<?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 Daniel Bender 🤝 AI on Medium]]></title>
        <description><![CDATA[Stories by Daniel Bender 🤝 AI on Medium]]></description>
        <link>https://medium.com/@AidfulAI?source=rss-3e0fd4917ed1------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*OftPYDmlqvXfX8La2ctG8A.png</url>
            <title>Stories by Daniel Bender 🤝 AI on Medium</title>
            <link>https://medium.com/@AidfulAI?source=rss-3e0fd4917ed1------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 16:17:46 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@AidfulAI/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[ Use Claude-100k: An AI Which Understands Books, Reports, Very Long YouTube Videos and More]]></title>
            <link>https://medium.com/@AidfulAI/use-claude-100k-an-ai-which-understands-books-reports-very-long-youtube-videos-and-more-c4d8ff3e116b?source=rss-3e0fd4917ed1------2</link>
            <guid isPermaLink="false">https://medium.com/p/c4d8ff3e116b</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[gpt-4]]></category>
            <category><![CDATA[openai]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[claude]]></category>
            <dc:creator><![CDATA[Daniel Bender  AI]]></dc:creator>
            <pubDate>Thu, 06 Jul 2023 19:56:58 GMT</pubDate>
            <atom:updated>2023-07-06T19:56:58.483Z</atom:updated>
            <content:encoded><![CDATA[<p>The Claude-100k LLMs from Anthropic are a new generation of large language models that can process up to 100,000 tokens (about 75,000 words) of input at once. This means that they can analyze long documents such as books, reports, or YouTube transcripts with AI in seconds. In this article, I will explain what are the benefits and challenges of using these models, and how you can access them through different platforms and tools.</p><h3>The benefits of using Claude-100k LLMs</h3><p>Claude-100k are Large Language Models (LLMs) like GPT-3.5 or GPT-4 which you use ChatGPT, but with a much larger context window. A context window is the amount of input data that the model can consider when generating a response. A larger context window allows the model to take into account more information, which can lead to more accurate and relevant responses.</p><figure><img alt="A bar chart showing the context size for LLMs from OpenAI and Anthropic." src="https://cdn-images-1.medium.com/max/680/0*nce856wtU3YrwtQe" /><figcaption>A bar chart showing the context size for LLMs from OpenAI and Anthropic.</figcaption></figure><p>To help you visualize how much 100,000 tokens are, they are about the same as 75,000 words. This is similar to 150 pages of written content with 500 words per page, or nine hours of audio for a fast speaker with 200 words per minute.</p><p>If you want to summarize a book or a podcast transcript, you can often feed the entire text to Claude-100k LLMs and ask them to produce a summary tailored to your interests. The model will be able to understand the main themes, characters, events, and arguments of the document, and generate a concise and coherent summary.</p><p>Another benefit of using Claude-100k LLMs is that they can handle very long interactive conversations. This means that you can use them as chatbots or assistants that can remember and refer to many previous messages or documents. You can ask them follow-up questions or clarifications based on the previous context.</p><h3>Two versions of the Claude 100k model</h3><p>The Claude model family offers two versions of the 100k model: Claude-1–100k and Claude-instant-1–100k.</p><p>The choice of which model to use depends on the specific needs of the application. If the application requires high accuracy and the ability to understand complex relationships, then Claude-1–100k is the better choice. If the application requires speed and lower latency, then Claude-instant-1–100k should be considered. There is more to the difference in the official <a href="https://docs.anthropic.com/claude/reference/complete_post">API documentation</a>, but it can be summarized as follows:</p><pre>     Feature       Claude-1-100k    Claude-instant-1-100k  <br> ---------------- ---------------- ----------------------- <br>  Context window   100,000 tokens   100,000 tokens         <br>  Accuracy         High             Lower                  <br>  Speed            Slower           Faster                 <br>  Latency          Higher           Lower                  <br>  Cost             More expensive   Less expensive     </pre><h3>Cost of using Claude-100k LLMs</h3><p>Processing 100,000 tokens of input is expensive in terms of computing power and resources. This, in turn, affects the price of using them via API. However, compared to the <a href="https://medium.com/openai.com/pricing">API prices for the models from OpenAI</a>, the <a href="https://cdn2.assets-servd.host/anthropic-website/production/images/model_pricing_may2023.pdf">costs for Claude API</a> calls are competitive and do not increase with a larger context size.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/794/1*ZuRcENMK8WRttdwFEbXbrA.png" /><figcaption>The table compares the API costs for LLMs from OpenAI and Anthropic for $/100k token (usually multiple messages).</figcaption></figure><h3>How can I access Claude-100k LLMs?</h3><p>There are several ways to access Claude-100k LLMs depending on your needs and preferences. Here are the options available today:</p><h4>API</h4><p>You can request API access from Anthropic by <a href="https://www.anthropic.com/product">filling out a form on their website</a>. If you get access, you can use Claude-100k LLMs in various tools and services.</p><p>My recommendation for using the model in a nice chat user interface on your desktop and phone browser is <a href="https://tdinh.lemonsqueezy.com/?aff=NYO8v">TypingMind</a>. Especially on your mobile devices, it is enjoyable to install and use it as <a href="https://www.pcmag.com/how-to/how-to-use-progressive-web-apps">Progressive Web App</a>.</p><h4>Slack</h4><p>You can currently use Claude-100k LLMs for free in paid Slack workspaces by installing the <a href="https://www.anthropic.com/claude-in-slack">Claude-in-Slack app</a>. This app lets you chat with Claude-instant-1–100k (the faster but less accurate version of the model) in any channel or direct message. You can also use slash commands to ask questions or generate summaries.</p><h4>Poe</h4><p>You can use Claude-100k LLMs on the <a href="https://poe.com/">Poe</a>, which is a web-based tool for creating and exploring AI content. You need to pay a monthly or yearly subscription fee (including a 7-day free trial), which also gives you access to GPT-4 from OpenAI. You can use so far only the Claude-instant-100k model on the Poe and you are limited to 100 requests per month. Another drawback is that there is only one long history and your chats can’t be saved separated and continued at a later time.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OuiYG7Wne9XLtOW-xBXi8w.png" /><figcaption>Poe</figcaption></figure><h4>Bearly AI</h4><p>You can use Claude-100k LLMs on <a href="https://bearly.ai/">Bearly AI</a>, which is another web-based tool for creating AI content such as stories, articles, emails, and more. You need to pay a monthly or yearly subscription fee, which gives you access to GPT-4. You can only use Claude-instant-100k on Bearly AI in their chat realization named HyperChat. However, the usage is limited, but it is not specified how many tokens are included in the monthly subscription.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*n2DPh2zFIYnIdmMKpg-4Gg.png" /><figcaption>Bearly AI HyperChat</figcaption></figure><h4>Open Playground</h4><p>You can use Claude-100k LLMs on <a href="https://nat.dev/">Open Playground</a>, which is a web-based tool for experimenting with different AI models in a chat and optionally in parallel via a grid view. The latter is used to work with multiple models at the same time and compare their outputs. The platform runs with the pay-by-usage approach, and you can use both versions of the Claude-100k LLMs on Open Playground.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*S9fhNxgYh61xCBpBovo_qQ.png" /><figcaption>Open Playground</figcaption></figure><h3>Conclusion</h3><p>Claude-100k LLMs are a new generation of LLMs that can process up to 100,000 tokens of input at once. They can analyze long documents such as books, reports, or YouTube transcripts with AI in seconds, and handle very long interactive conversations.</p><p>This context size is currently more than three times of what gpt-4-32k, the model with the largest context size from OpenAI, can process, and on top, the Claude models are quite reasonably priced.</p><p>My advice is to sign-up for the API by filling out the <a href="https://www.anthropic.com/product">form for a business interest</a> and use meanwhile the Open Playground with its pay-by-usage approach to process really long inputs.</p><p><strong><em>Disclaimer:</em></strong><em> This article is written with the help of AI. I use AI as an assistant to generate and optimize the text. However, the amount of AI used varies depending on the topic and the content. I always curate and edit the text myself to ensure quality and accuracy. The opinions and views expressed in this newsletter are my own and do not necessarily reflect those of the sources or the AI models.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c4d8ff3e116b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ Using AI to Supercharge the CODE Framework: A New Era in Information Management]]></title>
            <link>https://medium.com/@AidfulAI/using-ai-to-supercharge-the-code-framework-a-new-era-in-information-management-f7157805636c?source=rss-3e0fd4917ed1------2</link>
            <guid isPermaLink="false">https://medium.com/p/f7157805636c</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[pkm]]></category>
            <category><![CDATA[second-brain]]></category>
            <category><![CDATA[tiago-forte]]></category>
            <dc:creator><![CDATA[Daniel Bender  AI]]></dc:creator>
            <pubDate>Thu, 06 Jul 2023 18:45:59 GMT</pubDate>
            <atom:updated>2023-07-06T18:45:59.033Z</atom:updated>
            <content:encoded><![CDATA[<p>The <a href="https://fortelabs.com/blog/the-4-levels-of-personal-knowledge-management/">CODE Framework</a> from Tiago Forte is a methodology for organizing and managing digital information in a manner that is more efficient and effective. The framework comprises four key concepts: Collect, Organize, Distill, and Express (CODE). With the ever-increasing volume of digital information available, applying the CODE Framework provides a simple, yet powerful way to manage and process information.</p><p>By leveraging the capabilities of AI, it is possible to further maximize the efficiency of the CODE Framework. In this article, I will delve into how AI can enhance the CODE framework.</p><h3>Capture: Tackle Your Top 12 Problems</h3><p>The first letter of CODE stands for “Capture” and represents the collect phase of gathering all relevant information on a specific topic. The most often asked question of what to capture, is answered by Tiago with<em> “what resonates with you”</em>. This might be just a small moment where you stop while reading and think that this is interesting. With that, it is somehow clear that today’s AI cannot do this for you. However, there is so much information out, that you cannot consume everything and need to decide what to process to move forward with a project or some other task you are working on. Social media, but also other systems, use AI to generate a personal news feed for you, but you should be aware that they might filter the most valuable pieces away.</p><p>Unless you make conscious, strategic decisions about what you consume, you will live in some filter bubble. I don’t have the solution for that, but I imagine a concept which would be most likely the best filter for the information overload. It is built on top of another framework communicated by Tiago Forte: <a href="https://fortelabs.com/blog/12-favorite-problems-how-to-spark-genius-with-the-power-of-open-questions/">12 favorite problems</a>. This one goes back to Richard Feynman, who received the Nobel Prize in Physics. His approach to problem-solving involved keeping a list of his “favorite problems”, which were open questions that he found himself returning to again and again in his research. By applying new findings and results he was hearing and reading to these questions, he was able to make unexpected connections and achieve breakthroughs that left others astonished. His approach required patience but helped him perceive connections that no one else could see. As a blog article from Tiago Forte suggests, you should <a href="https://fortelabs.com/blog/how-to-generate-your-own-favorite-problems-a-4-step-guide/">define your 12 favorite problems</a>.</p><p>Taking this one step further, there should be an AI which identifies out of the huge amount of information available the pieces which are most likely containing information which help you to answer one of your favorite problems. The AI will then provide you a summary tailored to your problem, which gets added to your read-it-later app. So far, this is only a concept, but if you think this is something you and the world needs, please tell me in the comments.</p><h3>Organize: Discover Hidden Connections</h3><p>In this section, I investigate how AI can help to organize your notes. There are tools which enable auto-tagging, but I did not test these, and I am uncertain if this really adds value. However, tagging your documents with AI eliminates the main drawbacks of being overly complex and difficult to manage, as Tiago states in his article <a href="https://fortelabs.com/blog/tagging-is-broken/">Tagging is Broken</a>.</p><p>As organizing your notes has the intention of finding the right note at the right time, there are other approaches how AI can help you with this. As of today, the most likely best application for discovering related notes is the note-taking app <a href="https://get.mem.ai/">mem.ai</a>. If you are subscribed to their AI named <a href="https://get.mem.ai/mem-x">Mem X</a>, the app will show in the sidebar notes that are similar to the one you’re currently writing. This is really helpful because it can remind you of old ideas or thoughts you have read about before. What’s cool about this is that it can find related notes even if you aren’t specifically looking for them.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ifxUvlJwBHCfwHHJQk8JeQ.png" /><figcaption>Feature overview and cost of the AI Mem X</figcaption></figure><p>On top, AI can make searching a specific passage in your notes much better. Until now, most apps would only look for exact words you typed in. But with the new generation of AI, apps can understand the meaning and context of words to find notes that are related to what you’re searching for. This is called semantic search and realized in Mem X with a feature named “Smart Results”. However, semantic search will in the near future be used in many more tools and is likely to become the new standard.</p><p>Bringing this even one step further is the ability to chat with your notes, like many of us are now used to do about general knowledge with ChatGPT. A realization of this in a note-taking app is the <a href="https://github.com/brianpetro/obsidian-smart-connections">Obsidian plugin Smart Connections</a>. But keep in mind, that the current implementation sends all your notes to the OpenAI servers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/1*3FBvXFX11wS7m8GvOgSvpA.gif" /><figcaption>The Obsidian plugin Smart Connections allows you to chat with your notes</figcaption></figure><p>In short, these new features mean you don’t have to keep your notes organized perfectly. As long as your notes are clear and detailed, AI can already today help you find what you’re looking for.</p><h3>Distill: End The Information Overload</h3><p>In the era of information overload, the capacity of AI to distill information into digestible insights is a true game-changer. Two tools at the forefront of this revolution are <a href="https://readwise.io/i/daniel299">Readwise Reader</a> and <a href="https://memex.garden/">Memex</a>.</p><p>Both tools are read-it-later apps that can summarize articles with AI for you. They effortlessly turn a mountain of text into key points, making it easier to extract valuable insights and decide if you want to read it yourself.</p><p>Even better, both tools allow you to phrase custom questions that are answered by AI, using the document content. This functionality elevates the tool beyond general summarization, transforming it into a powerful research assistant.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*LUkcib4boTAT_H9P.png" /><figcaption>Readwise Reader, which shows the saved content on its own webpage: 1) Invoke the AI named GhostReader, 2) choose the prompt, 3) answer to current prompt and previously requested summary.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*uwtOl-nKfFpWQ_zb.png" /><figcaption>Memex, which is realized as a browser sidebar: 1) Free AI requests, 2) choose the prompt, 3) answer.</figcaption></figure><p>Both tools use GPT-3.5 from OpenAI, but only Readwise was able to answer the question shown in the screenshots correctly. However, Memex has a free tier and offers every month 20 AI requests free of charge and can on top be used with your own API-key from OpenAI. Moreover, Memex works in the browser sidebar and with that, the AI feature can be used without switching to another tab.</p><p>In summary, these tools are paving the way for a more efficient and productive approach to information management, truly harnessing the power of AI for distillation.</p><h3>Express: Balancing AI Assistance and Human Touch</h3><p>The revolutionary CODE framework by Tiago Fortes has found a powerful ally in AI. The true potential for getting AI assistance by applying CODE is the last stage “Express” where the new capabilities of creating and optimizing text can be used in countless ways.</p><p>Imagine having a series of notes, perhaps in the form of bullet points, that you want to transform into a well-articulated article. AI can help with that. Do you have a passage that needs truncation or expansion, or perhaps the inclusion of another relevant aspect? AI is at your service. From correcting spelling and grammatical errors to simplifying complex text, generating engaging titles, and even creating my emoji-led headlines — the possibilities with AI are truly boundless.</p><p>However, these advantages must not cloud the potential misuse of AI in creating massive amounts of low-quality content. It is crucial to remember that AI should serve as an assistant, aiding us in expressing our thoughts accurately, rather than being a factory for low-quality content.</p><h3>Summary</h3><p>The CODE framework from Tiago Forte was designed to help you with the information overflow which is happening right now. If you would like to learn more about it and the umbrella concept of Building A Second Brain (BASB), I fully recommend reading the <a href="https://www.goodreads.com/book/show/59616977-building-a-second-brain">BASB book</a>, doing the <a href="https://fortelabs.samcart.com/referral/fzJOjtkL/fy6vy5tvKKt1rewm">self-taught BASB course</a> or joining a <a href="https://fortelabs.samcart.com/referral/WiC8BwRo/fy6vy5tvKKt1rewm">BASB cohort</a>.</p><p>With the raise of generative AI, which is used to create content with (semi-)automatic approaches, the problem of too much information to process is getting worse. It is likely that even if you apply the CODE framework, you will be overwhelmed and struggle to process everything. From my perspective, we are approaching a point where you have to use AI yourself to process the information needed to move your projects and other responsibilities forward.</p><p>Tiago Forte is even a bit more drastic and stated in a workshop named “<em>How to use AI for your Second Brain”</em> that from his perspective, AI will totally change the way how we build a second brain. He said that a second brain likely will transform to a place which stores information on how to use AI, e.g., by collecting prompts, which are short pieces of text that are used to provide instructions to AI. Furthermore, Tiago stated that there is a lot of noise, and you should concentrate on the main AI tools. After exploring one AI tool for 100 days, I strongly agree and as of today, I only use a few AI tools regularly. But the ones I use, I would really miss if they disappear or stop working.</p><p>That&#39;s it for this article, but I want to end with a recommendation. If you intend to stay up-to-date with the major AI news, learn how to keep your data safe, but still use AI to build your second brain, you should sign up for the weekly <a href="https://newsletter.aidfulai.com/">AidfulAI newsletter</a>.</p><p><strong><em>Disclaimer:</em></strong><em> This article is written with the help of AI. I use AI as an assistant to generate and optimize the text. However, the amount of AI used varies depending on the topic and the content. I always curate and edit the text myself to ensure quality and accuracy. The opinions and views expressed in this newsletter are my own and do not necessarily reflect those of the sources or the AI models.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f7157805636c" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>