{"slug": "the-long-detour-three-part-book-series-on-pre-gpu-and-low-compute-ml", "title": "The Long Detour: Three Part Book Series on Pre-GPU and Low-Compute ML", "summary": "A new three-part book series on pre-GPU and low-compute machine learning has been published, drafted by a multi-agent AI system that verified every fact against primary sources and reproduced experiments. The 176,659-word work covers historical ML milestones like CMU's 1976 HARPY speech system and includes 398 executed experiments, with all construction data publicly available on GitHub.", "body_md": "[▶ watch it being written](replay.html) — the tape, all 13,539 events, replayed\n\n```\nHARPY · SPEECH UNDERSTANDING · CMU 1976\nVOCABULARY ......... 1,011 WORDS\nNETWORK STATES ..... 15,000\nCOMPILE TIME ....... OVER 13 HRS, DEC-10 (KL)\nTHROUGHPUT ......... 80 X REAL-TIME\nPROCESSOR .......... 0.35 MIPS PDP-KA10\nSTATUS ............. IT WORKED.\n```\n\nThis book was drafted, revised, and fact-checked by an event-sourced multi-agent system — an author, an adversarial critic, a planner, and an arbiter — working over a corpus of ~5,000 primary papers, directed and assembled by its human author. Nothing here rests on trust in the process: the process is published. Every sentence traces to logged reads of the primary sources; every quotation was machine-verified verbatim against the paper it cites; every number in an experimental rerun comes from an executed, independently reproduced computation; and the full construction record — the tape — replays to exactly this text.\n\n| Words | 176,659 |\n| Sections | 117 |\n| Construction events on the tape | 13,539 |\n| Executed experiments | 398 |\n| Critiques filed and resolved | 175 |\n\nPreprint, text only — figures are forthcoming. Source, tape, lab, and\nharness:\n[github.com/doInfinitely/long-detour](https://github.com/doInfinitely/long-detour).\n\nEven the jacket keeps receipts: the cover concepts and the\n[design\ntranscript](https://github.com/doInfinitely/long-detour/blob/main/transcript-cover-design-onward.md) are in the repository, including the quote check that caught a\nmocked-up cover inventing archival telemetry. Jacket copy passes the same\nquote check as the prose.", "url": "https://wpnews.pro/news/the-long-detour-three-part-book-series-on-pre-gpu-and-low-compute-ml", "canonical_source": "https://long-detour.com/", "published_at": "2026-07-16 13:05:37+00:00", "updated_at": "2026-07-16 13:25:38.597276+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "ai-tools", "ai-research"], "entities": ["CMU", "HARPY", "PDP-KA10", "DEC-10", "GitHub"], "alternates": {"html": "https://wpnews.pro/news/the-long-detour-three-part-book-series-on-pre-gpu-and-low-compute-ml", "markdown": "https://wpnews.pro/news/the-long-detour-three-part-book-series-on-pre-gpu-and-low-compute-ml.md", "text": "https://wpnews.pro/news/the-long-detour-three-part-book-series-on-pre-gpu-and-low-compute-ml.txt", "jsonld": "https://wpnews.pro/news/the-long-detour-three-part-book-series-on-pre-gpu-and-low-compute-ml.jsonld"}}