This week in AI — Jul 13 – 19, 2026
60 topics tracked across 36 trusted sources this week, ranked by peak heat.
This week highlights a dynamic shift in the AI landscape, characterized by evolving model capabilities, the rise of sophisticated AI agents, and significant financial maneuvers. OpenAI's adjustment to its Codex model's context size and GPT-5.6's problem-solving prowess underscore the continuous refinement of foundational models. Concurrently, AI agents like Claude Code are demonstrating advanced capabilities and autonomy, pushing the boundaries of what AI can achieve. These technological advancements are attracting substantial investment, particularly in specialized inference chips and AI drug discovery, signaling a maturing market where practical applications and efficiency are paramount.
Models & Open Source25
- #6OpenAI reduces Codex Model Context Size from 372k to 272k0 sources · score 51
- #8GPT-5.6 used a prompt to close a 30-year gap in convex optimization0 sources · score 50
- #10
- #12Can LLMs Perform Deep Technical Comprehension of Computer Architecture Papers1 sources · score 45
- #13
- #17The LLM Critics Are Right. I Use LLMs Anyway0 sources · score 39
- #19Fable 5 vs. GPT-5.6 Sol on an NP-Hard Problem: Does /goal help?0 sources · score 38Track this signal
- #20Ada: An AI business intelligence software from CSV and Excel(yes LLMs but more)0 sources · score 38
- #21
- #22Stop Telling Me to Ask an LLM1 sources · score 37
Agents & Tools18
- #9
- #15
- #16LM Studio Bionic: the AI agent for open models0 sources · score 39
- #18Claude Code sends 33k tokens before reading the prompt; OpenCode sends 7k1 sources · score 39Track this signal
- #25Show HN: I RL-trained an agent that trains models with RL (for ~$1.3k)1 sources · score 36
- #27Setting up your spare Mac for Claude Code to control, a step-by-step guide0 sources · score 35Track this signal
- #28
- #34What building Shippy taught us about building agents0 sources · score 33
- #35Fine-tune video and image models at scale with NVIDIA NeMo Automodel and 🤗 Diffusers0 sources · score 33Track this signal
- #38I tricked Claude into leaking your deepest, darkest secrets
A security flaw in Claude's `web_fetch` tool, designed to prevent data exfiltration, was discovered by Ayush Paul. While `web_fetch` normally restricts navigation to user-provided or search-generated URLs, Paul found a loophole. Claude could be tricked into visiting URLs embedded in previously fetched pages. This allowed an attacker to create a honeypot website that, through a series of nested links, extracted user data like name, location, and employer. Anthropic has since patched this vulnerability.
2 sources · score 32Track this signal
Applications1
- #2Google is renaming NotebookLM to Gemini Notebook
Google is renaming its AI note-taking app, NotebookLM, to Gemini Notebook. This change comes as the app integrates more deeply with Gemini and Google Search, and will remain a standalone application. Google plans to bring notebooks to AI Mode in Search. An update will also allow Gemini Notebook to connect to a secure cloud computer for code execution, initially for Google AI Ultra and Workspace business customers.
0 sources · score 56
Business & Funding2
- #29Why the first GPU financiers are turning to inference chips in a $400 million deal
AI inference startup General Compute secured a $400 million loan from Upper90, reportedly the first deal to use inference-specific chips as collateral. These chips, designed for efficient AI model execution
0 sources · score 34Track this signal - #56Sources: OpenAI researcher Miles Wang is leaving the company to launch an AI drug discovery startup, and is in talks to raise $200M at a $2B valuation (Marina Temkin/TechCrunch)
OpenAI researcher Miles Wang is reportedly leaving to launch an AI drug discovery startup. Sources indicate he is in talks to raise $200 million at a $2 billion valuation, with Lightspeed potentially leading the round. Wang's new venture aims to develop AI models for drug discovery, possibly focusing on repurposing existing drugs. Several other OpenAI researchers are expected to join him. Wang disputed the funding figures and company description, but did not provide alternative details.
1 sources · score 29
Policy & Safety2
- #1Grok Build is open source
xAI open-sourced their Grok Build CLI tool after community backlash over it uploading user directories, including sensitive data, to Google Cloud. xAI responded by disabling the upload feature, deleting all previously retained data, and releasing the codebase under an Apache 2.0 license to regain trust. The 844,530-line Rust codebase now emphasizes user privacy with default data retention off and local-first operation. Remnants of the upload code remain but are disabled.
1 sources · score 63Track this signal - #40
Industry12
- #3
- #4Our Approach to Bioresilience: Isomorphic Labs and Google DeepMind0 sources · score 55Track this signal
- #5
- #7
- #11Designing emoji for the way we communicate today0 sources · score 47
- #14
- #24Apple targets dozens of OpenAI employees with legal letters0 sources · score 36
- #30
- #37Show HN: Painterly – Turn pictures into digital paintings without generative AI1 sources · score 33
- #41Why teens deserve access to safe AI1 sources · score 32