I’m hoping that it’ll be able to work well as a portable workstation for coding and whatnot.

Sounds like a challenge

CUDA arguably works better on Linux than Windows, so you just have to find someone who needs it for accelerated computing purposes.

It seems more likely to me that any bugs present in that code just became features that old software relied upon over time, so they can’t change anything without breaking backward compatibility.

So I guess, in a sense, it’s bug-free.

LLMs are still a facet of AI though. It sounds like they’re saying it shouldn’t be categorized as AI at all.

Or any task change really. You tell me that I’m here for a writing task, then halfway through it becomes a math test? There’s no way I’m doing anywhere near as well as if they told me what was happening ahead of time.

It wasn’t “many employees close to Altman”. It was the entire company, including the people who initiated the process of getting him kicked out. The whole thing made absolutely no sense.

To make the difference you’d have to peddle thousands or hundreds of thousands of individualized discounts

That’s what they do here in Canada.

The person who was shot probably had some non-zero mass.

And so much of it is basically impossible to do with scheduled calls. People have tried making video call software to replicate in-person interaction dynamics, but none of them helped for the labs I worked in.

• When you’re discussing in person, conversations naturally split up into separate conversations each with their own interested participants.
• If you’re no longer interested in something, you can temporarily tune out and listen in on a different nearby conversation to gauge whether you want to move to a different group.
  • Video call softwares have introduced breakout rooms, but you need to leave a room to enter another. You can’t just quickly scan the room with your ears and pick out another topic you want to talk about. You also can’t do this scan without signaling to current participants that you might not want to talk to them.
  • There was another that created a virtual 2D room where you can walk around and get proximity chat, but things just get too noisy because you don’t get directional audio.
• Calls require scheduling and lack flexibility. Compared to an in person conference where the plan is basically just being around other researchers and interacting with them for the whole day. If it were remote, you would have to book calls to fill up your day and you can’t just decide to drop something because you’re tired and need some time to recover. If you schedule a call, you’re expected to be at the call.
• Scheduled calls come with much higher expectations. You don’t generally go into a call to exchange three minutes of small talk and hang up. You’re expected to know ahead of time that this is someone you want to talk to and what you want to talk about with this person. In person meetings let you go through this small talk process with multiple people to find the ones who share your research interests.
• Body language communicates a lot too, especially in this search process. Get enough experience with this and you have a much higher chance of picking out people who are of a compatible form of neuro-spicy.
• Considering that so many researchers are neurodivergent in some way, planning specific activities helps a lot as social lubricant. Going out for drinks as an evening activity is fairly standard at all conferences I’ve attended. There’s one guy who always organizes morning runs. Sometimes, we chat over board games. Etc.

I’ll add that things should also fail gracefully. If something breaks, they should all revert back to working like the dumb equivalent. Dumb switches, dumb thermostat, etc.

I don’t know where that quadruped character comes from but it’s often used to signify autism.

That’s easy to do. You just check that the username exists. If someone enters a wrong username/password pair, you can still check that the username exists, but how do you know that the user intended to log in with that username? You would also have to check every other username to see if the password matches, and that can’t be done with a simple search because you need to compute a different hash for each user you check. Then if the username exists and the password also happens to match someone else’s password, then what do you report? Should you even report it? Because doing so reveals that someone had that specific password, and if the list of usernames is publicly available (which they often are, or could become public through a leak of some sort), then you can brute force over a small set of usernames to match them up.

I don’t drink coffee, my breakfast is usually a meal replacement shake, and I work from home on most days. I wonder what long series of extremely improbable events would have to happen for me to end up owning any coffee at all and then somehow having it end up in my shaker bottle while I’m on route from the kitchen to my office.

That falls into the “not purchase” category. Regardless, their point still stands. If it’s not worth it for you at that price and you still want to play it, you might as well pirate it now because the price isn’t changing.

Being built suboptimally for a sport doesn’t mean that you’d be bad at the sport. The skills you develop would still be there and be equally impressive on anyone.

One of the things I find so fascinating about powerlifting is how the technique varies as you go across weight classes. It becomes a whole different game for people with different bodies. I’d love to see all the different ways that other sports can be played too.

A colloquial “not huge” can be equivalent to a statistical “no advantage”. I don’t know if that’s what they meant though.

It’s especially nice to see a comment go from -5 to +5 after you do so.

Right? Nuclear arms are supposed to be on arm sleeves, not vests.

Bring the money to Canada and build it here!