swyx

36 ms·

871.
by swyx 2y ago
hmm this is a tough name - conflicts with Liquid AI https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...
872.
by swyx 2y ago
https://web.archive.org/web/20250415160251/https://www.theve...
873.
by swyx 2y ago
what did you like about it?
874.
by swyx 2y ago
i bet if we talked to a real low level hardware systems/chip engineer they'd laugh and take another shot at how we put them on a pedestal
875.
by swyx 2y ago
the point is oai is saying they have a viable Claude Sonnet competitor now
876.
by swyx 2y ago
JSON is just MCP's transport layer. you can reformat to xml to pass into model
877.
by swyx 2y ago
yep. we address it in the podcast. presumably this is just a recent discovery and can be post-trained away.
878.
by swyx 2y ago
references for all the above + added more notes here on pricing https://x.com/swyx/status/1911849229188022278 and we'll be publishing our 4.1 pod later today https://www.youtube.com/@latentspa
879.
by swyx 2y ago
also sometimes if you get it wrong you catch unnecessary flak
880.
by swyx 2y ago
is there a shorthand/heuristic to calculate pvalue given n samples and effect size?
881.
by swyx 2y ago
don't miss that OAI also published a prompting guide WITH RECEIPTS for GPT 4.1 specifically for those building agents... with a new recommendation for: - telling the model to be persistent (+20%) - dont self-inject/parse toolcalls
882.
by swyx 2y ago
very sorry for your loss. i've been similar in Seattle and it was tremendously alienating. I left as soon as I could. Obviously not a seattle-only thing but I think Seattle is so full of picture perfect couples (working 9-5, big tech&#
883.
OpenAI gets ready to launch GPT-4.1(theverge.com)
3 points by swyx 2y ago|0 comments
884.
Anthropic: Circuit Tracing and on the Biology of a Large Language Model [video](youtube.com)
6 points by swyx 2y ago|0 comments
885.
by swyx 2y ago
as far as i know anthropic has zero voice stuff?
886.
by swyx 2y ago
pretrain data + rlhf
887.
by swyx 2y ago
itd be nice if you prettified your json in the blogpost fwiw i thought the message structure was pretty clear on the docs https://modelcontextprotocol.io/docs/concepts/architecture#m...
888.
by swyx 2y ago
Made in California. next question
889.
by swyx 2y ago
theres just no mechanism to my knowledge to ban links sitewide, esp small sites (vs normies) like lmarena
890.
by swyx 2y ago
evidence for this assertion please? its highly unlikely and maybe a bug
891.
Rethinking Reflection in Pre-Training(arxiv.org)
1 points by swyx 2y ago|0 comments
892.
by swyx 2y ago
brb destroying some magnetic tapes because i can just put them on the cloud
893.
by swyx 2y ago
a fun thought exercise - what would have to happen to HN for this to come true? basically all the old guard have to age out and not pass on the reference?
894.
by swyx 2y ago
i think with small codebases like this is less about speed and more about education of essentials - i actually often encourage juniors to do small clones like this, feel proud, and then study the diffs with the at-scale repros and either fe
895.
Paradigm Shifts(joincolossus.com)
1 points by swyx 2y ago|0 comments
896.
by swyx 2y ago
agree that TFA's advice is not really useful when open source libraries like these exist
897.
by swyx 2y ago
(but also they make 2m/mo as 18 year olds...)
898.
by swyx 2y ago
IBM execs would beg to differ
899.
by swyx 2y ago
god we are so spoiled. this is a successful, pretty well known commercial project. it's now source available. you have plenty of resources to get context in 2 minutes. lets appreciate the big stuff, have some agency for the rest. usual
900.
by swyx 2y ago
why is that impt to you? just trying to understand the problem you couldnt solve without a C-like

More ›