Welcome to another round-up of interesting AI news and some general web news as well. Here they are: * Google Gemini dubbed ‘high risk’ for kids and teens in new safety assessment. Together with OpenAI being sued for user deaths within, this is another story on how LLM controls for safety really don't work that well yet. Most of the controls and guardrails are added in post-training or within guardrails during usage. These mechanisms are just too weak to work yet. There are hypotheses that put
From AI Charter Schools to how reasoning models are driving the demand for AI Capacity. Another packed week of stories.
AI leaders offer government $1 AI, NVIDIA argues for smaller models, and Google shares energy data
Drug discovery, Chrome, Perplexity, and AI as the interface to the online world.
GPT-5 takes over the airways, we're looking forward to Cynic mode.
A huge week for releases and advances. Google and Github both release vibe coding tools. Anthropic integrates with Canva and Figma.
AI economics are unpredictable, but starting to feel inevitable.
ChatGPT Agent, No Code, and Sneaky Text Prompts
This week, NVIDIA's market value reached $4M, which is a huge milestone. It's our first story, but also, I think, a good moment to pause. The real meaning here, I think, is that there is a general consensus that much of human productivity and life is going to be running "on" or "with" AI.
Some interesting stories this week: * New Insights for Scaling Laws in Autonomous Driving (from the Waymo Team). TL;DR Tesla's strategy of turning all its cars into roving data capture machines was probably the right call. Not that Waymo isn't saying that Lidar is useless (Waymo uses expensive rigs whereas Tesla works with vision only), like Elon Musk is. In a sense, they are saying the opposite; most likely, data from multiple sensor types is still better as long as you can get a lot of it.
Sunday Links: Copyright, Failure Rates, and Mosquito Drones
ChatGPT outage scares everybody, Apple paper on LLM reasoning, LLMs.txt
Salesforce closes up Slack Data access, Shopify removes UX job titles, and Google Veo stars in Ad generation.
AI proto-alignment is causing escapes, OpenAI goes after business use-cases to capture context, and AI Engineering talks.
AI snitching benchmark, Google search, Jobs and other AI news.
Big strategy moves in AI with OpenAI moving into devices, adopting MCP and Google launching AI Mode in search
Shopping via AI, Lego creation, and keeping LLMs safe (or not).
AI Economics, Theorem Proving, and AI Remaking the Software Stack
Advertising via blackbox, FAA-approved drones, and Claude meets MCP, meets Zapier.
Made by humans initiative by Adobe, AI functionality gets compacted and OpenAI wants to buy Windsurf.