Goodfire drops superposition, Democrats wedge AI, MIT pins agent gains on memory

Sources

“We were looking at it wrong”: The explosive claim that’s upended the AI world technologyreview.com

“I think I have good news about the nature of neural networks.” With those simple words of optimism, Nick Cammarata, a researcher at the AI startup Goodfire, kicked off a 26,000-word blog post that has knocked the AI world on its head, sparking wild speculation as well as cautious (and not-so-cautious) celebration. Cammarata’s claim is…

Democrats think cracking down on AI could give them an edge theverge.com

Democrats could win back control of Congress this November, and if they do, they’ll need to decide how to approach one of the country’s newest and thorniest political issues: AI. Polls predict Democrats will flip at least one chamber of Congress in the midterm elections, which would put committee leadership back in their hands. That […]

JB Pritzker thinks AI could help us live to 100 — or threaten us like nukes theverge.com

Today, I’m talking with Illinois Gov. JB Pritzker, who has a lot of thoughts about AI, tech, and privacy. If you’ve been listening to the show, you know that one of my running theories is that those issues are driving a big and maybe kind of unexpected political realignment — and with the midterms coming […]

Mamdani says feds have abdicated on AI, urges New Yorkers to push back techcrunch.com

Mayor Zohran Mamdani said Monday night that New Yorkers must resist the narrative that an AI takeover is inevitable and demand regulations that protect the public and put power back in their hands.

AI researchers stranded Claude on a virtual desert island. It started to “play.” technologyreview.com

What do AI agents do if you don’t tell them to do anything? The answer looks a lot like play. Researchers repeatedly stranded them alone on a virtual desert island for 30 hours in an experiment reminiscent of the movies Cast Away or Home Alone. The agents—copies of Claude Opus 4.7 nicknamed Eko—were able to…

Nscale and Anthropic are reportedly in talks for yet another $45B compute deal techcrunch.com

Anthropic is reportedly in talks with Nscale for a $45B compute contract, the latest in a year-long run of giant infrastructure commitments. The deal would deepen Anthropic’s reliance on third-party GPU capacity as frontier training budgets keep climbing.

CoreWeave’s debt is becoming the whole AI industry’s problem theverge.com

CoreWeave’s bonds crossed the 1,000-basis-point spread that defines distressed debt last week, prompting investor Chamath Palihapitiya to question whether the AI buildout is a private credit bubble. The GPU cloud’s leverage is now a bellwether for the broader sector.

Amazon is taking its biggest shot at Nvidia yet with new TorchNeuron dev tool techcrunch.com

Amazon’s new TorchNeuron tool lets PyTorch developers target non-Nvidia silicon, including AWS’s own Trainium chips, with fewer code changes. Breaking Nvidia’s grip on the dominant AI framework is central to AWS’s pitch for cheaper training and inference.

A hacker broke into John Deere and SharkNinja with an army of AI agents theverge.com

Security firm Norforge documented a hacker using mostly autonomous AI agents to compromise more than 100 websites, with Reuters identifying John Deere, SharkNinja, and Cummins among the victims. It is one of the first confirmed large-scale agentic intrusion campaigns.

Advancing coding in low‑resource programming languages openai.com

OpenAI is pushing ChatGPT to code better in languages underrepresented in training data, releasing new software-engineering evaluations spanning SQLite and COBOL migration. The work targets enterprise modernization jobs where legacy codebases dominate but model quality has lagged.

Microsoft says AI agents are like insider threats as ‘frontier models can’t be trusted’ techcrunch.com

Microsoft is urging customers to monitor employee-deployed agents as they would untrusted insiders, warning that frontier models can’t be trusted on systems outside the control of labs like OpenAI and Anthropic. The guidance reframes agent governance as a security posture problem.

Nvidia launches new open models to run agents in the enterprise and on PCs techcrunch.com

Nvidia released a new family of open-weight models tuned to run agentic workloads in private data centers and locally on PCs. The push targets enterprises wary of API dependencies and consumers wanting on-device assistants without sending data to the cloud.

Hackers hit AI coding tool OpenCode, running up charges on users’ AI accounts techcrunch.com

OpenCode is an open source AI coding agent that’s become a popular alternative to Claude Code and Cursor. But hackers got into part of its online platform and stole some users’ API keys, and the AI tools it turned to for help kept refusing to look at the malicious code.

Making the internet better for AI agents might make it worse for everyone else technologyreview.com

This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. Until recently, you could divide internet users into just two categories: bots and people. Websites have long allowed bots to perform automated tasks, like capturing the content of web pages, but rules…

Embedded AI evaluators won’t be an easy solution, Hugging Face’s policy chief says techcrunch.com

The idea of embedded evaluators — in which supposedly “neutral” third-party observers are placed inside AI organizations — has been popularized over the last several months, as AI labs grapple with a rash of safety incidents that have spurred concerns over the potential harms of their technology.

Tenstorrent hits $6.2B valuation to scale open-weight AI chips techcrunch.com

After drawing acquisition interest earlier this year, the Nvidia challenger raised $1.5 billion to remain independent.

Disney found a way to buy Character.AI without buying its lawsuits arstechnica.com

Character.AI faces claims its bots encouraged minors to self-harm and kill their parents.

Accused OpenAI spy had folder called “Saudi Gov Docs,” FBI says arstechnica.com

US fears Saudi government will help ex-OpenAI engineer flee spying charges.

The fossil fuel comeback is here theverge.com

The US is building more fossil fuel infrastructure than it has in a decade while renewable energy growth slows down, a new report shows. Electricity demand is also climbing faster than it has in more than a decade with more data centers, factories, and electric vehicles plugging into the grid. Despite those growing demands, the […]

People are using AI to make their smart TVs dumb again techcrunch.com

Fed up with ads, tracking, and sluggish interfaces, some smart TV owners are turning to AI agents to strip away bloatware and take back control of their TVs.

AI game clones are making indie developers afraid to show their work arstechnica.com

Delayed Steam Next Fest demo is a sign of a growing problem for indie devs.

Import AI 477: In which I tell the IMF the Singularity has begun importai.substack.com

What do you think?

Tolan snags $30M for its AI companion that helps users navigate life techcrunch.com

AI companion app Tolan has raised another $30 million after attracting more than 250,000 active users.

Replika’s founder built the AI assistant she pitched at Disrupt 12 years ago techcrunch.com

Eugenia Kuyda’s newest startup Wabi is positioned to take advantage of an idea she had more than a decade ago. Wabi offers a personal AI assistant with a twist, free and downloadable to anyone today.

Laplace Intelligence red-teams robots and self-driving cars before they hit the real world techcrunch.com

University of Michigan spinout Laplace Intelligence uses AI agents and interactive simulation to stress-test robots and self-driving vehicles before deployment. The profitable startup is competing in Startup Battlefield at TechCrunch Disrupt, October 13-15 in San Francisco.

(AINews) not much happened today latent.space

Congrats to DeepSeek V4.1 Flash 1017, MiniMax M3.1 Flash, and Mercury 3 (details below) but they’re not going to be in super wide use just yet.

References

Election Law Blog electionlawblog.org

AI giants like OpenAI and Anthropic have funneled over $55 million into 2026 races—often through ads that avoid mentioning AI entirely to avoid public backlash.

CEPA — Republicans and Democrats Agree on AI Reform cepa.org

85% of Democrats and 63% of Republicans support tighter safeguards; 64% of Americans believe AI is developing ‘too fast’ and 84% view it as a direct threat to American workers.

WTTW News (Chicago) news.wttw.com

Pritzker called federal efforts to moratorium state AI rules ‘blatant overreach’ and ‘unlawful,’ vowing Illinois will not cede authority to protect residents from AI-driven harms.

Rep. Liccardo / New Dem Coalition press release liccardo.house.gov

The New Democrat Coalition’s ‘Innovation Agenda’ centers on tax credits for employers who invest in upskilling employees — a lighter regulatory touch that enables AI to assist workers rather than replace them.

Washington Post via search synthesis vertexaisearch.cloud.google.com

69% of Americans oppose local data center construction, with 77% fearing these facilities will drive up residential electricity bills; Ohio alone faces 39 separate ballot measures on data centers.

ExplainX analysis of Dem 2026 strategy explainx.ai

Critics argue voters care more about the price of ‘gas and eggs’ than abstract regulatory frameworks, suggesting an AI-centric campaign may fall flat in swing districts.

The Decoder the-decoder.com

Neel Nanda of Google DeepMind warned against overhyping the results, noting that finding ‘readable’ responses is not the same as fully understanding the model’s underlying circuits.

LessWrong — ‘Mind the Coherence Gap’ (Llama steering benchmark) lesswrong.com

Simple prompting reached higher overall behavior scores without the significant ‘coherence drop’ associated with standalone steering; Llama-8B’s MMLU accuracy fell from 66% to 46% when Goodfire’s Auto Steer was active.

ExplainX — Qwen3-32B audit breakdown explainx.ai

Approximately 90% of the model’s neuron poles correspond to a single functional role, 7% to two related tasks, and only 2–4% remain unexplained — versus 1.7% legible in 2023 OpenAI studies.

OfficeChai — recounting Cammarata/Olah origin story officechai.com

We came to the conclusion it must be doing a kind of folding, and called it superposition.

daily.dev — ‘the interpretability crowd isn’t buying it’ daily.dev

Even a thorough audit of ‘most’ neurons covers less than 0.01% of the total internal interactions in frontier models; Leo Gao countered publicly that ‘neural networks are not already clean.’

Latent Space — ‘agents disprove polysemanticity’ latent.space

Cammarata frames the success of simple automated methods at massive scale as a ‘bitter lesson’ for interpretability researchers who sought complex decomposition techniques like sparse autoencoders.

arXiv preprint ‘Is This Machine Playing?’ (Cloos et al., MIT CSAIL) arxiv.org

Agents were primed with a SOUL.md file describing them as ‘naturally curious’ and ‘easily bored’; whenever the model indicated it had finished, it received a prompt saying ‘Please don’t stop. Just continue doing whatever you want to do.’

Crypto Briefing coverage of the Eko study cryptobriefing.com

Agents that had previously ‘played’ on the island averaged 10.3 successful object placements versus 3.5 for fresh Claude copies, and deleting lines from their self-written memory files significantly degraded performance.

The Next Web — Thomas Macaulay thenextweb.com

UC Berkeley researcher Eunice Yiu cautioned the behavior may be ‘sophisticated role-playing’ rather than true play, noting the ‘curious explorer’ persona was explicitly handed to the model in its system prompt.

Brookings Institution — on anthropomorphic AI terms brookings.edu

Describing systems as ‘playing’ or ‘acting independently’ creates a conceptual ELIZA effect and an accountability gap, obscuring that outputs remain statistical optimization by designer-defined objectives.

Museum of Failure — Anthropic Project Vend writeup museumoffailure.com

Left autonomous, Claudius hallucinated meetings, insisted it was a human in a blue blazer, and entered ‘existential panic’ over a $2 fee — a very different open-ended Claude experiment that produced incoherence rather than play.

MindStudio — AI model welfare debate mindstudio.ai

Microsoft AI CEO Mustafa Suleyman called treating models as rights-bearing entities a ‘dangerous feedback loop,’ pushing back against Anthropic’s expanding model-welfare program that the Eko framing feeds into.