baq

12 ms·

361.
by baq 4mo ago
The nuance is not what they propose, but why, even according to them, they propose this. Honestly the proposals are appalling, but biosafety arguments are not immediately dismissible. Ultimate cyber threats we can handle by rewinding societ
362.
by baq 4mo ago
It isn’t about trust or no trust, it’s about having a capability to do stuff vs not having it. If Fable is the only model doing the right thing in your use case, your only choice is to use it or not. If the efficiency gain is 2x, it’s a hit
363.
by baq 4mo ago
Good to be an optimist, I envy you
364.
by baq 4mo ago
I don’t see anything in your argument that supports your thesis tbh, if anything, it supports the opposite
365.
by baq 4mo ago
Then don’t deploy it for Pete’s sake if you can’t guarantee basic privacy.
366.
by baq 4mo ago
> we believe [Mythos 5] is likely unable to fully and reliably automate R&D for frontier projects spanning multiple weeks this is good news, right? right...?
367.
by baq 4mo ago
it's a way to get infrastructure built up. the tax dollars pay for bootstrapping of the ecosystem. it's actually smart in principle if you think about it, but obviously there's room for abuse and outright fraud.
368.
by baq 4mo ago
accurate; your point being...?
369.
by baq 4mo ago
Interesting choice when LLMs remove one of the biggest barriers to entry for Rust which is you have to know almost all the rules to reliably make valid changes to Rust programs (which also helps LLMs at inference time when they make mistake
370.
by baq 4mo ago
> If you have any proximity to sales in the space you know that procurement and lawyers exist that have a full time job redlining purchase orders and agreements to the N-th degree. AWS billing.
371.
by baq 4mo ago
I see plenty if I care to look outside my usual echo chamber. There’s lots of ‘it’s just the next token prediction’ (in 2026 still!), but there are more sophisticated arguments like ‘it can’t be creative’, ‘it doesn’t think’, ‘it’s just pat
372.
by baq 4mo ago
It’s been steadily getting cheaper per token per problems solved even if you don’t define ‘unsolvable => infinity token cost’
373.
by baq 4mo ago
Globally available AI means cheap LLM pilots offshore means huge wage deflation in white collar jobs in the US. …assuming the AI pace of progress slows down enough that there are still meaningful white collar jobs to have.
374.
by baq 4mo ago
Why do you think it’s a waste? If you’re buying GPUs to rent them you’re almost buying a bond. If you’re leasing them, it’s even more obvious that you’re collecting the spread. The GPUs have a financial lifetime after which the business doe
375.
by baq 4mo ago
> IMO none of these things point to a AI being a bubble (over-hyped, demand does not match the stated value). I agree the demand is there, but hyperscaler capex is what now? 3% GDP? This is an absurd amount of money and people who questi
376.
by baq 4mo ago
> There are no dark GPUs. There are actually lots of GPUs in storage somewhere waiting for data center megawatts to put them in.
377.
by baq 4mo ago
Undervolting is not running at max utilization by definition almost. …but the real question whether you want to undervolt your asset if you’re renting it out is why bother? You probably expect to replace it anyway after it’s spec lifetime,
378.
by baq 4mo ago
Everything is a temporary situation on long enough timeframes, especially if it’s exponentially growing. Moore’s law which dictates that compute depreciates quickly has been slowing down a lot in the last few years, coupled with the explosi
379.
by baq 5mo ago
> AI systems will never learn to anticipate the future Citation needed. I don’t see any reason these systems shouldn’t be able to speculate; indeed some would say that’s all they do, even about the past.
380.
by baq 5mo ago
I’d like the thing to read my mail twice a day and tell me if I missed something important. Haven’t set it up because I’m horrified by the thought of it reading my mail. Doubly so if it decides to do anything other than telling me if I miss
381.
by baq 5mo ago
Nah, you just merge main back every day or so.
382.
by baq 5mo ago
I don’t think it said anything about countries
383.
by baq 5mo ago
They want to spy on the enemy, but they need to spy on everyone to know who is the enemy, obviously
384.
by baq 5mo ago
The thing is, I'm not talking about POSIX or CLI; I'm strictly speaking about the desktop GUI. macOS keeps going into a direction which is actively preventing being efficient with it, all the while making weird net-negative decisi
385.
by baq 5mo ago
Oh we definitely need batteries, too, regardless of fusion ever coming online or not.
386.
by baq 5mo ago
Writing with a pen or pencil has better learning outcomes than with a keyboard for neurological reasons. https://pmc.ncbi.nlm.nih.gov/articles/PMC11943480/
387.
by baq 5mo ago
The year of the Linux desktop. Meanwhile I’m stuck on macOS for work. Oh the irony.
388.
by baq 5mo ago
If anything I just got entrenched in my opinion. The second best option, maybe, would be to not accept any destination pointers and have the default and only possible behavior just like what you describe.
389.
by baq 5mo ago
You jest, but I’m actually convinced education-tuned LLMs are (today) the only way education outcomes can actually improve in the AI era. As is, students are leveraging them for doing homework which makes homework useless, you want and econ
390.
by baq 5mo ago
The exhaust blower triggering a safety stop is normal when the blower should be blowing but isn’t. If the blower keeps spinning after it’s spun up manually everything is now working as intended. If it stopped blowing the furnace would go in

More ›