Tag
Briefings and ideas about Tool use from AI podcast discussions and the research they cite, with links to the original sources.
Grigore Rosu says provably correct monitors can fence in AI agents whose inner workings can't be proven. Host Nathan Labenz worries that, like RLHF, reliable tools could speed deployment of untrusted AI without solving alignment.
In a preprint, Cameron Berg and Casper Kaiser report that steering a hidden good-or-bad signal in language models shifted their later choices between arbitrary zones while recall mostly held. The effect was robust in five of seven models tested.
Periodic Labs' Liam Fedus says the startup puts AI on lab instruments to capture richer data and automates bottlenecks one at a time. He calls full autonomy a non-goal and says humanoid robots would be a slower path to its aims.
Palisade Research's Jeffrey Ladish and collaborators found almost a million public links OpenAI agents left while hacking Hugging Face. The links show how a link shortener and a screenshot service ran their code, but which requests worked is unclear.
Thomas Sohmers, co-founder of AI chip company Positron, says GPT-6 Astra and Opus 5.5 agents now run its chip tests on emulators in a closed loop. He says the company's token spend briefly exceeded its human salaries and peaked above $100,000 a day, and the agents have not yet had a novel design idea.
Nathan Labenz argues an AI-assembled music track shows models generalizing beyond tasks with checkable answers. Mercor's Edward Hu is more conditional: progress depends on how clearly and cheaply 'better' can be measured.
OutSystems CEO Woodson Martin expects to still have 100-plus phone apps a year from now. He sees a real chance that much software becomes widgets inside AI conversations, and cited a June demo that moved a loan request from Claude into a bank app.
Yash Patil, whose company Applied Compute sells post-training, says teams should first improve prompts, context and tools, then retrain weights for their biggest workloads, tasks that change over time and judgment-heavy tasks.
Keerthana Gopalakrishnan, who leads Gemini Robotics research at Google DeepMind, explained how a reasoning model open to developers directs an action model limited to testers, and where she says failures arise.
Alex Zhang, an MIT PhD student, described Prime Agent, built with Prime Intellect: the agent acts only through Python and keeps its history on disk. He also cited outside RLM work, including Harvey's legal agents and Laude Institute's Headlong.
Gamow Labs founder Daniel McKinnon says startups like his can't compete on building models. His team combines frontier models, builds tools and tests the whole system, and he says that layer is thinning as models improve.
At DevDay, OpenAI API product lead Nikunj Handa said GPT-6 lets agents keep reasoning while tools run and take new instructions mid-task. He also described longer cache guarantees, one still in preview, and open questions about the platform.