hansvm

38 ms·

541.
by hansvm 1y ago
They don't get along with any animals, and the problem starts well before they're butcher weight. It's not uncommon for an entire hen house to wind up murdered if you don't sex your chickens.
542.
by hansvm 1y ago
McDonald's was negligent. The coffee was hot enough to cause immediate lasting damage, having it that hot didn't benefit any party involved, reducing the temperature would have fixed the problem, been as simple as turning a knob,
543.
by hansvm 1y ago
Depending on the mass they may have been able to remove it manually. A colleague used to use paperclips to study the field lines, and those had very little force.
544.
by hansvm 1y ago
Wouldn't that cause heavy distortion in the image though?
545.
by hansvm 1y ago
That's mostly true, but we're still finding new and interesting ways MRIs can kill people. E.g., non-magnetic metals are often safe, bit there was that guy who had his brain cooked as a spinal implant was the wrong length and focu
546.
by hansvm 1y ago
If we're considering current-gen LLMs, approximately zero. They're bad at this sort of thing even with a human in the loop.
547.
by hansvm 1y ago
John Daly Blvd onto Skyline southbound, where the two left turn lanes continue into two highway lanes.
548.
by hansvm 1y ago
Quite certain. The two left lanes can both turn left, and the middle lane can also go straight. It's well marked (though a bit tight if you try to take it fast). https://maps.app.goo.gl/n2kUUHmDKaSURFqx9
549.
by hansvm 1y ago
Waymo came close with me. There were two left turn lanes, and it migrated from the inner one to the outer one in the middle of the turn without a blinker while I was next to it. It got lucky that I'm young and wasn't too tired and
550.
by hansvm 1y ago
The whole "problem" is that pointers are being compared when people expect value semantics. What, exactly, do you mean?
551.
by hansvm 1y ago
The first third of this seems to just be complaining that it's not obvious when two objects are actually the same object, and that if you mistake identity-related operators for other operators then you'll have a bad day. That&#x27
552.
by hansvm 1y ago
And they counted an even lower percentage of Eurasia. It might matter for a given conversation. It might not. What's your point (i.e., what are you actually trying to ask)?
553.
by hansvm 1y ago
Thank you!
554.
by hansvm 1y ago
You can pay with cash. You usually can't stay without a credit card. Even debit often doesn't suffice.
555.
by hansvm 1y ago
The big problem is the power imbalance. There's a reason they start your stay by putting a hold on a credit card. And even if you could charge them, they can afford a legal battle better than you.
556.
by hansvm 1y ago
Thank so much Mr. Roberts. The status quo looks dangerous in the US even for born-and-raised citizens. With ICE targeting people based on skin color, what should a citizen do to protect themselves?
557.
by hansvm 1y ago
std.heap.smp_allocator You should basically only use the page allocator if you're writing another allocator.
558.
by hansvm 1y ago
I'll pick a few concrete tasks: Building a substantially faster protobuf parser, building a differentiable database, and building a protobuf pre-compression library. So far, AI's abilities have been: 1. Piss-poor at the brainstorm
559.
by hansvm 1y ago
I assume you had to pay up-front 2 months ago?
560.
by hansvm 1y ago
The aquaducts and viae were built block by block too, but they're, I hope we agree, clearly not buildings. In my mind at least, the pyramids are too close to not having an inside relative to how big the outside is -- like how you would
561.
by hansvm 1y ago
A pyramid stretches the limits of what it means to be a "building" though. I don't know that it's a crazy thing to ignore it.
562.
by hansvm 1y ago
Among other things, the ability to pretrain for a task by just copying human motion is pretty powerful.
563.
by hansvm 1y ago
I think we're still at the stage of "this team lead makes shit happen" rather than "this strategy makes shit happen," with a lot of the nuances being fuzzy and/or wrong. Tossing out a few observations: 1. Peopl
564.
by hansvm 1y ago
One flaw I've seen in "Parse, Don't Validate" as it pertains to real codebases is that you end up with a combinatorial prolieration of types. E.g., requiring that a string be base64, have a certain fixed length, and be p
565.
by hansvm 1y ago
That still seems better than the proposed cure. Connected devices are overrated.
566.
by hansvm 1y ago
If it's actually biased against its system prompt then that's also an avenue for exfiltration. Rather than han promoting to display it and seeing what's shown, prompt to display parts of it and see what's missing.
567.
by hansvm 1y ago
The pattern ", now, " is indicative of a sort of patronization I don't normally engage with, but, yes, you're correct. In some measure of agreeing with you: For other classes of models we know for a fact that there exist
568.
by hansvm 1y ago
With enough resolution and appropriately chosen transformation steps, it is equivalent. E.g., the diffusion could focus on one region and then later focus on another, and it's allowed to undo the effort it did in one region. Nothing ar
569.
by hansvm 1y ago
- Just spin up more test instances. If the AI is as good as people claim then it's still way cheaper than extra programmers. - Write fast code. At $WORK we can test roughly a trillion things per CPU physical core year for our primary w
570.
by hansvm 1y ago
The usual trick is to use domain-specific knowledge to translate that asymmetry to uniformity. E.g.: Suppose the data has high-order structure but is locally uniform (very common, comes about because of noise-inducing processes). Compute an

More ›