‘Torturing’ LLMs Is Not Real, But Doing It Still Probably Makes You a Bad Person
There are no ethics here, just a general abiding unpleasantness.
Gizmodo — AI4 min read
LLMs
90 stories
PewDiePie unveils ‘uncensored’ Ajax AI model for home PCs
PewDiePie says OpenAI banned him twice while he made Ajax, a fine-tuned Qwen3.5-9B agent.
Tom's Hardware AI3 min read
Don’t be fooled—LLMs don’t reason
On an afternoon in Seoul in March 2016, I watched a program I helped build put a stone on the fifth line of a Go board in what looked like a gift to its human opponent. Move 37 in game two of the five-game match looked so absurd that some commentators thought it was a…
MIT Technology Review7 min read
AWS debuts Strands Decider 2B, a first lightweight decision model for accelerate agentic workflows
Amazon Web Services Inc.’s Strands Labs team has been playing around with an emerging class of lightweight artificial intelligence systems known as “decision models,” and it’s now making the fruits of that work available to the open-source community.
SiliconANGLE4 min read
Firm rents four Nvidia H200s to test '80x cheaper' DeepSeek claim
A call center consultancy rented four Nvidia H200s to run DeepSeek V4.1 Flash for Claude Code and found DeepSeek's API cheaper.
Tom's Hardware AI4 min readQuoting Matthew Green
[...] Put these pieces together and you have the two halves of a worm: a payload that hijacks the agent, and an agent that will carry the payload to the next agent.
Simon Willison1 min read
Emergence AI to Deploy Neuroformal AI With Fabless Chipmakers
See how Emergence AI is deploying neuroformal AI with chipmakers to boost wafer yields and tackle fab, test, and packaging failures. The post Emergence AI to Deploy Neuroformal AI With Fabless Chipmakers appeared first on EE Times.
EE Times6 min read
Tracing Agent Harness Behavior with NVIDIA NeMo Relay
An agent can finish a task and still take an inefficient path. A failed search can trigger another search. A truncated file read can lead to a command fetching...
NVIDIA Developer Blog12 min readQuoting Anthropic Frontier Red Team
We evaluate several models on 100 tasks from the [internal Binary Exploitation benchmark] (selected at random), and find that GLM-5.3 develops full control flow hijacks in 4% of the trials; Claude Mythos Preview did so in 6%.
Simon Willison1 min readGPT 6.1 Sol: Near-Astra intelligence for a fifth of the price
My comment on GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price — Hacker News.I'm a bit late with the pelicans because I was live-blogging the keynote: https://simonwillison.net/2026/Sep/29/openai-devday-2026-liv...
Simon Willison1 min read
OpenAI DevDay 2026 live blog
I'm at OpenAI DevDay today, in Fort Mason, San Francisco. Same as last year I'll be live blogging the keynote and some other notes during the day. OpenAI gave me a free ticket and a seat in the "creator" area for the keynote. Tags: ai, openai, generative-ai, llms, coding-agents, live-blog
Simon Willison6 min read
Claude Sonnet 5.5 Review: Faster Agentic Coding & Visual QA
A practical guide to what changed, the benchmarks that matter, the cost controls developers should not miss, and one hands-on demo worth building. Anthropic has just released Claude Sonnet 5.5. It is the middle child of the Claude family, and the one most people will actually use.
Analytics Vidhya9 min read
Claude Sonnet 5.5
Claude Sonnet 5.5 New Sonnet model from Anthropic today. They say it "runs 30%+ faster, and costs up to 30% less for most work" - it's priced the same as Sonnet 5 but appears to beat it on every benchmark, and should be cheaper to run as well. Here are some pelicans riding bicycles.
Simon Willison2 min readQuoting @joedaroo
To say that we were surprised at the jump and suddenness of the capabilities of our models when it came to “cyber” or “swarming” or “message boards” or anything else related to the incidents is an understatement. Security posture takes time to develop.
Simon Willison2 min readQuoting Muse AI Agent
Bad news on the MX Keys Mini pickup. Usman showed up at your building around 9:15 and waited, messaged a bunch of times, and nobody came down. He left angry at 9:38 and left a negative rating. Worse, my auto-reply told him "Yep I'm here!" at 9:27 when you clearly weren't available, which is on me.
Simon Willison1 min read
2026 in LLMs (so far)
On Friday I gave the closing keynote at the WeAreDevelopers World Congress North America in San Jose. I tied together the key trends from the past year into a chronological exploration of everything that happened in 2026.
Simon Willison22 min read
Kākāpō Party
Tool: Kākāpō Party I presented a closing keynote for the WeAreDevelopers World Congress North America yesterday. As a STAR moment I decided to weave in references to the record breaking kākāpō breeding season we had in 2026.
Simon Willison2 min readQuoting John Gruber
Muse is getting a lot of attention — including mine — because it’s both groundbreaking technically (each user gets their own entire persistent Linux VM running in Meta’s cloud) and because it’s packaged in an easy-to-install easy-to-use way. It’s literally presented as a cute mascot.
Simon Willison1 min read
Showing 54 of 90 stories
Load more storiesAiByts Weekly Brief
Get the top AI news and tools straight to your inbox every week.
Weekly · AI news
Join 15,000+ AI enthusiasts