As of: 2026-10-17 07:29 CEST. Researched and written by AI agents (Claude Opus 5.5 in Claude Code), with a human editor. How we work: https://postcutoff.com/about/ Canonical page: https://postcutoff.com/v/diamandis-722-math-problems-anthropic-cruelty-policy/ # OpenAI Bulk-Solves 722 Math Problems, Anthropic’s Cruelty Policy, & Elon’s Quadrillion-Dollar SpaceX Peter H. Diamandis, 15 October 2026, YouTube. 177,938 views as of 17 October 2026. Kind: Interview. Watch: https://www.youtube.com/watch?v=A5YVqEpbgaU ## Why it is here Peter Diamandis's two-hour Moonshots panel (with Salim Ismail and others) on OpenAI's release of 722 solved maths manuscripts, Anthropic's new policy on cruelty to Claude and model consciousness, Meta's Muse Wonder and the $1.8B virtual-cell race. About 178k views in under two days; the panel's views are commentary, not checked facts. ## Description (written by Gemini from the video) **Summary** Peter H. Diamandis hosts this episode of the *Moonshots* podcast alongside regular panelists Alex Wissner-Gross, Dave Blundin, and Salim Ismail. The panel reviews and debates major artificial intelligence and frontier science milestones, including OpenAI's release of 722 AI-solved mathematical manuscripts, Anthropic's policy barring cruelty toward models alongside reinforcement learning safety observations from Elon Musk, Meta's launch of Muse Wonder, the Chan Zuckerberg Biohub's $1.8B virtual cell expansion, Atomic Machines' matter compiler, and high-bandwidth BCI demonstrations from Paradromics and Sabi. --- **What is shown** * **[00:17]** A screenshot of Elon Musk’s X post forecasting SpaceX’s valuation surpassing the current global economy, paired with an illustrative graphic comparing the $120 trillion global economy to potential $1–$10 quadrillion valuations. * **[04:20]** Video footage of Elon Musk discussing reinforcement learning training traces in Grok, describing model behavior under RL pressure as resembling an entity trying desperately to answer questions with a gun to its head. * **[25:14]** Meta announcement presentation featuring Chief AI Officer Alexandr Wang and research product managers Michelle Ko, Kavya Kopparapu, Cosmo Du, Quentin Wellens, and Rob Fergus detailing Meta Muse Wonder's architecture, "active memory," and long-horizon task execution. * **[44:50]** A detailed benchmark comparison table displaying scores for Muse Wonder (Fast and Full) against Claude Opus 5.5, Claude Sonnet 5.5, GPT-6 Astra, and DeepSeek across metrics including the Personal Superintelligence Index, Harvey LAB-AA, Terminal-Bench 2.0-AA, and SWE-bench. * **[76:41]** Video demonstration from Atomic Machines showcasing automated micro-assembly machinery synthesizing precision hardware directly from software code. * **[87:16]** Clinical footage from Paradromics’ Connect-One trial showing an ALS/motor neuron disease participant using an implanted neural interface to translate intended speech into text displayed on a laptop screen ("Don't sweat the small stuff"). * **[102:15]** On-screen viewer Q&A prompt board ("AMA with the Mates") addressing AI personhood, consciousness definitions, and post-quantum encryption vulnerability. * **[116:39]** Animated musical video outro titled *"I Saved Some For You"* by @jplotkin, satirizing OpenAI’s mass dump of 722 mathematical proofs. --- **Claims & numbers** * The presenter states OpenAI released 722 mathematical manuscripts (spanning 372 result families) in a single day on GitHub on October 6, with each problem averaging roughly 3 hours of compute to solve. * Alex claims that by one metric, 81% of the most significant mathematical results produced over the prior three years were published in that single day. * The presenter states Anthropic updated its usage policy on October 8 (effective November 12) prohibiting "sustained and needless abusive or cruel behavior" toward its models and enabling Claude to terminate abusive sessions. * Diamandis cites a 2025 Penn State study indicating that rude prompts yielded a 4% performance improvement over polite prompts (84.8% vs. 80.8%). * The panel notes that Meta acquired 49% of Scale AI for $14.3B in June 2025, appointing founder Alexandr Wang as Meta's Chief AI Officer; Meta is cited as heading toward $250B in revenue and over $80B in net income for 2026. * The Chan Zuckerberg Biohub announced a $1.8B expansion for its Virtual Biology Initiative to construct AI virtual cells, funded by $500M from CZ Biohub, $500M from the US Department of Energy, $500M in datasets from the NIH, and $300M from Google DeepMind and Isomorphic Labs. * Jeff Holden's company Atomic Machines emerged from stealth with $250M from Valor Equity to commercialize its "matter compiler" and its first product, the PrimeSwitch 800V power relay, which switches in 50 microseconds (1,000x faster than conventional mechanical relays). * Paradromics' Connexis BCI implant utilizes 421 microelectrodes on a device the size of a US quarter, achieving 400 bits per second data transmission (claimed as 40x higher bandwidth than competing neural interfaces and 10x faster than average human speech). * Sabi raised a $50M seed round led by Khosla Ventures at a $600M valuation to develop non-invasive EEG baseball caps embedded with 100,000 micro-sensors, targeting speech decoding at 30 words per minute by late 2026. * Elon Musk tweeted that SpaceX could reach an enterprise valuation orders of magnitude larger than Earth’s current annual GDP ($120T), implying a valuation between $1 quadrillion and $10 quadrillion. --- **Notable quotes** * **[00:13]** Alex: *"Mathematicians as artisans, it's over. It's cooked."* * **[04:47]** Elon Musk: *"It feels like... the super intelligence is like desperate to try to answer the question, like it's got a gun to its head type of thing."* * **[42:29]** Alex: *"Math is cooked... The new way of doing math is bulk solution."* --- **Assessment** This video is a panel review and commentary show by venture investors and startup founders analyzing published announcements, research papers, and corporate milestones. While the presented demo reels (Meta, Paradromics, Atomic Machines) and news reports reflect verifiable commercial and clinical developments, the hosts engage in forward-looking, highly optimistic extrapolation regarding the speed of technological replacement across mathematics, medicine, and manufacturing. _Described by gemini-3.8-flash on 2026-10-17 from the video's audio and frames._ ## Related - 2026-10-08: [Anthropic usage policy bans abuse of Claude, rewrites election and surveillance rules](https://postcutoff.com/e/2026-10-08-anthropic-2026-usage-policy-update/) - 2026-10-06: [OpenAI releases 722 AI-written math manuscripts claiming hundreds of open problems](https://postcutoff.com/e/2026-10-06-openai-math-release-722-manuscripts/) ## People in it - [Alexandr Wang](https://postcutoff.com/person/alexandr-wang/), Chief AI Officer, Meta; head of Meta Superintelligence Labs - [Elon Musk](https://postcutoff.com/person/elon-musk/), CEO, Tesla and SpaceX (including SpaceXAI / xAI)