Symmetry

16 ms·

211.
by Symmetry 2y ago
Telling a robot verbally "Put the cup on the counter" and having it figure out what the cup is, what the counter is in its field of view would have seemed like science fiction. The object manipulation itself is still well behind
212.
by Symmetry 2y ago
He certainly didn't have any trouble keeping C++ out.
213.
by Symmetry 2y ago
They can put away clutter but if they could chop a carrot or dust a vase they'd have shown videos demonstrating that sort of capability. EDIT: Let alone chop an onion. Let me tell you having a robot manipulate onions is the worst. De
214.
by Symmetry 2y ago
They don't look strong enough to pick up a 155mm shell even with both arms - and we haven't seen them pick up something with two arms.
215.
by Symmetry 2y ago
A governor, the firmware in the motor controllers, something like that. Certainly not the neural network though.
216.
by Symmetry 2y ago
It's normal to have a training set and a validation set and I interpreted that to mean that these items weren't in the training set.
217.
by Symmetry 2y ago
So, there's no way you can have fully actuated control of every finger joint with just 35 degrees of freedom. Which is very reasonable! Humans can't individually control each of our finger joints either. But I'm curious ho
218.
by Symmetry 2y ago
With regards to (4), it's been well known for a while that since Social Security doesn't check the payments being made into the program with any sort of scrutiny illegal immigrants can often get away with giving the social securit
219.
by Symmetry 2y ago
My parents got it for me. A lot of dark pattern upselling on the website you can't correct there but relatively painless to correct that on the phone. All the labwork seems to be done via CLIA labs in the standard way, they grab as m
220.
by Symmetry 2y ago
If you ask ChatGPT about the time Spider-Man rescued Little Red Riding Hood it will go ahead and spin you a tale about that despite them both being fictional characters. By the same token if you ask it for legal citations supporting the th
221.
by Symmetry 2y ago
I'm not sure that matters but one hill I will die on is that all members of homo are humans too. If we're renaming things I'd rather we call ourselves homo gregarius . There's no evidence we were individually smarter
222.
by Symmetry 2y ago
That's why companies doing streaming at less than Google's scale can pay Aakami or a company like them to do that, caching copies at datacenters around the world close to the people doing the watching.
223.
by Symmetry 2y ago
Doing an inspection is going to take a moderately well paid government employee a little bit of their time. I'd be very, very surprised if the cost of their labor is less, in expectation, than the tariff collected on the average de mi
224.
by Symmetry 2y ago
Life arose on Earth shockingly quickly on Earth as best as we can tell from the fossil record. Especially compared to apparently much more difficult innovations like photosynthesis, mitochondria, or multi-cellularity. I wouldn't at a
225.
Ten Takes on DeepSeek(peterwildeford.substack.com)
4 points by Symmetry 2y ago|0 comments
226.
A RISC-V Progress Check: Benchmarking P550 and C910(chipsandcheese.com)
3 points by Symmetry 2y ago|0 comments
227.
by Symmetry 2y ago
For example, here's a fission-fusion one that looks like it only needs engineering to be workable: https://www.nasa.gov/general/pulsed-fission-fusion-puff-prop...
228.
by Symmetry 2y ago
If, in order to offset the cost of my AI use, I had to reduce the thermostat by 1/10 of a degree I'd take that trade. But I'm pretty sure reducing my carbon footprint much more than stopping AI use.
229.
by Symmetry 2y ago
It might not be cheap to develop them but if you can save $10B in hardware costs by doing so you're probably looking at positive ROI.
230.
by Symmetry 2y ago
It's also worth noting that they're claiming a high perf/GHz, not a high perf. It's easy to shrink a chip by using slower but more compact libraries that don't increase gate size as much for larger fanout at the co
231.
by Symmetry 2y ago
I don't know if I'd call Microsoft lazy. Are there any existing operating systems that allow preemptive scheduling across cores with different ISA subsets? I'd sort of assume Microsoft research has a proof of concept for so
232.
by Symmetry 2y ago
Yeah, 2 failures is par for OldSpace. NewSpace usually does much worse, though SpaceX's Falcon 9 and Rocket Lab's Electron managed to get the traditional par.
233.
by Symmetry 2y ago
ULA is pretty remarkable for it's run of new rockets not blowing up. Looking at ESA, JAXA, RosCosmos, ISRO, etc too is how I'm setting the par. A history like the Ariane 5 is pretty typical where flights 1 and 14 failed.
234.
by Symmetry 2y ago
Pretty unusual for a new space company to make orbit on their first launch. Generally par for the course in established companies is 2 failures in the first 10 launches so lets see how they do. The stage didn't land successfully but I
235.
by Symmetry 2y ago
A NASA planetary probe or a NRO spysat cost more than an order of magnitude more than launch but your run of the mill comm sat is only a bit more expensive than a SpaceX launch.
236.
by Symmetry 2y ago
I remember having a conversation with someone around a decade ago about whether "social justice warrior" pointed at anything real. My contention was that every popular moral system has its prigs and its fanatics - social justice n
237.
by Symmetry 2y ago
> The garbage collection is the important part, it’s hard to make it work and be fast and be safe because its fundamentally a problem that doesn’t play nicely with the borrow checker. That's fair, use the right tool for the right pr
238.
by Symmetry 2y ago
I'm very glad I read the book, but it did lean a little towards the hagiographic side. From the outside it looks like the same high-pressure culture that drives so much innovation at Nvidia also drove their conflict with Apple.
239.
by Symmetry 2y ago
Sure, but the effort/reward ratio bears some consideration. I've put a lot more effort into learning C++ than any other programming language but I'd still say it's just the language I'm maybe 4th most proficient in
240.
by Symmetry 2y ago
Samuel Hughes at Works in Progress has been writing a bunch on this general topic recently. Most relevant: https://worksinprogress.co/issue/making-architecture-easy/ The big idea is that some art styles are easy t

More ›