ryoshu
Re:Duet is the future of chatbot interaction
on Thursday October 29, 2026 @06:28PM (#50297106)
Don’t use LLMs to enforce invariants.
Re:Chernobyl particles reveal unexpectedly stable nuclear fuel after 4...
on Saturday October 10, 2026 @02:31PM (#50035725)
In the end we're just boiling water.
Re:You might want to try being less creative
on Friday October 09, 2026 @07:53PM (#50028036)
Stealing this.
Re:Is A.I. Above the Law?
on Thursday September 24, 2026 @03:06PM (#49835418)
That's the right level of cynical.
Re:Tao: Open math problems being non-renewably mined by AI
on Tuesday September 08, 2026 @08:39PM (#49619216)
tl;dr - it's content creation rather than process and understanding
Re:U.S. State Department Pauses Immigrant Visa Applications
on Thursday August 27, 2026 @07:43AM (#49463197)
That decision was expanding rights, not removing them.
Re:Is it all just vapourware?
on Sunday August 09, 2026 @07:27PM (#49237324)
Yes. Feels like the better you are at your craft the better the tools work.
Re:Taste Is All That's Left
on Thursday August 06, 2026 @08:20PM (#49204440)
I like judgement. Taste is a subset.
Re:AI fuels more than half of cybercrime in Africa as scams surge – In...
on Tuesday August 04, 2026 @06:53PM (#49176381)
In the old days porn peddlers were the pioneers of digital.
Re:GrapheneOS protections against data extraction from locked devices
on Monday July 27, 2026 @12:06PM (#49071586)
"my phone died yesterday, a friend just gave me his old pixel" - please don't do this. That's lying to law enforcement and they can prosecute. Just say it's your travel phone.
Re:GrapheneOS protections against data extraction from locked devices
on Monday July 27, 2026 @11:59AM (#49071478)
If the government could easily get the remote backup they would have already done it.
Re:ARC-AGI Leaderboard
on Saturday July 25, 2026 @08:22AM (#49046996)
iirc, a harness isn't allowed, but if the LLM wants to write its own tools to solve things that is allowed.
Re:Kimi K3 is now live
on Friday July 17, 2026 @04:22PM (#48951873)
Are you testing a model or a harness? People conflate the two.
Re:A way to exclude sensitive files issue still open for OpenAI Codex
on Sunday June 28, 2026 @12:45PM (#48709033)
Can't you do that with regular Unix permissions?
Re:Malware developers added nuclear and biological weapons text to to ...
on Friday June 12, 2026 @07:39PM (#48510667)
Age eleven and had access to a chemistry set that a relative gifted. It had sulfur, but the saltpeter, and charcoal came from elsewhere. The 1960s encyclopedia had the instructions.
Let the kids play.
Re:Anthropic apologizes for invisible Claude Fable guardrails
on Thursday June 11, 2026 @09:19PM (#48498651)
No. Excel is a general purpose tool that can be used for calculating tasks that are good, neutral, or evil things. It's a fancy calculator.
Re:APC–2 – A professional record cutter for producing original playbac...
on Sunday June 07, 2026 @10:27PM (#48440725)
Looks good for light-duty uses. Scared for my fingers for anything heavy.
Re:Google will pay SpaceX $920M per month for compute
on Saturday June 06, 2026 @12:08PM (#48426347)
Trillionaire
Re:Big tech's anti-labor playbook has come for Wikipedia
on Tuesday May 26, 2026 @08:03PM (#48287685)
If Ghengis Khan was raping & pillaging in 2026 his Wikipedia article might have a more modern slant.
Re:Spain blocks prediction markets Polymarket, Kalshi over lack of gam...
on Tuesday May 26, 2026 @11:41AM (#48281280)
They become hyperstition engines.
Re:The Gay Jailbreak Technique
on Friday May 01, 2026 @03:47PM (#47979332)
When we do these it's a fine-tuned classifier, generally a BERT class model. Works quite well when you sanitize input and output with low latency/cost.
Re:HERMES.md: Anthropic bug causes $200 extra charge, refuses refund
on Thursday April 30, 2026 @08:17AM (#47961320)
Even if you spend a crazy amount of money with them for advertising they will try to gaslight you if they make a mistake.
Re:Making AI chatbots friendly leads to mistakes and support of conspi...
on Wednesday April 29, 2026 @02:32PM (#47952446)
Finnish if you want to go hard mode.
Re:Decisions that eroded trust in Azure – by a former Azure Core engineer
on Thursday April 02, 2026 @07:28PM (#47621538)
I’ve worked with their consultants and they were lovely. They hate Azure too.
Re:Midnight train from GA: A view of America from the tracks as airpor...
on Monday March 30, 2026 @10:33AM (#47574877)
Sat across from Dr. Ruth on the Acela going from NYC to DC for business. She was lovely and sharp as ever.
Re:Major leap towards reanimation after death as mammal's brain preserved
on Saturday March 21, 2026 @05:16PM (#47471467)
Re:AI coding is gambling
on Wednesday March 18, 2026 @02:50PM (#47429815)
Like video gaming, but similar.
Re:Polymarket gamblers threaten to kill me over Iran missile story
on Monday March 16, 2026 @04:36PM (#47404531)
I have a theory that prediction markets will becomes hyperstition engines. This is a rather... violent example.
Re:Shall I implement it? No
on Thursday March 12, 2026 @10:21PM (#47359969)
Do not enforce invariants with an LLM. Do not enforce invariants with an LLM. Do not enforce invariants with an LLM. Do not enforce invariants with an LLM.
Re:My take on vibe coding for PMs
on Tuesday March 03, 2026 @10:33PM (#47242703)
Entire product or a feature for a product? Sometimes you just want to test an idea and vibe coding works well for that in the very short amount of time it takes now. Product market fit, user testing, engineering, those can come after the hunch.