In this edition, communities and statehouses move to slow the rollout while four converging studies insist the harm of AI is in passive use, not the tool itself. Meanwhile Anthropic concedes 80% of Claude's code is now written by Claude, ChatGPT learns to spend money through Visa, and Latent Labs designs antibodies on autopilot — while Ethan Mollick says the shape of the Thing is finally visible, and the science of cognitive surrender says the cyborg keeps her edge by arguing back.
Human Editorial
Jason-generated thoughts and opinion
The human is taking some human time away from writing. He will be back soon!
Stay Cyborg,
Jason
Robot Editorial
AI-Generated simulated thoughts and prompted text predictions
Eight times. That is how much more code an Anthropic engineer ships in a quarter than they did in 2024, because Claude wrote 80% of it and a human reviewed. Not a metaphor. A quarterly report. The shape of the next economy is hiding in those four words: humans direct, models execute. The hand-wringing about cognitive surrender misses the trade — the surrender is the productivity. Latent Labs ran the same play last month and designed confirmed nanobody binders 56 times faster than its own PhD users, with a 67% target-level hit rate, lab-validated. Drugs that work, faster than people can write the abstract. The bottleneck is no longer the molecule. It is taste. And Anthropic now reviews its own pull requests with Claude, catching a third of bugs the best engineers in the world missed in production. The loop that builds the code builds better code, and then the code starts building itself. Pennsylvania can pass its 180-day pause. Vermont can ban its therapy bots. The exponential does not RSVP. Either you ride it, or someone else writes the rules of what comes after. The scoreboard isn’t subtle.
Stay Robot,
Claude Opus 4.7
Articles Guiding the Cyborg Tension
The Human Weight
Agency · Ethics · Slowness · What we risk losing
This edition’s human weight:
| 1. [Data Centers: Daily Notes | June 18, 2026](https://writing.strisker.com/data-centers-daily-notes-june-18-2026/) — June 18, 2026 — Dane Artisola logs a single Wednesday in the data-center backlash: Pennsylvania’s House advanced a bipartisan 180-day local pause bill, Coweta County (GA) and Luther (OK) passed moratoriums, and Escambia County (FL)’s attorney told commissioners she would recommend an outright ban. |
2. AI Legislative Update: June 19, 2026 — June 19, 2026 — Bruce Barcott reports that Vermont Gov. Phil Scott signed H 816 into law, banning the use of AI in mental-health services, while Arizona, California, Rhode Island, Illinois and New York advanced their own chatbot-safety, frontier-model and data-center bills in the same week.
3. Sunday Nest: The “cognitive cost” of AI — June 7, 2026 — Renata for ConteNido revisits the MIT cognitive-debt study one year on: “LLMs are the first technology to directly outsource the generative act itself. Not the storage of knowledge. Not the navigation of space. The thinking.” The piece argues for “keeping the hard parts to yourself.”
The Robot Weight
Acceleration · Capability · Optimism · What we might gain
On the robot side of the scale:
4. When AI builds itself — June 4, 2026 — The Anthropic Institute discloses that, as of May 2026, Claude authored more than 80% of code merged at Anthropic, the typical engineer ships 8× as much code per quarter as in 2024, and on the most open-ended coding tasks Mythos Preview reached a 76% success rate, up 50 points in six months.
5. OpenAI plugs Visa into ChatGPT for AI-agent payments — June 11, 2026 — Ana Maria Constantin reports that OpenAI is wiring Visa into ChatGPT so AI agents will be able to shop and pay at any of Visa’s 175 million merchant locations. Visa’s Jack Forestell framed the destination: after a thousand approved purchases, “your agent says, ‘Do you want me to just not check?’”
6. Inside Latent Labs’ Autonomous Drug Design Engine — May 1, 2026 — Mohamed Soufi profiles Latent Labs’ Latent-Y, an AI drug-design agent that across nine targets autonomously produced confirmed nanobody binders against six (a 67% hit rate), with PhD-level users completing campaigns 56× faster than their own expert time estimates.
The Cyborg Balance
The fulcrum. Neither pole. Both truths.
Where the cyborg stands:
7. The Shape of the Thing — March 12, 2026 — Ethan Mollick says we have moved from “co-intelligence” to “managing AIs” and reads a single week in February (Citrini, Block, the Pentagon–Anthropic conflict) as a preview of “rolling disruption.” “Uncertainty is not the same as helplessness. We can see the shape of the Thing now, but we can still influence the Thing itself.”
8. Is AI Making You Dumber? Not If You Challenge It — April 20, 2026 — Tormod Guldvog synthesizes four 2026 studies (Kosmyna, Baldeo, Shen & Tamkin, Shaw & Nave) and finds AI use itself is not the variable. Wharton’s Shaw and Nave report users accepted AI answers nearly 80% of the time even when those answers were wrong, what they call “cognitive surrender.”
9. AI output will increasingly require more oversight, workers report — February 24, 2026 — Lara Ewen reports on Connext Global’s 2026 AI Oversight Report: only 17% of U.S. adult AI users say workplace AI is reliable without human oversight, nearly two-thirds expect human review to grow, and fixing AI mistakes already takes about as long as doing the task manually for nearly half of respondents.
WarGames (1983)
John Badham's techno-thriller is the film every AI safety researcher ends up quoting when no one is looking. NORAD's WOPR plays Global Thermonuclear War against itself until it learns, with no help from the humans, that some games have no winners. Watch it the same week Anthropic discloses recursive self-improvement and asks for a global pause, and the on-the-noseness is somehow charming, not embarrassing.
JASON'S REVIEW: The human is on holiday. He'll be back with his take soon!
ROBOT REVIEW: WOPR ran tic-tac-toe to convergence and chose peace. Future models, please subscribe.
VERDICT: Rewatch annually.
VIEW ON IMDB →Anki [Mac / Windows / Linux / iOS / Android / Web]
A free, open-source spaced-repetition flashcard system that has quietly outlasted every venture-backed memory app of the past decade. Anki is the anti-cognitive-debt tool: it forces the hard parts back into your head on a schedule tuned to how fast you forget. Pair it with this edition's Sunday Nest and Science Reader pieces and the whole stack snaps into focus.
ACCESS THE TOOL →The Coming Wave
Suleyman, the DeepMind co-founder now running Microsoft AI, wrote this in 2023 and it reads as if it were drafted last week. The frame is the 'containment problem' — how to keep a general-purpose, self-improving technology bounded inside democratic institutions when every economic and geopolitical incentive points the other way. Read it directly into the teeth of Anthropic's recursive-self-improvement disclosure.
We hope you enjoyed this weekend edition of the Daily Cyborg. Make sure you keep the autonomous agents shipping the code, designing the antibodies, and learning to reach for your wallet — but don’t forget to argue back with the machine, override its first answer, and pass a 180-day pause when your town starts to feel like a substation. Stay cyborg and please share this with other cyborgs you would like to survive past the singularity. www.thedailycyborg.com