dominotw

12 ms·

361.
by dominotw 4mo ago
Saltman prbly had to beg and bribe for this. imagine fable getting banned and this just going though. That would be like accepting defeat.
362.
by dominotw 4mo ago
sol has no soul
363.
by dominotw 4mo ago
current players in the space love the regulatory capture
364.
by dominotw 4mo ago
vibe versioning
365.
by dominotw 4mo ago
ppl buy iphones over cheap android phones . android phones can do everything that an iphone does ( and better)
366.
by dominotw 4mo ago
what is 'bad data' and 'perfect data' according to you?
367.
by dominotw 4mo ago
solution is to always do what it has seen in training data and how it was RL. But ai companies dont tell you that. so you have to reverse engineer its training and stick to to that. These are no general purpose machines. They are shipping a
368.
by dominotw 4mo ago
> Generally, you don’t want to punish people for making decisions. riff-raff cogs get fired for making bad decisions all the time. also if not punished for making decisions. how do execs ever get punished because all they do is make deci
369.
by dominotw 4mo ago
my comment from 12 yrs ago came true lol https://news.ycombinator.com/item?id=8517186 take that top responder > From philosophy? Are you kidding? There's simply no way AI is ever going to come from a bunch of people
370.
by dominotw 4mo ago
5 year olds and ai both have jagged intelligence. also its AI not "artificial code generation intelligence" . Ship is your view of the product to shoehorn into something specific.
371.
by dominotw 4mo ago
how do you have access to their financials? are you an insider? Edit: to the commenter below . It was widely reported that these companies were unprofitable 1 from last year. I am asking question to this specefic comment because they made
372.
by dominotw 4mo ago
starting y axis from 50 and 95 is a bit mileading
373.
by dominotw 4mo ago
There is even like an american flag flying high in the background
374.
by dominotw 4mo ago
> Am I right in thinking this is a tiny model which has been trained well to reason, and that's it? i remember karpathy mentioning in dwarkesh podcast. But is reasoning really possible without all the knowledge.
375.
by dominotw 4mo ago
standard compactions doesnt really distinguish between long term vs short term ephemeral facts ?
376.
by dominotw 4mo ago
i think he scrwed holes into helment to setup his go pro
377.
by dominotw 4mo ago
airbnb asked me why i didnt go iit in india and interviever soured after that. it was a chinese guy.
378.
by dominotw 4mo ago
oh yea skiing and mountian biking helmets have had mips for years
379.
by dominotw 4mo ago
safetymaxxing
380.
by dominotw 4mo ago
i worked with them at sears. They were always so smug holier-than-you douchebags. There were some good talented ppl but as group they were complete douchebags. They were not even smart not sure why they were so smug.
381.
by dominotw 4mo ago
"very good" 99 percent of time and hallucinating 1 percent makes the "very good" part untrustworthy.
382.
by dominotw 4mo ago
no one is making money side-hustling ai models. This is like reddit wet dream. get real, dont get scammed by ppl selling you these dreams.
383.
by dominotw 4mo ago
also helped my chronic acid reflux/lpr but its more of diaphragmatic breathing. its backed by research. using phones and laptops all day stuns your brain into shallow breathing all day.when i was kid i remember my dad taking naps in th
384.
by dominotw 4mo ago
i can almost tell then you have not done anything like this in production scale. context window size is irrelevant.
385.
by dominotw 4mo ago
spent way too much time looking for raw data for the graphs . is it in the article?
386.
by dominotw 4mo ago
article specifically talks about this. deepseek spending significant test time with worse results than klm
387.
by dominotw 4mo ago
meta has reallocated a significant protion of their staff to genrating this
388.
by dominotw 4mo ago
let me take down armchair analysis with my armchair analysis
389.
by dominotw 4mo ago
you mixed two random quotes from the article to create a strawman. ofcourse you knew what you were doing but disappointing that this was top comment.
390.
by dominotw 4mo ago
is there any proof that all these shenanigans impove agent performance

More ›