<?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:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd" xmlns:googleplay="http://www.google.com/schemas/play-podcasts/1.0"><channel><title><![CDATA[Think4.Me]]></title><description><![CDATA[Think4.Me is ]]></description><link>https://www.think4.me</link><image><url>https://substackcdn.com/image/fetch/$s_!S2o-!,w_256,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88b649ac-b5b6-4fe4-9da0-bc528b4fd0ed_800x800.png</url><title>Think4.Me</title><link>https://www.think4.me</link></image><generator>Substack</generator><lastBuildDate>Tue, 21 Jul 2026 02:44:43 GMT</lastBuildDate><atom:link href="https://www.think4.me/feed" rel="self" type="application/rss+xml"/><copyright><![CDATA[Greg Cowin]]></copyright><language><![CDATA[en]]></language><webMaster><![CDATA[gcowin@substack.com]]></webMaster><itunes:owner><itunes:email><![CDATA[gcowin@substack.com]]></itunes:email><itunes:name><![CDATA[Greg Cowin]]></itunes:name></itunes:owner><itunes:author><![CDATA[Greg Cowin]]></itunes:author><googleplay:owner><![CDATA[gcowin@substack.com]]></googleplay:owner><googleplay:email><![CDATA[gcowin@substack.com]]></googleplay:email><googleplay:author><![CDATA[Greg Cowin]]></googleplay:author><itunes:block><![CDATA[Yes]]></itunes:block><item><title><![CDATA[Weekly Link: Monday, July 13]]></title><description><![CDATA[AI Links]]></description><link>https://www.think4.me/p/weekly-link-monday-july-13</link><guid isPermaLink="false">https://www.think4.me/p/weekly-link-monday-july-13</guid><dc:creator><![CDATA[Greg Cowin]]></dc:creator><pubDate>Sat, 11 Jul 2026 12:44:20 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!S2o-!,w_256,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88b649ac-b5b6-4fe4-9da0-bc528b4fd0ed_800x800.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><strong>AI Links</strong></p><p>Microsoft is heavily investing with a $2.5 billion and 6,000 employees dedicated to AI implementation, aligning with data showing that companies spending the most on AI are also hiring more humans - emphasizing job transformation and augmentation over widespread job loss. On the model front, OpenAI rolled out GPT-5.5 Instant Mini. Research highlights non-invasive Brain2Qwerty technology advancing brain-computer interfaces for communication, and Microsoft continues enhancing its Foundry platform for developers.</p><p><strong>Use Cases</strong></p><p>Scaling generative AI in the cloud: Enterprise use cases for driving secure innovation</p><p><a href="https://azure.microsoft.com/en-us/blog/scaling-generative-ai-in-the-cloud-enterprise-use-cases-for-driving-secure-innovation/">Scaling generative AI in the cloud: Enterprise use cases for driving secure innovation | Microsoft A&#8230;</a></p><p><strong>Techniques and Training</strong></p><p>Harness Engineering for Coding Agent Users</p><p>&#8220;Details the agent harness as a cybernetic governor with feed-forward/feedback loops for regulating codebases and ensuring reliable agent execution.&#8221;</p><p><a href="https://martinfowler.com/articles/harness-engineering.html">Harness engineering for coding agent users</a></p><p><strong>Impacts: Business, Tech, Partnerships</strong></p><p>Microsoft commits $2.5 billion and 6,000 employees to new AI implementation unit</p><p><a href="https://www.cnbc.com/2026/07/02/microsoft-commits-2point5-billion-6000-employees-ai-implementation-unit.html">Microsoft commits $2.5 billion and 6,000 employees to new AI implementation unit</a></p><p>The Companies Spending the Most on AI Are Also Spending the Most on Humans</p><p>&#8220;The dire predictions of job loss are not happening according to the data. We&#8217;ve seen many predictions that all knowledge workers will lose their jobs in 2 years for about 4 years. It&#8217;s about Job replacement not Job loss.&#8221;</p><p><a href="https://www.reveliolabs.com/news/ai-and-work/greater-ai-investment-more-hiring/">Does AI Cause Job Losses? New Data Says Heavy Adopters Hire More | Revelio Labs</a></p><p><strong>Models/Agents</strong></p><p>GPT-5.5 Instant Mini (July 6, 2026)</p><p>&#8220;Rolled out in ChatGPT as the new fallback model (replacing GPT-5.3 Instant Mini). It offers better intent tracking, tone calibration, personalization, and fewer factual issues.&#8221;</p><p><a href="https://help.openai.com/en/articles/6825453-chatgpt-release-notes">https://help.openai.com/en/articles/6825453-chatgpt-release-notes</a></p><p><strong>Research/Papers</strong></p><p>From Brain Waves to Words: Brain2Qwerty Offers a New Path to Communication Without Surgery</p><p>&#8220;Programming: 2021 (fingers), 2026 (voice), 2030 (brain)?&#8221;</p><p><a href="https://ai.meta.com/blog/brain2qwerty-brain-ai-human-communication/">https://ai.meta.com/blog/brain2qwerty-brain-ai-human-communication/</a></p><p><strong>APIs/Coding</strong></p><p>What&#8217;s New in Microsoft Foundry | June 2026</p><p><a href="https://devblogs.microsoft.com/foundry/whats-new-in-microsoft-foundry-june-2026/">What&#8217;s New in Microsoft Foundry | June 2026 | Microsoft Foundry Blog</a></p>]]></content:encoded></item><item><title><![CDATA[FEB 18: Weekly AI Links and Prompts]]></title><description><![CDATA[Use Cases]]></description><link>https://www.think4.me/p/feb-18-weekly-ai-links-and-prompts</link><guid isPermaLink="false">https://www.think4.me/p/feb-18-weekly-ai-links-and-prompts</guid><dc:creator><![CDATA[Greg Cowin]]></dc:creator><pubDate>Wed, 18 Feb 2026 16:42:38 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!S2o-!,w_256,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88b649ac-b5b6-4fe4-9da0-bc528b4fd0ed_800x800.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><strong>    Use Cases</strong><br>        Learnings from 800+ GenAI and ML use cases<br>        <a href="https://www.evidentlyai.com/blog/gen-ai-applications">https://www.evidentlyai.com/blog/gen-ai-applications</a><br><br>    <strong>Techniques and Training</strong><br>        13 Foundational Types of AI Models<br>       <a href="https://www.turingpost.com/p/13-foundational-types-of-ai-models"> https://www.turingpost.com/p/13-foundational-types-of-ai-models</a><br><br>    <strong>Impacts: Business, Tech, Partnerships</strong><br>        OpenClaw Creator Peter Steinberger Joins OpenAI, OpenClaw To Continue As Open-source Project<br>        <a href="https://officechai.com/ai/openclaw-creator-peter-steinberger-joins-openai-openclaw-to-continue-as-open-source-project/">https://officechai.com/ai/openclaw-creator-peter-steinberger-joins-openai-openclaw-to-continue-as-open-source-project/</a><br><br>        DALL-E 3: Scheduled for retirement on February 18, 2026. Users are urged to migrate to gpt-image-1-mini to avoid service disruption.<br><br>    <strong>Models/Agents</strong><br>        Grok 4.2 from xAI (beta)<br>        "4 competing agents; not in Microsoft Foundry yet"<br><br>        GLM-5 from Z.AI<br>        "GLM-5 is designed for complex systems engineering and long-horizon agentic tasks. Open weights and scores very high."<br>        <a href="https://z.ai/blog/glm-5">https://z.ai/blog/glm-5</a><br><br>        Azure Copilot from Microsoft<br>        "Get assistance in Azure"<br>        <a href="https://azure.microsoft.com/en-us/products/copilot">https://azure.microsoft.com/en-us/products/copilot</a><br>        <a href="https://learn.microsoft.com/en-us/azure/copilot/overview">https://learn.microsoft.com/en-us/azure/copilot/overview</a> <br> <br>    <strong>Research/Papers</strong><br>        Prompt Repetition Improves Non-Reasoning LLMs<br>        <a href="https://learn.microsoft.com/en-us/azure/copilot/overview">https://arxiv.org/pdf/2512.14982</a><br><br>        Self-Feedback&#8211;Driven LLM Reasoning<br>        <a href="https://arxiv.org/pdf/2602.09000">https://arxiv.org/pdf/2602.09000</a><br><br>    <strong>APIs/Coding</strong><br>        Building a LLM-from-scratch<br>       <a href="https://github.com/rasbt/LLMs-from-scratch"> https://github.com/rasbt/LLMs-from-scratch</a><br> <br>    <strong>Little Prompts:</strong><br>        ...<br>        Where could an AI system fail in this scenario?</p>]]></content:encoded></item><item><title><![CDATA[Weekly: AI Links and Little Prompts]]></title><description><![CDATA[Feb 13 2026]]></description><link>https://www.think4.me/p/coming-soon</link><guid isPermaLink="false">https://www.think4.me/p/coming-soon</guid><dc:creator><![CDATA[Greg Cowin]]></dc:creator><pubDate>Fri, 13 Feb 2026 18:46:37 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!S2o-!,w_256,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88b649ac-b5b6-4fe4-9da0-bc528b4fd0ed_800x800.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<h1><strong>AI Hours: Links and Prompts</strong></h1><p>    <strong>Use Cases</strong></p><p>        Transformative AI starts with clear use cases</p><p>       <a href="https://aws.amazon.com/ai/generative-ai/use-cases/"> https://aws.amazon.com/ai/generative-ai/use-cases/</a></p><p>    <strong>Techniques and Training</strong></p><p>        Calling AI a Stochastic Parrot is wrong - Geoffey Hilton - known for his work on artificial neural networks, which earned him the title of &#8216;Godfather of AI&#8217;</p><p>        &#8220;Calling it stochastic parrot lacks an understanding of the mechanisms; it is not statistics.&#8221;</p><p>        <a href="https://medium.com/@jk1849716/from-statistical-parrots-to-topological-cognition-beyond-feature-spaces-in-generative-models-54b44dd875e">https://medium.com/@jk1849716/from-statistical-parrots-to-topological-cognition-beyond-feature-spaces-in-generative-models-54b44dd875e</a>3</p><p>    Impacts: Business, Tech, Partnerships</p><p>        From magic to malware: How OpenClaw&#8217;s agent skills become an attack surface</p><p>        &#8220;One of the top skills for OpenClaw is malware&#8221;</p><p>        <a href="https://1password.com/blog/from-magic-to-malware-how-openclaws-agent-skills-become-an-attack-surface">https://1password.com/blog/from-magic-to-malware-how-openclaws-agent-skills-become-an-attack-surface</a></p><p>    <strong>Models/Agents</strong></p><p>        Anthropic&#8217;s Opus 4.6 built a C compiler with team of parallel Claudes</p><p>        <a href="https://www.anthropic.com/engineering/building-c-compiler">https://www.anthropic.com/engineering/building-c-compiler</a></p><p>        Qwen3-Coder-Next</p><p>        &#8220;Alibaba&#8217;s open weights model; 80B with 3B active; scores better than previous 235B model&#8221;</p><p>        <a href="https://huggingface.co/Qwen/Qwen3-Coder-Next">https://huggingface.co/Qwen/Qwen3-Coder-Next</a></p><p>    <strong>Research/Papers</strong></p><p>        Scaling Long-Horizon LLM Agent via Context Folding</p><p>        &#8220;Context Folder and Agent Folding are emerging useful techniques</p><p>        Context folding is a pattern in AI system design where you compress, transform, or selectively carry forward prior interaction state into a structured representation that fits within the model&#8217;s context window &#8212; while preserving what matters.&#8221;</p><p>        <a href="https://arxiv.org/pdf/2510.11967">https://arxiv.org/pdf/2510.11967</a></p><p>        Recursive Language Models: the paradigm of 2026</p><p>       <a href="https://www.primeintellect.ai/blog/rlm"> https://www.primeintellect.ai/blog/rlm</a></p><p>    A<strong>PIs/Coding</strong></p><p>        Palantir&#8217;s Foundry and Ontology Architectural Patterns for Data Engineers</p><p>        &#8220;Taking AI &amp; Data to a higher level&#8221;</p><p>        <a href="https://medium.com/@sachtekchanda90/im-a-data-engineer-what-is-palantir-foundry-7e7568aed87e">https://medium.com/@sachtekchanda90/im-a-data-engineer-what-is-palantir-foundry-7e7568aed87e</a></p><p>        <a href="https://medium.com/@sachtekchanda90/ontology-catalog-design-in-palantir-foundry-part-1-21904cebd7d3">https://medium.com/@sachtekchanda90/ontology-catalog-design-in-palantir-foundry-part-1-21904cebd7d3</a></p><p>    <strong>Little Prompts:</strong></p><p>        ...</p><p>        Show the decision points.</p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.think4.me/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://www.think4.me/subscribe?"><span>Subscribe now</span></a></p>]]></content:encoded></item></channel></rss>