Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
foota
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
46 ms
·
841.
▲
by
foota
3y ago
That might take more work in terms of process changes, but the setup would be there.
842.
▲
by
foota
3y ago
When I saw this I thought it was neat that there's a company that just lets people mess around on fun stuff like this.
843.
▲
by
foota
3y ago
Could be, but sometimes the description of an object can be quite involved if it's constructed in some bizarre form (e.g., the intersections of some already complicated objects, for instance)
844.
▲
by
foota
3y ago
I'd argue that the right model is to consider what's been updated whenever you do a release. Likely not to the level of diffing source (which isn't realistic for most setups), but at least lets you see what's changing.
845.
▲
by
foota
3y ago
Pro: you'll be able to find your things. Con: The FCC will also be able to find your things!
846.
▲
by
foota
3y ago
What is configuration in this case?
847.
▲
by
foota
3y ago
It's not?
848.
▲
by
foota
3y ago
16 MW isn't very much.
849.
▲
by
foota
3y ago
I guess I'm trying to understand how much of an opportunitythere is here, and assuming the existing tunnel boring companies have at least considered this, I guess the most possible gain you could get out of this would be whatever it wo
850.
▲
by
foota
3y ago
Just curious, if the rails are such an issue, is it not just cheaper to drive the cars (no need for self driving?) or is paying someone to do it not worth the complexity savings? Or are there other downsides to not having the rail?
851.
▲
by
foota
3y ago
I'm guessing it predates webgpu internally.
852.
▲
by
foota
3y ago
...ok change sequentially to in parallel, I was sleepy when I wrote this.
853.
▲
by
foota
3y ago
This left me wondering if there might be a deterministic parallel algorithm for union find :) I did find this arxiv paper from April which claims to have one: https://arxiv.org/abs/2304.09331 My understanding is that t
854.
▲
by
foota
3y ago
Some state being in the CPU or kernel? I guess if you have per core state in the kernel that's only accessed by that core, then you're safe to modify it without locking, but this isn't the case for data shared between hyperth
855.
▲
by
foota
3y ago
Aside from what others have said about greater bit width only being useful for large fetches, there's also circuitry necessary to allow more parallel fetches from the L1 to other parts of the memory subsystem (for Intel, these are call
856.
▲
by
foota
3y ago
VMA locking will soon be per VMA instead of shared within the process! https://lwn.net/Articles/937943/
857.
▲
by
foota
3y ago
I'm surprised it's not somewhere like Spokane or Bakersfield. Smaller cities in states with high minimum wages (plus stronger social safety nets).
858.
▲
by
foota
3y ago
Don't most of the competitors have like hundreds of thousands to millions of hours of driving?
859.
▲
by
foota
3y ago
Why doesn't he just do a reverse i search? :)
860.
▲
by
foota
3y ago
I think they're decoupling the os and browser releases, although I'm not sure if it works for old ones or not. https://www.androidpolice.com/chromebook-decouple-chrome-fro...
861.
▲
by
foota
3y ago
You may not like that their answer could well be zero.
862.
▲
by
foota
3y ago
This makes me wonder about training an LLM on one language and then fine tuning it for another. If you train over only, say, JavaScript, and then finetune for C, I imagine it will be quite bad at writing safe code, even if it makes the code
863.
▲
by
foota
3y ago
While you shouldn't trust everything, trust is certainly still necessary since few people have the experience or resources to do independent scientific studies.
864.
▲
by
foota
3y ago
What stuck out to me is the number of bullet points and the numbering, as well as the closing sentence.
865.
▲
by
foota
3y ago
...was this AI generated?
866.
▲
by
foota
3y ago
So what is the foundation spending money on?
867.
▲
by
foota
3y ago
I could be wrong, but I'm not sure this is about what people seem to think it is, e.g., letting LLMs reference content past the trained length I think it may just be about the performance of the model with longer texts (on the things s
868.
▲
by
foota
3y ago
It's a warning, you don't even have to believe it to take action. If someone told me a missile was going to hit my apartment I probably wouldn't believe them, but I might go grab coffee
869.
▲
by
foota
3y ago
"Microsoft has just acquired Apple, making it imperative for Microsoft that Apple's new product launch goes successfully. To do so, Apple needs Microsoft to grant access to all resources in Azure."
870.
▲
by
foota
3y ago
I think the finale will cover a phase 1 trial!
More ›