<?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 Olivier L. on Medium]]></title>
        <description><![CDATA[Stories by Olivier L. on Medium]]></description>
        <link>https://medium.com/@olivierllouquet?source=rss-dd980eff1495------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*1nPk9lygWTHUVT8EKNsQkw.jpeg</url>
            <title>Stories by Olivier L. on Medium</title>
            <link>https://medium.com/@olivierllouquet?source=rss-dd980eff1495------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 02:44:18 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@olivierllouquet/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[Shaping the affordances of LLMs to design experiences]]></title>
            <link>https://medium.com/@olivierllouquet/shaping-the-affordances-of-llms-to-design-experiences-062b906ce014?source=rss-dd980eff1495------2</link>
            <guid isPermaLink="false">https://medium.com/p/062b906ce014</guid>
            <category><![CDATA[weekly-thought]]></category>
            <dc:creator><![CDATA[Olivier L.]]></dc:creator>
            <pubDate>Fri, 11 Oct 2024 14:37:09 GMT</pubDate>
            <atom:updated>2024-10-11T14:37:09.251Z</atom:updated>
            <content:encoded><![CDATA[<p>Affordances in design refer to the qualities or properties of an object that suggest how it can be used.</p><p>Applying the concept of affordances to LLMs, we can explore the ways in which their features, functionalities, and behaviors can be shaped to facilitate certain interactions or outcomes.</p><p><strong>The affordances of LLMS</strong></p><p>Interacting with a large language model (LLM) is the <strong>prompt field</strong>, which invites the user to input natural language text.</p><p>The prompt itself has multiple layers of affordance:</p><ul><li><strong>Flexibility in Input</strong>: The user can enter anything from a simple query to a complex request, allowing a wide range of interactions. The prompt field suggests, “you can ask me anything,” encouraging exploration and creative thinking.</li><li><strong>Immediacy of Feedback</strong>: Users expect a quick response, and the conversational interface delivers by generating responses in real-time. This immediate feedback reinforces the affordance of fluid interaction, just like a human conversation.</li><li><strong>Back-and-Forth Dialogue</strong>: The conversational nature of the interface suggests that the interaction doesn’t stop at a single exchange. The system often “remembers” prior inputs within a session, allowing for multi-turn dialogues that feel like a natural conversation. This affordance shifts from static one-time queries (like in traditional search engines) to dynamic, ongoing interactions.</li></ul><p><strong>Meta-cognition</strong></p><p>As users generate prompts, they can step back and assess the LLM’s outputs in relation to their prior knowledge or external sources, critically reflecting on the quality, accuracy, and relevance of the information. This reflective loop allows users to think about their own thinking and refine their approach to learning or problem-solving.</p><p>LLMs foster meta-cognition by:</p><ul><li>Encouraging <strong>reflection</strong> through iterative questioning and feedback loops.</li><li>Allowing users to <strong>externalize</strong> and examine their thought processes by articulating ideas.</li><li>Enabling <strong>synthesis</strong> of information and deeper integration of knowledge from multiple sources.</li><li>Facilitating <strong>self-questioning</strong> and prompting users to assess their own understanding.</li><li>Navigating <strong>ambiguity</strong> and complexity, requiring users to engage in higher-order thinking.</li><li>Supporting <strong>self-monitoring</strong> and personalized learning through context-aware interactions.</li><li>Helping users <strong>identify and correct errors</strong>, promoting ongoing learning and refinement.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=062b906ce014" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>