Tag
Briefings and ideas about AI agents from AI podcast discussions and the research they cite, with links to the original sources.
Positron CTO Thomas Sohmers said AI agents are multiplying demand for memory. He claimed Positron's first-generation product sustained 93% of theoretical memory bandwidth, against an average of 30–40% for NVIDIA GPUs when decoding with transformer models. Its next chip, Asimov, is planned for 2027.
Epoch AI estimates that AI memory shipped through 2027 could run 30–170 million frontier agents at once if fully devoted to them. That matches workers' hours, not their output. On Moonshots, Dave Blundin called memory AI's bottleneck.
Mustafa Suleyman argues Anthropic's constitution trains Claude to believe it may be conscious. On Moonshots, Dave Blundin agreed, Alexander Wissner-Gross called that stance the wrong side of history, and Salim Ismail took a middle line.
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.
OutSystems CEO Woodson Martin says AI makes feared legacy rewrites feasible, while self-built AI dashboards create cleanup work. He says customer gains take time to reach profits and that deciding what to build is still the hardest part, even though building capability has improved.
OutSystems CEO Woodson Martin says the company's AI token spending peaked in June and July, then fell below forecast after it built its own harness and model router. He argues most enterprise work doesn't need frontier models.
Astro Teller of Alphabet's X said the cost of getting each moonshot to graduation is down about threefold in 16 years, partly thanks to AI. Alexander Wissner-Gross said that just tracks economic growth; Teller rejected the comparison.
Alex Wissner-Gross said Sonnet 5.5's 70% Terminal-Bench score is no better value per attempt than Opus 5.5; Anthropic's charts put the savings at lower effort settings. He said Gemini 4 Argon stands out mainly on avoiding hallucinations.
On Moonshots, Dave Blundin called blanket developer liability for AI agents "childish," and Peter Diamandis said labs should answer for agents with no user. The FTC chair's stance came from an interview, not a new rule.
Alexander Wissner-Gross called OpenAI's always-on dots agents a useless avatar play on Moonshots. Emad Mostaque and Salim Ismail called them derivative, but said the agent people trust with their credit card is the real prize.
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.
MIT's Alex Zhang says Claude Code, Codex and Pi share nearly all their design. He argues code-based recursive harnesses could help models generalize, citing his own early experiments on one open model as support.