buildbot

40 ms·

631.
by buildbot 2y ago
Thank you!
632.
by buildbot 2y ago
Anyone know what sensor they are using behind all that nice Canon glass?
633.
by buildbot 2y ago
As an MS employee working on LLMs, that entire saga is super weird. We need approval for everything! Releasing anything without approval is quite weird. We can’t just drop papers on arxiv. There is no way running your own twitter, github, e
634.
by buildbot 2y ago
Without doing anything, I have access to GPT-4o in chatgpt and the api already (on a personal account, not related to work). Maybe I’m just super lucky, but it’s certainly not vaporware.
635.
by buildbot 2y ago
And your frame must be much stiffer, as now your tool is being used to apply force vs just lay down plastic.
636.
by buildbot 2y ago
The demo is impressive but I get very odd/cringe "Her" vibes from it as well.
637.
by buildbot 2y ago
Groq is serving an LLM from (100s of chips worth of) SRAM, so the effective bandwidth thus token generation speed is an order of magnitude higher than HBM. This would 3.5x their speed as well, it is orthogonal.
638.
by buildbot 2y ago
Excellent, now I can write docker files to run containers as VMs on FreeBSD!
639.
by buildbot 2y ago
I'm renting a box from Hetzner - 4x 22tb HD, 2x 1.5TB NVME SSD, Ryzen 3600, 64gb ecc ram... for 100$ per month! It's nice :)
640.
by buildbot 2y ago
Note - this is a pretty bad study. Quoting the top comment from Benito Baldauf Apr 26, 2024 at 08:43 AM: "The findings presented here leave little room for speculation. Instead the study authors suggest several potential links between
641.
by buildbot 2y ago
Shake a fleece jacket under the light. All those fibers are plastic, going into your lungs...
642.
by buildbot 2y ago
Exactly! Sniffing passwords out of coffee makers is hard to scale. Lots of tech needed/knowledge. Wrenchs scale linearly with people given wrenchs, and typically one does not need training to apply brute force with wrench. You may be a
643.
by buildbot 2y ago
If someone is targeting you that precisely they are sorting through your trash for a coffee maker, then I would posit you are already in deep trouble and they'd likely do something easier like wait for you to leave and insert physical
644.
by buildbot 2y ago
Unrelated to the article, but did Andruril pay the Tolkien estate for their name? “Flame of the west…”
645.
by buildbot 2y ago
Yeah them and Boeing would fit right in as Arasaka and Militech...
646.
by buildbot 2y ago
Even offered as a joke, this is wild "haha, so funny how we could just kill you all!" Well, he's probably a bit jealous of Boeing.
647.
by buildbot 2y ago
That's not any different to me, personally.
648.
by buildbot 2y ago
Patents are odd, is probably all I can say - https://patents.google.com/patent/US20230281046A1/en
649.
by buildbot 2y ago
At the risk of sounding stereotypically skeptical, how do we know there's not a person behind the AI curtain, geo-guessing?
650.
by buildbot 2y ago
Lol, if he is planning on training on his cars Mercedes/Daimler had that idea back in 2018. I built the demo for the idea using nomad and a bunch of Jetson/Drive boards, and gave a quick presentation to the new CEO (of Daimler!).
651.
by buildbot 2y ago
Sure, it's really simple - Implement a Pytorch module for the DropBack continuous pruning while training algorithm:
652.
by buildbot 2y ago
Good point,, in school the undergrad lab machines where very locked down and secure. In grad school, every prof had their own few machines in the lab that where shared, limited sudo access to me, I'm sure some shared the machines more
653.
by buildbot 2y ago
Shared lab machines come to mind! Often many people have sudo and sometimes there is just one shared account.
654.
by buildbot 2y ago
Isn’t this still actually wormable to some degree? Just among system you have access too. But imagine if any of those at all are shared…
655.
by buildbot 2y ago
Sadly, still fails my test of reproducing code that implements my thesis (Dropback Continuous Pruning), which I used because it's vaguely complicated and something I know very well. It totally misses the core concept of using an PRNG a
656.
by buildbot 2y ago
With things like these, I like to think how much money went into the development of the random PR assets like the weird progression diagram. Apparently, that money is better spent of these than say, someone on the Python Language Committee,
657.
by buildbot 2y ago
Why does Quantum===intuition?
658.
by buildbot 2y ago
Ah yes, install some random (closed source?) software
659.
by buildbot 2y ago
It’s interesting how it always the big vendors that seem to suffer attacks like this - is it the more positive angle that simply nobody important uses stuff like opnsense? Or more negatively, not enough people paying attention to catch thes
660.
by buildbot 2y ago
…Nobody said they were going to… Apple could request a sensor with the polarsens mask. It’s just not worth it, from a resolution & light gathering perspective. Big tradeoffs improvements in specific scenarios is not a path Apple has tak

More ›