NOT Takeover the Workforce in the Next 5-10 Years]]></title>
            <link>https://dkhundley.medium.com/five-reasons-why-ai-will-not-takeover-the-workforce-in-the-next-5-10-years-222231b857fa?source=rss-82498630db6------2</link>
            <guid isPermaLink="false">https://medium.com/p/222231b857fa</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[genai]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[David Hundley]]></dc:creator>
            <pubDate>Sat, 14 Jun 2025 11:52:40 GMT</pubDate>
            <atom:updated>2025-06-14T11:52:57.540Z</atom:updated>
            <content:encoded><![CDATA[<h4>Analyzing both technical and non-technical reasons why it is naive to believe that AI will take human jobs in the near term future</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8cgkMfCALsObnPYb6X4Ptw.png" /><figcaption>Title card created by the author</figcaption></figure><p>Working in the AI space, one of the top questions I get is, “Will AI takeover my job?” To be completely transparent, I’m not comfortable with answering that question from a long term viewpoint. Shoot, I’m not sure that anybody could have predicted the rise of social media and smartphones in the 1990s, so I’d be naive to say I know how AI will affect the workforce in the next 20+ years.</p><p>What specifically triggered me to write this post is a quote I came across from <a href="https://www.axios.com/2025/05/28/ai-jobs-white-collar-unemployment-anthropic">Anthropic CEO Dario Amodei in a recent interview with Axios</a>. Here is a specific line from that article that is currently making rounds across the internet: “AI could wipe out <em>half</em> of all entry-level white-collar jobs — and spike unemployment to 10–20% in the next one to five years, Amodei told us in an interview from his San Francisco office.”</p><p>At face value, these are pretty frightening numbers! Even a 10% reduction in the workforce would be quite devastating, so that 50% number is downright catastrophic. Amodei is not alone in his thinking. With a quick internet search, you can find many who will agree with this sentiment.</p><p>I personally do not share this opinion. In fact, I’ll go as far to say that I think that anybody who thinks AI will take over even 5% of the workforce in the next 5–10 years is almost laughably wrong.</p><p>But these individuals are wrong for reasons you may not think, and I’m not sure even folks like Dario Amodei are necessarily thinking about this in the right way. Without trying to read Amodei’s mind, it seems to me that he’s approaching this from the perspective of what an AI <em>can</em> do. In other words, I think I would agree with him that the AI technology is probably already intelligent enough to takeover jobs today, but I think where folks like Amodei might be losing the thread is everything else that surrounds what it would take to replace a workforce with AI.</p><p>In the remainder of this post, I’m going to provide some reasoning around what I’m thinking. What you’ll quickly see that an AI workforce takeover has a lot more to consider than just the pure AI capability itself.</p><h3>1. Almost all jobs are privy to some level of sensitive information</h3><p>Consider the onboarding process for almost every corporate company today. Because almost all jobs are privy to some level of sensitive information, HR and legal departments require new employees to do things like sign nondisclosure agreements (NDAs) or go through training on what they are or are not allowed to do with certain sensitive pieces of information.</p><p>The reason this is important to consider is because most companies leveraging AI are dependent on third party model providers like OpenAI or Anthropic to meet their AI needs. This means in order to effectively leverage an AI to replace a human employee, you’re <em>obligated</em> to hand over a ton of sensitive information to the third party model provider. Frankly, that’s a nonstarter for most companies and understandably so. Even when companies do their best in good faith to protect another company’s data, we live in a world where bad actors exist who try to exploit security vulnerabilities to extract sensitive information. It’s unfortunately too common to hear about this or that data leak. I imagine companies like OpenAI and Anthropic are sitting on a gold mine of sensitive information, and if anybody is ever able to successfully able to penetrate their systems, that could be quite catastrophic to many companies.</p><p>What many companies are choosing to do is dipping their toe into the water by easing their way into Generative AI technologies slowly, starting with projects that leverage less sensitive information. I respect and agree with this action, but in order for an AI to effectively takeover a human’s job, it really needs access to all the data it can get. This problem may be alleviated in the future when companies can effectively run AI on their own local hardware, but we are arguably at least a decade away from hardware becoming commoditized that way. Again, until this hurdle is overcome, this reason alone will be the largest nonstarter for companies interested in replacing their human workforce with AI.</p><h3>2. We are still in the very early days of AI legislation</h3><p>At the time I write this post, I am not aware of any major AI legislation that has been passed aside from what the European Union (EU) has done, but almost everybody (including myself) feels like it is bound to happen. We are starting to see inklings of this in proposed bills, particularly where many bills are urging the mandate to inform a human user when they are interacting with an AI.</p><p>Until we more hardened legislation is passed, many companies will take a more conservative approach in how they approach AI. In my opinion, the most non-conservative action you can take with AI is replacing a human’s job. Let’s imagine that we’re able to find solutions or workarounds to every other reason in this post. Even if you can effectively replace a human’s job with AI today, the government could turn around in the next few years and slap your hand for your actions.</p><h3>3. Managing AI memory remains a steep challenge today</h3><p>When I refer to memory here, I’m not talking about computer memory like RAM but rather memory as we understand it from the human perspective. Under the hood, AI largely manifests as statically trained mathematical models. In other words, every time you interact with an AI like ChatGPT, it has no memory of who you are nor your previous interactions. Now you might be thinking, “But when I have a conversation with an AI chatbot, it remembers my conversation history.” Technically speaking, the AI model itself doesn’t remember a thing. Rather, we’ve given the illusion of memory by adding extra engineering behind the scenes.</p><p>For example, let’s say I’m having a long back-and-forth conversation with an AI chatbot. There is the illusion that the AI has a memory of our conversation because I can continue to ask questions where the AI will be able to refer to some previous part of the conversation. But the only reason this is possible is because there is engineering behind the scenes that is constantly taking those previous interactions and feeding them back into the AI model each time you get a new response. Without that extra engineering, the AI model itself has no capability of remembering anything at all.</p><p>As you might imagine, managing memory for a chatbot like ChatGPT is doable, and OpenAI for example has done a pretty effective job at giving the illusion the AI model can remember your interactions with all the engineering they’ve done on their backend. But it’s important to keep in mind that AI models have a limited context window, meaning that I can’t feed in 100% of my data in a single instance so that the AI model has 100% context of everything it needs to take an appropriate action. This is very different from a human being, where we have quite an astounding memory recollection. (Funny story: My kids played the song “Baby” by Justin Bieber in front of me the other day, and even though I probably haven’t heard that song in 10 years, I still knew many of the lyrics by heart. 😂)</p><p>Now try to extrapolate that need for memory to a person’s job. I don’t know how you’d even begin to manage memory at that level. It’s already challenging enough for a standard chatbot to engineer memory in something like ChatGPT, so trying to do that for a person’s job seems like a very challenging task. Again, it’s not impossible, but I don’t see us solving this problem within the next few years.</p><h3>4. Technical debt remains a barrier to effective AI implementation</h3><p>If you’re not familiar with the concept of technical debt, it’s this idea that you have a bunch of existing systems that you have to contend with in order to implement anything new. For example, let’s say I have historically had all my systems on an on-premises environment and want to leverage any capability offered by a cloud service, even non-AI. In order to make use of that cloud service, I might need to migrate a bunch of my data from my on-prem servers to the cloud, which could be quite the challenge for me.</p><p>Technical debt is not a new problem, and it’s one that is especially hard to overcome for AI. Consider the previous point about managing AI memory. If I work in an environment with a bunch of legacy systems and databases, how can I engineer a solution that will effectively integrate with those systems to provide all the context and AI would need to do its job? Again, this is already a hard problem even without AI, so when you try to approach this problem for replacing a human’s job, it becomes an even more challenging task.</p><h3>5. AI can’t play office politics</h3><p>Of all the reasons shared in this post, this one is admittedly the ickiest to think about. There’s a famous quote in the poker world that goes along these lines: “Poker is not about the cards you’re dealt, it’s about how you play the person across from you.” It’s a recognition that we as humans are largely irrational and act on irrational reasons. To be clear, this is not necessarily all bad. Consider having to give a presentation on a complex topic to different kinds of audiences. You may choose to communicate a message to one audience (like a group of interns) versus a different audience (like a group of executives).</p><p>But again, this can quickly get quite icky and potentially even unethical. Consider how you feel about all your coworkers. You may quietly be biased toward liking or disliking an individual for reasons you would never want to admit aloud, let alone write down on paper. So the question becomes, do you want AI analyzing other people’s psyches? Keep in mind the previous point about AI memory: in order to manage AI memory, you have to capture that in the form of hard data. So do you want an AI analyzing another person’s psyche AND getting hard evidence of that in the form of data? I personally cringe even thinking about that. It seems extremely messy, yet in order to effectively replace a human’s job, I’m not sure you can necessarily get around this.</p><p>I hope these reasons help to illustrate how challenging of a matter this can be. Trying to solve all these problems within a 5 to 10 year window seems virtually impossible to me, but I would be naive to say I know what the future looks like beyond that point. As somebody who still probably has 30 years left in the workforce myself, I wouldn’t necessarily say this is something that keeps me awake at night but is something I do ponder a lot, especially as a dad to two little girls. By the time I retire, the world may look very, very different. But what I do feel confident about is this shorter timeframe. Honestly, I’m not sure if this post provides any comfort for anybody except maybe folks who anticipate retiring within the next 10 to 15 years. These reasons may be challenges today, but there is a very real possibility of overcoming all these problems in the longer term future.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=222231b857fa" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Five Things to Consider Before Fine-Tuning a GenAI Model]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/data-science-collective/five-things-to-consider-before-fine-tuning-a-genai-model-32079e384665?source=rss-82498630db6------2"><img src="https://cdn-images-1.medium.com/max/1842/1*GItM2duo9nBkeA7KLCPHgA.png" width="1842"></a></p><p class="medium-feed-snippet">Things to consider before approaching fine-tuning as a final option for your GenAI project</p><p class="medium-feed-link"><a href="https://medium.com/data-science-collective/five-things-to-consider-before-fine-tuning-a-genai-model-32079e384665?source=rss-82498630db6------2">Continue reading on Data Science Collective »</a></p></div>]]></description>
            <link>https://medium.com/data-science-collective/five-things-to-consider-before-fine-tuning-a-genai-model-32079e384665?source=rss-82498630db6------2</link>
            <guid isPermaLink="false">https://medium.com/p/32079e384665</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[llm]]></category>
            <category><![CDATA[genai]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[David Hundley]]></dc:creator>
            <pubDate>Wed, 16 Apr 2025 21:28:02 GMT</pubDate>
            <atom:updated>2025-04-18T03:02:36.313Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Five Engineering Skills Every Data Scientist Should Learn]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/data-science/five-engineering-skills-every-data-scientist-should-learn-e7a347fada6f?source=rss-82498630db6------2"><img src="https://cdn-images-1.medium.com/max/1842/1*P3Q5YloIwDDUBXCuQ5UT_g.png" width="1842"></a></p><p class="medium-feed-snippet">Rounding out tactics to help you stay competitive as a &#x201C;full stack&#x201D; data scientist</p><p class="medium-feed-link"><a href="https://medium.com/data-science/five-engineering-skills-every-data-scientist-should-learn-e7a347fada6f?source=rss-82498630db6------2">Continue reading on TDS Archive »</a></p></div>]]></description>
            <link>https://medium.com/data-science/five-engineering-skills-every-data-scientist-should-learn-e7a347fada6f?source=rss-82498630db6------2</link>
            <guid isPermaLink="false">https://medium.com/p/e7a347fada6f</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[David Hundley]]></dc:creator>
            <pubDate>Tue, 01 Oct 2024 00:34:12 GMT</pubDate>
            <atom:updated>2024-10-01T14:54:11.235Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Using an iPad for Data Science in 2024]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/data-science/using-an-ipad-for-data-science-in-2024-97bd9ef484b8?source=rss-82498630db6------2"><img src="https://cdn-images-1.medium.com/max/1842/1*fylQMULCLmJg2oF83TZ8kA.png" width="1842"></a></p><p class="medium-feed-snippet">Exploring the benefits and pitfalls of using Apple&#x2019;s popular tablet for data science work</p><p class="medium-feed-link"><a href="https://medium.com/data-science/using-an-ipad-for-data-science-in-2024-97bd9ef484b8?source=rss-82498630db6------2">Continue reading on TDS Archive »</a></p></div>]]></description>
            <link>https://medium.com/data-science/using-an-ipad-for-data-science-in-2024-97bd9ef484b8?source=rss-82498630db6------2</link>
            <guid isPermaLink="false">https://medium.com/p/97bd9ef484b8</guid>
            <category><![CDATA[editors-pick]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[ipad]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[David Hundley]]></dc:creator>
            <pubDate>Sun, 01 Sep 2024 15:16:28 GMT</pubDate>
            <atom:updated>2024-09-01T19:35:39.262Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Evaluating RAG Pipelines with Ragas]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/data-science/evaluating-rag-pipelines-with-ragas-5ff28aa27984?source=rss-82498630db6------2"><img src="https://cdn-images-1.medium.com/max/1842/1*pYo7Ab7mxUaIb-biapfXqg.png" width="1842"></a></p><p class="medium-feed-snippet">Leveraging the Ragas framework to determine the performance of your retrieval augmented generation (RAG) pipeline</p><p class="medium-feed-link"><a href="https://medium.com/data-science/evaluating-rag-pipelines-with-ragas-5ff28aa27984?source=rss-82498630db6------2">Continue reading on TDS Archive »</a></p></div>]]></description>
            <link>https://medium.com/data-science/evaluating-rag-pipelines-with-ragas-5ff28aa27984?source=rss-82498630db6------2</link>
            <guid isPermaLink="false">https://medium.com/p/5ff28aa27984</guid>
            <category><![CDATA[raga]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[rags]]></category>
            <category><![CDATA[editors-pick]]></category>
            <category><![CDATA[genai]]></category>
            <dc:creator><![CDATA[David Hundley]]></dc:creator>
            <pubDate>Sun, 30 Jun 2024 18:54:39 GMT</pubDate>
            <atom:updated>2024-06-30T22:07:54.312Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[How to Think About Using Your Company’s Information with GenAI]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/data-science/how-to-think-about-using-your-companys-information-with-genai-e4ba63af38b8?source=rss-82498630db6------2"><img src="https://cdn-images-1.medium.com/max/1842/1*AQ8uvZUNZlI7Rg_uwVB-9w.png" width="1842"></a></p><p class="medium-feed-snippet">Assessing the options for GenAI information integration along with considerations for your specific company</p><p class="medium-feed-link"><a href="https://medium.com/data-science/how-to-think-about-using-your-companys-information-with-genai-e4ba63af38b8?source=rss-82498630db6------2">Continue reading on TDS Archive »</a></p></div>]]></description>
            <link>https://medium.com/data-science/how-to-think-about-using-your-companys-information-with-genai-e4ba63af38b8?source=rss-82498630db6------2</link>
            <guid isPermaLink="false">https://medium.com/p/e4ba63af38b8</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[llm]]></category>
            <category><![CDATA[editors-pick]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[genai]]></category>
            <dc:creator><![CDATA[David Hundley]]></dc:creator>
            <pubDate>Sun, 07 Apr 2024 15:19:29 GMT</pubDate>
            <atom:updated>2024-04-07T15:43:34.834Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[A High Level Guide to LLM Evaluation Metrics]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/data-science/a-high-level-guide-to-llm-evaluation-metrics-fbecd08f725c?source=rss-82498630db6------2"><img src="https://cdn-images-1.medium.com/max/1842/1*v0bJtzzJLR10ZNJEBu8gKQ.png" width="1842"></a></p><p class="medium-feed-snippet">Developing an understanding of a variety of LLM benchmarks &amp; scores, including an intuition of when they may be of value for your purpose</p><p class="medium-feed-link"><a href="https://medium.com/data-science/a-high-level-guide-to-llm-evaluation-metrics-fbecd08f725c?source=rss-82498630db6------2">Continue reading on TDS Archive »</a></p></div>]]></description>
            <link>https://medium.com/data-science/a-high-level-guide-to-llm-evaluation-metrics-fbecd08f725c?source=rss-82498630db6------2</link>
            <guid isPermaLink="false">https://medium.com/p/fbecd08f725c</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[deep-dives]]></category>
            <category><![CDATA[llm]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[David Hundley]]></dc:creator>
            <pubDate>Tue, 27 Feb 2024 03:15:12 GMT</pubDate>
            <atom:updated>2024-02-27T15:56:58.097Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[An Intuition for How Models like ChatGPT Work]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/data-science/an-intuition-for-how-models-like-chatgpt-work-c7f01616bd6d?source=rss-82498630db6------2"><img src="https://cdn-images-1.medium.com/max/1842/1*SihSvVABQrHxLV7cttfQrg.png" width="1842"></a></p><p class="medium-feed-snippet">Providing an intuition on the ideas behind popular transformer models like ChatGPT and other large language models (LLMs)</p><p class="medium-feed-link"><a href="https://medium.com/data-science/an-intuition-for-how-models-like-chatgpt-work-c7f01616bd6d?source=rss-82498630db6------2">Continue reading on TDS Archive »</a></p></div>]]></description>
            <link>https://medium.com/data-science/an-intuition-for-how-models-like-chatgpt-work-c7f01616bd6d?source=rss-82498630db6------2</link>
            <guid isPermaLink="false">https://medium.com/p/c7f01616bd6d</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[llm]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[David Hundley]]></dc:creator>
            <pubDate>Sat, 30 Dec 2023 15:40:24 GMT</pubDate>
            <atom:updated>2023-12-31T13:08:19.886Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Five Things GenAI Can and Can’t Do]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/data-science/five-things-genai-can-and-cant-do-d8117aad82f4?source=rss-82498630db6------2"><img src="https://cdn-images-1.medium.com/max/1842/1*WDrJXyNsCxnUYUjgzKJuoA.png" width="1842"></a></p><p class="medium-feed-snippet">An introductory guide for business leaders as to what Generative AI can or cannot do</p><p class="medium-feed-link"><a href="https://medium.com/data-science/five-things-genai-can-and-cant-do-d8117aad82f4?source=rss-82498630db6------2">Continue reading on TDS Archive »</a></p></div>]]></description>
            <link>https://medium.com/data-science/five-things-genai-can-and-cant-do-d8117aad82f4?source=rss-82498630db6------2</link>
            <guid isPermaLink="false">https://medium.com/p/d8117aad82f4</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[large-language-models]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[llm]]></category>
            <dc:creator><![CDATA[David Hundley]]></dc:creator>
            <pubDate>Sat, 07 Oct 2023 22:06:36 GMT</pubDate>
            <atom:updated>2023-10-08T16:05:55.925Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[An Introduction to OpenAI Function Calling]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/data-science/an-introduction-to-openai-function-calling-e47e7cd7680e?source=rss-82498630db6------2"><img src="https://cdn-images-1.medium.com/max/1842/1*eD4DSSn-GeVliFFsXSFCPQ.png" width="1842"></a></p><p class="medium-feed-snippet">No more unstructured data outputs; turn ChatGPT&#x2019;s completions into structured JSON!</p><p class="medium-feed-link"><a href="https://medium.com/data-science/an-introduction-to-openai-function-calling-e47e7cd7680e?source=rss-82498630db6------2">Continue reading on TDS Archive »</a></p></div>]]></description>
            <link>https://medium.com/data-science/an-introduction-to-openai-function-calling-e47e7cd7680e?source=rss-82498630db6------2</link>
            <guid isPermaLink="false">https://medium.com/p/e47e7cd7680e</guid>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[deep-dives]]></category>
            <category><![CDATA[genai]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[David Hundley]]></dc:creator>
            <pubDate>Sun, 09 Jul 2023 18:56:55 GMT</pubDate>
            <atom:updated>2023-07-28T21:50:35.314Z</atom:updated>
        </item>
    </channel>
</rss>