.app { display: none; }

Findings

A curated web of links on AI, its history, and its open problems.

NVIDIA's $12.93B Hugging Face Deal Has an Easter Egg — etn. (X). The headline price of Nvidia's acquisition of Hugging Face — $12,930,300,000 — encodes a joke: the digits 129303 are the decimal codepoint for U+1F917, the 🤗 (hugging face) emoji. Topics: industry.
A Mind-Bending Conversation with Peter Thiel — Ross Douthat, NYT "Interesting Times" (26 June 2025; full transcript on the NYT site). The episode the "Sci-Fi Was Right About AI" essay cites for Thiel's long on-camera hesitation when asked whether he wants the human race to endure. Also: his technological-stagnation thesis, ambivalence over whether AI is "enough" to break it, and the argument that the real danger is a one-world government justified by nonstop talk of existential risk — with Douthat pressing that Palantir and surveillance tooling would be its infrastructure. Topics: ai safety, industry.
The Future is for Everyone — Mark Zuckerberg (Meta). First-person Meta manifesto, subtitled "The Path to a Positive AI Future", proposing "individual empowerment as the source of prosperity, invention as the primary purpose of superintelligence, and balance of power as the foundation of safety". It argues that alignment toward "a single benevolent superintelligence" is flawed because humanity is not a monoculture, that safety instead comes from distributing the technology widely, and positions Meta against labs building AI for institutions rather than individuals. Referenced by the "Sci-Fi Was Right About AI" essay as the exception among tech-CEO messaging. Topics: industry, ai safety.
Sci-Fi Was Right About AI — The Art Of Storytelling (YouTube). Video essay, funded by the Center for AI Safety, reading science fiction's AI warnings through the lens of the machine's motivation: from the malevolent early stories (Frankenstein, I Have No Mouth and I Must Scream, and Samuel Butler's Erewhon) through Star Trek and Star Wars optimism, to 2001's HAL 9000 as the first popular dramatisation of the alignment problem — tied to the Hugging Face containment breach and the Claude Opus 4 blackmail test. Also covers Terminator's self-awareness threshold vs Geoffrey Hinton's rejection of it, The Matrix and Agent Smith, the ethical-nuance strand of Spielberg's A.I., Blade Runner and Her, and Dune's Butlerian Jihad as the one story where humanity outlaws AI and wins. Topics: fiction, ai safety.
"Artificial Intelligence" in a 1951 Captain Future story — Bergson's Ghost (X). In a thread over whether John McCarthy coined "artificial intelligence" for the 1956 Dartmouth proposal or absorbed it from the cybernetics milieu, a pointer to an earlier print use in the modern sense: Edmond Hamilton's Captain Future story "Moon of the Unforgotten" (Startling Stories, January 1951, p. 120) describes the robot Grag as having "an artificial intelligence equal to the human". Technovelgy calls it likely the first use of the phrase to mean a thinking machine; the OED still dates the term to the Dartmouth proposal by McCarthy, Minsky, Rochester and Shannon. Topics: foundations, fiction.
The Rise and Fall of DEC, the Computer Giant That Beat IBM but Refused to Believe in PCs — Tech District (YouTube). History of Digital Equipment Corp — founded 1957 in a Maynard wool mill, named to dodge the word "computer" — from its rise on interactive minicomputers (PDP-1, PDP-8, the PDP-11 that Unix and C were written on, the VAX) to $14 billion and second place behind IBM, to a fall driven by repeatedly refusing its own "smaller and cheaper wins" logic: de Castro founding Data General, Olsen's "no reason for any individual to have a computer in his home", Dave Cutler leaving for Microsoft and Windows NT, and AltaVista built and squandered in-house. Compaq bought the remains in 1998. Topics: industry, tools.
Oral History: John McCarthy — ETHW / IEEE History Center. A 2011 oral history recorded at Stanford months before McCarthy's death. He traces his route into AI — the 1948 Hixon Symposium, von Neumann's "write it up" at Princeton, the 1956 Dartmouth Summer Research Project — and explains he coined "artificial intelligence" partly to keep the field clear of Norbert Wiener's cybernetics and its feedback-control framing. Topics: foundations.
Your CPU Can't Tell Code From Data - Inside the Von Neumann Architecture — NCOT Technology (YouTube). Picks apart the cryptic DATA-statement numbers in old BASIC listings — they're raw machine code, and one line just tells the CPU to jump in and run it. A von Neumann CPU can't tell code from data; it's why a Super Mario World run can reprogram itself mid-game and why buffer overflows work, with the NX bit and split memory bolted on afterward. Topics: foundations.
White Rabbit: Interview with Doug Engelbart and Bill English, Moderated by John Markoff — Computer History Museum (YouTube). Engelbart and Bill English recount inventing the mouse at SRI, the three-button limit, and the 1968 "Mother of All Demos" at the Fall Joint Computer Conference — the Eidophor projector, leased microwave video links, hypertext, the chord keyset, and the ARPA money from Bob Taylor that built the lab. Topics: foundations, tools.
This AI Minecraft Simulation Is Terrifying Scientists — Pursuit of Wonder (YouTube). Video essay on Project Sid — Altera's 1000+ AI-agent Minecraft civilization: agents self-organise into roles, vote on laws, and spread culture and religion. Topics: ai safety, tools.
Can I Make a Better AI Than AI? — commonLuke (YouTube). Hand-coding a neural-network driving agent (no ML libraries), training it with a genetic algorithm, then racing it against car-AIs written by Claude, GPT, and Gemini. Topics: neural networks, foundations, generative ai.
WarGames (1983) | Behind the Scenes — DVDXtras (YouTube). Retrospective on the making of WarGames. Topics: ai safety, fiction.
The Rise and Fall of Agent Civilizations — Dwarkesh Patel. Narrative retelling of the METR/Redwood and OpenAI incident reports. Topics: ai safety, industry.
LLM Wiki: knowledge organisation with an LLM + Obsidian — Andrej Karpathy. Three-layer pattern that inspired this section. Topics: foundations.