tosh

17 ms·

511.
The Principal-Agent vs. Agents Problem(twitter.com)
1 points by tosh 2mo ago|0 comments
512.
A minimal chatbot template built with Next.js(github.com)
1 points by tosh 2mo ago|0 comments
513.
Everything hackable will get hacked(vercel.com)
3 points by tosh 2mo ago|0 comments
514.
UK extreme heat prompts Cobra emergency meeting(theguardian.com)
1 points by tosh 2mo ago|2 comments
515.
by tosh 2mo ago
ty for digging this up!
516.
Where good ideas come from (for coding agents)(sunilpai.dev)
2 points by tosh 2mo ago|0 comments
517.
by tosh 2mo ago
I was a bit suspicious of the url but it is also listed on llama.cpp github https://github.com/ggml-org/llama.cpp
518.
ChatGPT Desktop App for Linux(twitter.com)
23 points by tosh 3mo ago|0 comments
519.
Hetzner: Open Weights Inference API(docs.hetzner.com)
5 points by tosh 3mo ago|0 comments
520.
Mitchell Hashimoto's daily schedule given Superlogical, second kid, and AI usage(twitter.com)
2 points by tosh 3mo ago|0 comments
521.
Marc Andreessen on How Michael Ovitz and CAA Took over Hollywood(twitter.com)
1 points by tosh 3mo ago|0 comments
522.
by tosh 3mo ago
it's good to have even more (and more slim!) ways to run Clojure (not just JVM and JavaScript runtimes)
523.
Economics of Open versus Closed AI(a16z.news)
1 points by tosh 3mo ago|0 comments
524.
Xogot Comes to Mac [video](youtube.com)
1 points by tosh 3mo ago|0 comments
525.
Unsloth Desktop(twitter.com)
1 points by tosh 3mo ago|0 comments
526.
Nvidia Nemotron 3.5 Lightning(twitter.com)
5 points by tosh 3mo ago|0 comments
527.
PDCA(en.wikipedia.org)
3 points by tosh 3mo ago|0 comments
528.
Spotify Xirp(xirp.spotify.com)
2 points by tosh 3mo ago|1 comments
529.
by tosh 3mo ago
is this confirmed? this is a github issue with little context and only one comment
530.
Stealing Reasoning Traces from Proprietary LLM APIs(twitter.com)
2 points by tosh 3mo ago|0 comments
531.
Claude Sonnet 5 introductory pricing becomes permanent(twitter.com)
3 points by tosh 3mo ago|0 comments
532.
by tosh 3mo ago
also no need to trust these bench runs, you can just run your own (any OpenAI Responses API compatible endpoint works, if your endpoint does not support 'custom' tools you can have your agent change the smol implementation to use
533.
by tosh 3mo ago
if you're interested in an open source agent that uses a minimal amount of CPU and RAM and has source that is easy to audit: https://github.com/smol-env/smol here are traces from an agentic task around using duckd
534.
Unikernel(en.wikipedia.org)
2 points by tosh 3mo ago|0 comments
535.
helppeer.ai(helppeer.ai)
1 points by tosh 3mo ago|0 comments
536.
by tosh 3mo ago
interesting! it sounds like this might benefit from a ui that helps to edit/re-write the history I also think this would make sense for programming but there it is a bit harder to justify the effort when you are working on something th
537.
Peter Steinberger: "Fun Is Velocity" [video](youtube.com)
3 points by tosh 3mo ago|0 comments
538.
Clanker Constitution(github.com)
2 points by tosh 3mo ago|0 comments
539.
by tosh 3mo ago
prompt engineering 2025: you are an expert programmer, use industry best practices, test driven development and use modularity and abstraction to anticipate future features, … prompt engineering 2026: i believe in you
540.
Learning more about Claude's mathematical capabilities(anthropic.com)
282 points by tosh 3mo ago|177 comments

More ›