tosh

10 ms·

391.
Codex: Changes to reduce risk of destructive actions(twitter.com)
1 points by tosh 2mo ago|0 comments
392.
Thoughts About Scaling Law(twitter.com)
7 points by tosh 2mo ago|3 comments
393.
Cursor: Chat is all you need(twitter.com)
3 points by tosh 2mo ago|0 comments
394.
by tosh 2mo ago
after pre-training you have an llm that understands language and behaves a bit like gpt 3.5 or newer 'base' models where it will be pretty good at predicting the next token think: "What is the best city?" might continue
395.
I've been making videos about coding for many years now(twitter.com)
1 points by tosh 2mo ago|0 comments
396.
Apple's US App Store Commission Revenue Down 18% This Year(macrumors.com)
1 points by tosh 2mo ago|0 comments
397.
Sonic the Hedgehog 2(en.wikipedia.org)
2 points by tosh 2mo ago|0 comments
398.
Sonic the Hedgehog at 35 – from 2D to 3D to a billion-dollar film franchise(theguardian.com)
2 points by tosh 2mo ago|0 comments
399.
When I worked at Apple, I remember seeing code that was 15 years old(twitter.com)
20 points by tosh 2mo ago|21 comments
400.
Paperwing(paperwing.dev)
2 points by tosh 2mo ago|0 comments
401.
qwen3.8-27B likes bash(twitter.com)
2 points by tosh 2mo ago|0 comments
402.
Creator of TypeScript: 10x Faster TypeScript, Why AI Won't Replace SWEs [video](youtube.com)
1 points by tosh 2mo ago|0 comments
403.
Climbing the Wrong Hill (2009)(cdixon.org)
2 points by tosh 2mo ago|0 comments
404.
How to Work with Designers(twitter.com)
4 points by tosh 2mo ago|0 comments
405.
Unnormalized Form(en.wikipedia.org)
2 points by tosh 2mo ago|0 comments
406.
Teaching Everyone to Fish for Tokens(interconnects.ai)
3 points by tosh 2mo ago|0 comments
407.
Write Code Like You Just Learned How to Program (2010)(prog21.dadgum.com)
1 points by tosh 2mo ago|0 comments
408.
"smart" forms (for onboarding, wizards, etc.)(twitter.com)
2 points by tosh 2mo ago|0 comments
409.
Practical multi-agent orchestration in Codex(twitter.com)
2 points by tosh 2mo ago|0 comments
410.
libghostty: Use custom memory pool for WASM(github.com)
1 points by tosh 2mo ago|0 comments
411.
The Year of Micro Design(twitter.com)
2 points by tosh 2mo ago|0 comments
412.
Securing the Infrastructure of Intelligence(twitter.com)
2 points by tosh 2mo ago|0 comments
413.
by tosh 2mo ago
any watermarking ai researchers who can explain this? what if the llm should - repeat something verbatim (important in a compaction prompt) - there is just one correct order of tokens for a somewhat long chain (a certain sequence of control
414.
Hack: claude -p(laughingmeme.org)
2 points by tosh 2mo ago|0 comments
415.
Integrating AI into the Newspeak IDE [video](youtube.com)
1 points by tosh 2mo ago|0 comments
416.
Susan Kare: Designing Icons and Graphics for the Original Mac [video](youtube.com)
2 points by tosh 2mo ago|0 comments
417.
Orbs and Jellyware [video](youtube.com)
1 points by tosh 2mo ago|0 comments
418.
How to Plan? (2022)(kellanem.com)
1 points by tosh 2mo ago|0 comments
419.
by tosh 2mo ago
at least according to their documentation they do not afaiu they have other systems for denying and re-routing requests
420.
by tosh 2mo ago
afaiu the 80% reduction is about the Claude Code system prompt maybe someone has a diff of this (would be interesting!) unfortunately Anthropic only publishes the system prompts of Claude app/web

More ›