uoaei

16 ms·

181.
by uoaei 11mo ago
Tangential question related to the example kernel: in GPU programming is it idiomatic/standard to initialize the out array as zeros rather than empty? are the performance savings negligible?
182.
by uoaei 1y ago
I missed the part about a boat?
183.
by uoaei 1y ago
This is exactly the kind of out-of-the-box thinking that will get us past some of the limitations of the current crop of AI architectures. Bravo to the authors.
184.
by uoaei 1y ago
To repeat myself: if there is no life to experience doom, then whatever happens still happens, but it is not "doom". In other words, doom is a moral construct. Morality only exists when a being draws a line between "good&quot
185.
by uoaei 1y ago
"Far be it from me"
186.
by uoaei 1y ago
I think the point is more to posit that our civilization will come to normalize AI as a ubiquitous tool extremely quickly like the other ones mentioned, and to analyze it from that perspective. The breathless extremist takes on both sides a
187.
by uoaei 1y ago
Humanist, maybe. The anthropic argument is tautological: nothing is a doomsday without there being someone for whom the scenario spells certain doom.
188.
by uoaei 1y ago
In tech circles there is a naive utopianism which comprises the "good intentions" currently paving the road to hell. I was just at a talk where they had "AI experts" judge a startup pitch for an AI call center company. N
189.
by uoaei 1y ago
My pet theory is that frontend devs have so little on the critical path, and are usually overqualified for the kind of work they do, that they keep reinventing these interesting paradigms for managing state in GUIs basically just so they ca
190.
by uoaei 1y ago
That's an interesting question, the question of whether engineering per se is strictly pragmatic. I personally think drawing a hard line between research and engineering is a misstep and relies too heavily on a bureaucratic kind of met
191.
by uoaei 1y ago
That's simply not true. Engineering can exist outside industry. "Stuff costs money" is not a governing aspect of these kinds of things. FOSS is the obvious counterexample to your absurdly firm stance, but so are many artistic
192.
by uoaei 1y ago
That's not engineering, that's industry. It's important to distinguish the two.
193.
by uoaei 1y ago
Physics prevents detected objects from jumping unrealistically. Current systems seem not to account for that at all, reacting to objects which appear and disappear spontaneously. Sensor fusion is exactly the solution to this: use a variety
194.
by uoaei 1y ago
That's ok, they're supposed to be. That's no excuse to rush a bad job.
195.
by uoaei 1y ago
This impulse to limit robots to the capacities, and especially the form factors, of humans has severely limited our path to progress and a more convenient life. Robots are supposed to make up for our limitations by doing things we can'
196.
by uoaei 1y ago
That is of course the ideal, however local economic conditions (low/subsidized water prices, etc.) may dictate a different path by incentivizing another way of doing things (open cooling systems, etc.).
197.
by uoaei 1y ago
What's unclear, exactly? Whether video documentation obtained privately belongs to the person who owns the camera and storage media? Because I think it's about as clear as it gets.
198.
by uoaei 1y ago
"stuff" being technically, legally, IP
199.
by uoaei 1y ago
Features present in all IDEs over the last 5 years or so are better and more verifiably correct for this task than probabilistic text generators.
200.
by uoaei 1y ago
Plex shows you both critic and audience scores from RT (IMDB also) and they indeed diverge consistently on the lines you suggest. In general I trust the audience scores a lot more because I'm trying to have fun watching movies rather t
201.
by uoaei 1y ago
What is the primary reason for hard-coding these kinds of things? Gotta imagine CScientists would have debated this during the design phase. I have some hypotheses, are any of these right? * correctness/verifiability analyses * securit
202.
by uoaei 1y ago
Love the latent pedantry of HN. No shade! Bravo/brava!
203.
by uoaei 1y ago
This is obviously the direction of persistent storage: SD meets SSD. Technological innovations are reliably predictable at this relatively high level. If you didn't see it coming, better get in quick before everyone else catches on!
204.
by uoaei 1y ago
GDP doesn't represent much about output so much as how much money people pay for what outputs. It follows directly from this that if you want to increase GDP, you start commodifying activities that previously were not measured in econo
205.
by uoaei 1y ago
The article says that but it's overcomplicating to the point of being actually wrong. You could, I suppose, argue that the big innovation is the application of vectorization to the chain rule (by virtue of the matmul-based architecture
206.
by uoaei 1y ago
Calling the implementation of chain rule "inventing" is most of the problem here.
207.
by uoaei 1y ago
It is remarkable that Transformer-based feedforward models can be appropriated for a whole lot, though at some point it should be considered a stretch to call such models "language models" when they are so heavily adapted to geome
208.
by uoaei 1y ago
I'm also a visual learner and my class on dynamical systems really put a lot into perspective, particularly the parts about classifying stable/unstable/saddle points by finding eigenvectors/values of Jacobians. A lot of
209.
by uoaei 1y ago
I encounter similar sentiments constantly across the internet. One thing we as humans have a lot of trouble with is understanding the scope of experience that isn't commented on in public forums. When we consider comments on the inte
210.
by uoaei 1y ago
Hiring bad engineers will always be a problem, even (especially) in the age of LLMs.

More ›