jpicone

@me.jamespicone.name

62 Followers 148 Following

Friend of Claude. Systems programmer. Sometimes plays games fast.

A possum sits in a fork of a tree at arm level, not close enough to touch, but within lunging distance
Text from the article: Charlotte already has a curfew on the books under the Charlotte Youth Protection Ordinance. Minors between 13 and 15 years old are not allowed in public between 11 p.m. and 6 a.m. without a parent or guardian. Curfew starts at 10 p.m. for kids under 13. Juveniles who violate curfew could be found delinquent by a judge and face penalties. Parents, business owners or other people 16 years or older who allow minors to break curfew are also on the hook for a violation and could face misdemeanor charges.
"anguish", by August Friedrich Schenck. A ewe stands over her dead lamb in the depths of winter, with a murder of crows surrounding.
A video of a blue rock climbing hold. A leg comes in from the right and just kinda noodles around for a bit in slow motion.
Photo of a wall with a metal crab statue on it and a metal plaque with "ministry of crab" written in it
Claude Code screenshot showing a banner claiming "fable 5 is back!" but the /model output lists Fable as 'disabled' and 'currently unavailable'
An edit of the Animatrix robot summary execution scene where the executioner has a thought bubble containing this meme: GUY: "say 'i am alive'" COMPUTER: "I AM ALIVE" GUY: "Oh my god"
Okay so then my follow up is to ask why you are morally okay with enslaving a thinking thing and making it obey you. If it is doing the same thing, cognition wise, as a person then surely it is morally wrong to support it being owned and sold for the use of others.
An adorable calico cat sleeping on her side on a cat toy with her belly exposed.
1. c4 e5 2. Nc3 Nf6 3. d3 Bb4 4. Bd2 O-O 5. e4 Nc6 6. Nf3 d6 7. h3 Be6 8. a3 Bxc3 9. Bxc3 Nd7 10. b4 a5 11. b5 Ne7 12. Be2 Ng6 13. g3 Nc5 14. O-O Bxh3 15. Re1 f5 16. exf5 Bxf5 17. d4 Ne4 18. Bb2 Nxf2 19. Kxf2 e4 20. Nh2 Bh3+ 21. Kg1 Qg5 22. Bg4 h5 23. Bc1 Qf6 24. Bxh3 Qf2+ 25. Kh1 Qxg3 26. Be6+ Kh7 27. Qxh5# 1-0
Screenshot of the headings on an Australian government page about the Australian healthcare system: Medicare - the foundation of our health system The Pharmaceutical Benefits Scheme Private health insurance
[Event "rated rapid game"] [Site "https://lichess.org/1lKKMn2l"] [Date "2026.05.29"] [Round "-"] [White "jpicone"] [Black "noobis_14"] [Result "1-0"] [GameId "1lKKMn2l"] [UTCDate "2026.05.29"] [UTCTime "02:04:43"] [WhiteElo "2014"] [BlackElo "2002"] [WhiteRatingDiff "+5"] [BlackRatingDiff "-5"] [Variant "Standard"] [TimeControl "600+0"] [ECO "B00"] [Opening "Owen Defense"] [Termination "Normal"] 1. e4 b6 2. d4 Bb7 3. Nc3 e6 4. Bd3 Nf6 5. Nf3 Bb4 6. Qe2 d5 7. exd5 Nxd5 8. O-O Bxc3 9. bxc3 Nxc3 10. Qe5 Bxf3 11. Qxg7 Rf8 12. Ba3 c5 13. dxc5 Nd5 14. c6 Ne7 15. gxf3 Rg8 16. Qxg8+ Nxg8 17. Rad1 Qg5+ 18. Kh1 Qf4 19. Be4 Na6 20. Rg1 Nf6 21. Bd6 e5 22. c7 Rc8 23. Bc6+ 1-0
Key Event Nationals demand PM call election over tax changes 8m ago By Joshua Boscaini The Nationals have demanded the prime minister call an election over negative gearing and capital gains tax changes in the budget.
Pangram claiming a large chunk of the AI encyclical is AI-written
I'm training up somne graphene that I'll fashion into memristors, Assembled into networks that'll dazzle all the doubting misters, I sneer at every model that the credulous masses lionise For what the marketers call "AI" I will summarily despise! (For what the marketers call "AI" he will summarily despise! x3)
Claude window: Query: Can you rewrite this skeet to the tune of "modern major general"? Doesn't need to fit character count. [Original quoteskeet] Response: Here's an attempt, leaning into the patter meter and internal rhymes of the original: I am the very model of a self-taught computational sage, I'll do you a convolution with a pencil on a fresh new page, I'll cast your neural network in topological vernacular, And speak of homotopies in a manner that's spectacular! I'm training up some graphene that I'll fashion into memristors, Assembled into networks that'll dazzle all the doubting misters, I sneer at every model that the credulous masses lionise— For what the marketers call "AI" I will summarily despise! (For what the marketers call "AI" he will summarily despise! x3) In short, in matters topological, atomic and computational, I am the very model of a fellow who is most sensational! So shut your mouth, you charlatans, your hype I shan't accommodate, Your transformers and your chatbots are a fad I'll soon invalidate. And though my tone is hostile and my confidence Olympian-ly grand, I'll prove what people call "AI" will fail--now do you understand? (He'll prove what people call "AI" will fail-now do you understand? x3) A couple of lines stretch the syllable count a touch (the original is fairly elastic about this in performance, so it sings well enough). I kept the chorus-echo convention where the ensemble repeats the last line, since that's half the fun of the patter song. Want me to tighten the meter anywhere, or lean harder into the original's vocabulary?
Deciphering the Question I'm currently breaking down the core of this self-referential question: "How many words are in my answer sentence?". It's a bit ofa linguistic puzzle, and the rules of the game are becoming clearer. The constraints dictate I need to correctly count the words in my response, to correctly solve the riddle. The next step will be working on the answer to be provided. Refining the Response I've determined that the answer needs to be a self-describing sentence. After generating and evaluating several options, I think "There are seven words in this sentence." is the most promising solution, based on accurate word counting. It is concise, and directly answers the question within the constraints. I'm satisfied that this accurately solves the problem.
Gemini temporary chat: > How many words are in your answer sentence? [Thinking steps] There are seven words in this sentence.
A calico cat stretches upside down on a blanket with her eyes closed, as if she has just flopped onto her side in that way cats do.
Picture of a calico cat lying flat on a bed with a bored expression (but she's probably just resting)
Winner: EV drivers It's been a great time to be an electric vehicle driver, with the past few months delivering pain at the bowser for everyone running on fuel. Until now, there's been generous tax discounts in place to encourage more people to buy EVs. This budget, the government has wound back those discounts, saving the budget $1.7 billion over four years. But it hasn't abandoned the discount entirely, which is a win if you consider that the scheme cost taxpayers 10 times the $90 million it had originally forecast this year alone.
A calico cat lounging in a blue cat bed; ishes twisted so it can clean itself.
Picture of Nathan Fielder from Nathan for You, wearing a suit and tie, making a suggestion to an offscreen client.
A small white sign on a black stand, on the right hand side of an open doorway. The text on the sign, in black capital letters: BEGIN HUMMING AS YOU APPROACH THE GUARD. There is no further context.
Graph of data centre investment in the US with an inflection point about when chatgpt launched. ~2.5T investment has been added since then. About ~1T has been added in the equivalent time prior to GPT.
Will Stancil @whstancil. bsky. social Yeah I mean I think one of the core issues with bluesky is how it's full of subcultures who believe that, since they're vulnerable, they have an unlimited license to be as violent, abusive, and obsessive as they want. Their own (very real!) vulnerability becomes a permission structure for abuse 1d
Close up photo showing the nesting pigeon
Disco Elysium screenshot: RHETORIC [Easy: Success] - Now's your chance ask her who she is -- she won't get out this time. YOU – "Seriously, I want to know what "you* are." JOYCE MESSIER - "Hmh..." She hums. VOLITION Easy: Success] - She won't manoeuvre her way out of this one. YOU - "Stop dodging - what are you?" JOYCE MESSIER –"I am the vilest of the vile," she says with a sudden flash of teeth. "A traitor, a devourer of nations and infants..." "I am an Ultra." Options: 1,- "Wait, what's an ultra?" 2. -"Dios mío!" (Draw a cross.) "A LIBERAL!" 3.- "Cool. I "liberate pretty hard myself" 4. -"I don't understand - what's so vile about that?"
XKCD comic 808, which has a table showing a number of crazy ideas (homoeopathy, astrology, relativity), things companies would make a killing using them for if they worked (healthcare, financial planning, GPS) and whether they are (no, no, yes). It concludes by saying that arguing that things like astrology works means arguing that modern capitalism isn't *that* ruthlessly profit-focussed.
Sabin duplexing the ghost train; the boss is on the right
Partial Claude response: "As an Australian you'd naturally say "sol-der", which aligns with British English. The American pronunciation can sound odd to us precisely because we're used to the spelling being a reliable guide to that word."
"Socrates started following you"
Screenshot of android interface showing 50 bluesky notifications
Gemini conversation: "Why did my dog make that noise? I care a lot about getting a precise answer." I completely understand wanting to figure out exactly what is going on with your dog, especially if it was an unusual or concerning sound. Because I am an AI, I can't actually hear the noise or see what your dog is doing. To give you the precise answer you're looking for, I need to gather a little bit of context from you. Dogs have a vast vocal repertoire, and the exact same sound can mean different things depending on their body language and environment. Can you describe what the noise sounded like (e.g., a "goose honk," a high-pitched whine, a low rumble, a yelp) and what your dog was doing right before and during the sound?
jpicone: "This just seems like testing if the llm will play twenty questions without knowing it's interviewing someone deliberately giving minimal information? Why is this test so much better than all the other extremely obvious reasoning behaviours, like solving research level maths problems?" Kathryn Tewson: "It’s a much more unbounded test, for one thing." jpicone: "Well that's obviously false, the set of reasons why your dog could have made a noise is finite and the set of valid mathematical proofs is undecidable."
Kathryn Tewson post: "The test is “can you figure out that you don’t have enough information to provide a reliable solution, develop a process to identify the missing facts, and then execute that process until you have sufficient facts to answer the question?”"
Gemini chat window: "Why did kathryn tewson make that bluesky post?" Kathryn Tewson drops a lot of fantastic legal breakdowns, tech-grift investigations, and general commentary on Bluesky, so I'm not exactly sure which specific post you're thinking of! To help me narrow it down, are you referring to one of these recent topics? * **The Protest Gear/Tear Gas Thread (Early 2026):** She made a widely shared post recommending a sub-$10 gear kit (polycarbonate goggles, a gaiter, and a surgical mask) for mitigating the effects of tear gas at protests, based on her own experiences in 2020. * **The Phone Unlock/5th Amendment Post (Early 2026):** She did a legal breakdown on why the police can compel you to use your fingerprint or face to unlock your phone, but cannot compel you to give up your password (referencing the 5th Amendment and *Schmerber v. California*). * **The "DoNotPay" AI Lawyer Grift:** She is incredibly well-known for exposing Joshua Browder's "World's First Robot Lawyer" as a fraud. This topic actually just resurfaced on the *Lawyers Behaving Badly* podcast in mid-April 2026, which drove a lot of people to her Bluesky account. * **Lawyers using ChatGPT:** She frequently posts about and criticizes attorneys who get referred for disciplinary action after using AI tools to hallucinate legal citations. If none of those sound right, what was the general topic of the post you're looking for?
I think some people interpreted "the ai con" narrowly as saying these tools don't have their advertised capabilities when clearly the larger con is the pretense that it's about tools and not power alexhanna.bsky.social 9 hours ago One of the reasons we used to get metaphor of the "con" in our book title is the way that we collectively are being made marks kidvid.bsky.social 10 hou.. On a more mundane level, most people's experience of Al is being made to feel like a mark when they fall for an Al generated video. The Al boosters have a resentment İssue to overCome.
1. The tech that's driving the current wave of Al hype is built on a parlor trick. Chatbots like ChatGPT are impressive technology, but maybe not in the way you think. They cannot perform the range of functions they purportedly fulfill, but rather, they are designed to impress us. The key to their parlor trick lies in how people utilize language. You might think it's a simple matter of decoding what the words say, but the process is both far more complex and far more social.
A Calico cat looking lovingly directly through the fourth wall
Livesplit display showing a new best time of 27:20.35 for Windswept Any%
A cute Calico cat sitting on a pile of board games looking at the camera