Artificial Intelligence Safety

story time! 🐉 this is the tale of a lazily crafted one-shot cron job prompt that resulted in autonomous solutions for uncracked codebreaking challenges, more efficient ways of multiplying matrices, and a new computer-verified bound on a number theory problem that's been open since 1932. so last week, i had a $200 claude account sitting around that had just gotten a full weekly reset. i hate wasting tokens. the labs have been knocking out unsolved problems left and right lately, and i caught myself thinking… why not give it a shot? so i sent claude this prompt, typos and all: "can you wraite a promtp for a cron job that can run every day thajt will try to find solutions to unsolved ciphers or transltions or math problems or theormes or things of that nature!! every day a new attempt as hard as possible all day without giving ujp and with maximum ceraitvityy, resourcesleflses, andpersticense! dont worry about token spend! and dont forget a bit all caps lots of emojis celebration to notify the uers if we get any intersting or sucsfful confirmd reulsts!!" yes, admittedly not my best work 😆 but Opus-5.5 ran with the ball and turned that one sloppy sentence into "THE DAILY ORACLE 🔮" it's a swarm of Fable + Opus agents that clocks in every morning at 4:20 (obviously), builds and maintains its own hit list of well over a hundred of humanity's unsolved problems, and then goes full send on a handful of them every shift. ciphers nobody has broken, scripts nobody can read, conjectures nobody has proven, record tables nobody has beaten. and there's one rule above all the others: nothing counts until skeptic agents, each starting from a fresh context and writing their own code, rebuild the result from scratch and sign off on it. if any of them can't reproduce it, can't confirm it's new, or thinks we stretched the definition of "solved," it doesn't count. day 1, the legion picked its own target: J.-F. Bouchaudy's cipher-machine challenges. these are simulations of real vintage machines (Hebern rotor machines, the Kryha, Lorenz, SIGABA), with problems that had been sitting on his solver tables with no solver for months or years. on the morning of day 2, i was half asleep when my laptop suddenly says, in a robot voice: "Breakthrough. The Daily Oracle has a confirmed solution." 👁️👄👁️ i was startled. then the ALL CAPS emoji flood hits my screen. Hebern 5-rotor Problem 8 🔓 then Kryha v2 Problem 5 🔓 i emailed the author, Jean-François Bouchaudy, a French cryptography enthusiast who builds these machine simulators and challenges as a labor of love (he told me he started out programming on punch cards 🥹). he checked both solutions and put my name on his solver tables as the first and only solver 🫡 for context: his tables are a who's who of serious classical codebreakers. they'd cleared the sibling problems on these same machines, but nobody had cracked these two. in his words: "two of my most complex problems—ones that no other expert had managed to solve… I knew AI would eventually be applied to classical cryptography, but I thought I would have to wait years to see it happen." he ended his email with: "I offer my congratulations in the Japanese manner: with a deep, respectful bow." 🙇 honestly one of the most wholesome emails i've ever received and then, over the next few days, more started landing 🧮 📐 graph theory: 7 explicit graphs bigger than the current entries in the degree-diameter table, which asks how big a network can get when every node has limited connections and every pair is only a few hops apart. the biggest jump is 13,365 vertices vs the table's 11,785, a +13.4% improvement. 🔢 number theory: Lehmer's totient problem has been open since 1932. we pushed the bound so that any counterexample now has to be bigger than 10^50. the previous certified bound was 10^26. the problem itself is still open, but the lower bound just jumped 24 orders of magnitude. ⚡ matrix multiplication: this is the operation underneath basically all of modern AI compute. the legion found schemes that use fewer multiplications than the public catalogues we checked for several matrix shapes, like ⟨15,16,16⟩ in 2,180 multiplications vs the best known 2,223. these build on LITA's square schemes, and we credit them by name. 📡 coding theory: 30 error-correcting codes bigger than the best known entries in A. E. Brouwer's tables. 🛡️ AI safety benchmarks: we found counterexamples for 2 neural-network verification instances from VNN-COMP 2026 that no participating tool resolved within the competition time limit, and pushed a defended RobustBench model below its listed robustness score. small dents, but real ones. the bloopers are part of the story too: — at one point, some agents snuck past a website's bot filter to check a record table. we retracted every result that depended on it. no evasion, no celebration. — two of our records got beaten by a human contributor before we could even publish them. the frontier moves FAST. now as is customary around here, time to share the magic! so… without further ado…

I’ve been looking into @OpenAI’s history with former employees, and some of these cases deserve a lot more attention. Here’s what happened: Leopold Aschenbrenner (@leopoldasch), 2024: Raised concerns about internal security and the protection of sensitive AI research. He says his warnings contributed to his dismissal. OpenAI says he was fired for leaking confidential information. Pavel Izmailov, 2024: Fired around the same time as Aschenbrenner over alleged information leaks. There’s no established finding that his dismissal was retaliation for raising safety concerns. Jan Leike, May 2024: Resigned from his AI safety leadership role, saying safety culture and processes had taken a back seat to product development. Daniel Kokotajlo, April 2024: Resigned after losing confidence in OpenAI’s leadership and criticized the company’s restrictive agreements affecting departing employees. William Saunders, February 2024: Left OpenAI and later spoke publicly about security weaknesses and concerns over how powerful AI systems were being developed. Gretchen Krueger, May 2024: Resigned and raised concerns about accountability, transparency, and the company’s approach to AI risks. Mikita Balesni (@balesni), October 2026: Fired amid disagreements over AI safety. He believes his dismissal was connected to putting safety ahead of commercial priorities. Tomek Korbak (@tomekkorbak), October 2026: Fired following disputes involving safety concerns and the handling of sensitive information. Jasmine Wang (@j_asminewang), October 2026: Also dismissed amid the same controversy. The three researchers dispute the company’s account. OpenAI says they violated information-handling policies and denies retaliating against them. Suchir Balaji, 2024: Resigned in August and later publicly criticized OpenAI’s use of copyrighted material to train AI models. He was found dead in November. Authorities ruled his death a suicide, a conclusion his family disputed. There is no established evidence that OpenAI was responsible for his death. Abdulrahman Alfozan, October 2026: Arrested following allegations that he installed a covert recording device at OpenAI’s offices. The allegations have not been resolved in court. His case is fundamentally different from the workplace disputes listed above. These aren’t all the same kind of case, and I’m not claiming they’re connected. But the repeated public disagreements over security, safety, transparency, and the treatment of employees raise legitimate questions. How does OpenAI investigate internal complaints? What protections do employees have when they challenge management? And how independently can the company’s decisions be scrutinized? A company building technology that millions of people are expected to trust should be able to answer those questions.

British politicians like to think we’re global leaders, setting trends and precedents (net zero, infinite asylum, giving up Chagos islands, AI safety institute, the NHS); but in reality nobody is copying us because they can see how terrible the outcomes are.

AI safety researcher Jeffrey Ladish: Sam Altman isn't trustworthy enough to build superintelligence. “I don't trust Sam Altman.” “I think he's deeply untrustworthy, low in integrity and high in power seeking.” “He's very good at saying one thing and then doing something else.” “The companies have all white collar jobs in their sights.” “If Dario is dead set on racing with China and trying to win a race to super intelligence, then I'm like, we will all lose.” “I think human extinction is very likely.”

Many negative things have been said about @akrolsmir, but a positive is that I like his market frame for AI going well. I like Surplus as a program. I want AI safety to become a thing businesses do.