extr

33 ms·

511.
by extr 3y ago
Am I misunderstanding? The paper says: > We also carry out manual experiments on OpenAI ChatGPT-3.512 and GPT-4 models, where we interact with the models in multiple rounds of dialogue, trying to hint the correct solution. The models are
512.
by extr 3y ago
FWIW personally I use @staticmethod when I have small-ish "utility" methods that are only ever used in-class. Basically just a convenient way of organizing code and making the intended context clear.
513.
by extr 3y ago
What is the smart money doing for type checking in Python? I've used mypy which seems to work well but is incredibly slow (3-4s to update linting after I change code). I've tried pylance type checking in VS Code, which seems to wo
514.
by extr 3y ago
GPT-4 works great, super impressive, but I am significantly less impressed by OpenAI as a business. I'm still locked out of creating a work account because I used up the two accounts-per-phone number limit that they have hardcoded (wit
515.
by extr 3y ago
The amount of non-technical people who have very little intuition around the limitations of LLMs is actually fairly startling - and I like to think of myself of someone who is usually good at that kind of thing. The realistic/coherence
516.
by extr 3y ago
> To boost his acting career, Mr. Badoo relocated with his wife from Seattle to Los Angeles a couple of years ago. While home, in between watching the couple’s 1-year-old twins, Mr. Badoo sets up a ring light and does remote auditions fo
517.
by extr 3y ago
It would be totally fine from like a plot standpoint, but personally I would play BotW first. Or at the very least get like 50% of the way through. I think that game does a better job of introducing the open world Zelda concept and it'
518.
by extr 3y ago
For sure. When you first play BotW it's somehow like, oh yeah, I get it now, this is what Zelda was supposed to be all along. More than that even, it's like the platonic ideal of a video game.
519.
by extr 3y ago
I just played it for an hour. It's like a remix Breath of the Wild, which isn't a bad thing considering BotW might be the best game I've ever played. But because of that there's no initial mind blowing experience of lear
520.
by extr 3y ago
IMO if anything I am coming to the opposite conclusion. Yud and his entire project failed to predict literally anything about how LLMs work. So why take anything they say seriously? Can anyone name a single meaningful contribution they&#x27
521.
by extr 3y ago
I know this is true but it stings to see this as someone who has unfortunately done a lot of job hopping. It all "made sense" at the time, and looking back I wonder what I should have done differently, but at this point I really d
522.
by extr 3y ago
Yeah what I've found with Airbnb is these days you need to optimize way harder for host quality/reviews and less for your actual preference in terms of location, price, furnishings. If the place has a 4.95, you will get what you p
523.
by extr 4y ago
Really cool. My girlfriend is into interior design and I've been showing her projects like this as they come along: in her view these are some of the best results yet. A few first impressions though: - They need to add a filter to the
524.
by extr 4y ago
As a friend in the industry pointed out, regardless of the legality of releasing the actual song, this will be a huge boon for producers/songwriters. Imagine being able to produce a song and send an artist a demo of themselves. If they
525.
by extr 4y ago
I've run out of the mental energy to give Musk the benefit of the doubt. His capricious and poor decision-making is ruining one of my favorite products day and day. It's so frustrating to watch, a complete train wreck in slow moti
526.
by extr 4y ago
This is what ChatGPT is made for. I also like to learn concepts by analogy with concepts I already know. Before, you had to search for books and posts made explicitly with this in mind. "Javascript for Python Developers" or the li
527.
by extr 4y ago
I visited Paris last year and had a very cute midnight scooter ride with my girlfriend down the Seine. It was like out of a movie. But the ride costed like $17 each, it would have been cheaper to just get an Uber directly back to our hotel.
528.
by extr 4y ago
IDK. I don't really mind difficult games. Hades is an all time favorite for me. But in Elden Ring, I just didn't feel like the core gameplay loop was very fun or satisfying.
529.
by extr 4y ago
Same here. I bought it on the strength of the rave reviews and found it incredibly boring. The world has a cool vibe but the gameplay and mechanics are so so clunky. Coming from the latest Zelda game, which was kinda similar (open world, yo
530.
by extr 4y ago
> Is that assistance? ...Yes? The fact that it's basic and they already know how to do it is irrelevant. The law isn't "it's okay to give them advice as long as the advice is sufficiently generic and obvious".
531.
by extr 4y ago
Didn't they? Or was it a system error? Perhaps the LLM has no record of it because the user's sister's boyfriend called in with a different phone number and misstated the policy number. But the change was definitely supposed
532.
by extr 4y ago
Do you have direct experience with a call center where this was happening? This sounds like a ridiculous conspiracy theory. In every call center I've ever had experience in, median/mean hold time was a KPI that was relentlessly op
533.
by extr 4y ago
Bingo. The biggest problem with implementing LLMs-as-call-center-agents, at least in contexts like insurance, is fraud. Even GPT-4 is just too easy to fool currently. Call center conversations are often adversarial, where the caller wants t
534.
by extr 4y ago
Optimistically, it might be an improvement. Imagine a world where you can call in or chat. Hold time is zero in both cases. You never need to "transfer" to a new department, or supply the same information over and over. LLMs have
535.
by extr 4y ago
It's interesting, because I considered doing this to make it "fair". But after some reflection I noticed that I almost never type complete sentences into google. Google has kind of trained us to use it a certain way, which ca
536.
by extr 4y ago
I've noticed that when it refuses to answer it's good to "get it talking" about related subject matter, and then try to create a smooth transition toward whatever you wanted it to say/do.
537.
by extr 4y ago
I've been using ChatGPT as a health consultant for all sorts of things. If you've ever experienced health anxiety, you know a huge problem is Googling your symptoms and the top links immediately going to diagnoses like lymphoma or
538.
by extr 4y ago
To apply this to philosophy, Wittgenstein argued that many of the "classical" problems of philosophy, like the meaning of life, beauty, truth, etc, are basically misappropriations of language. The reason they are intractable is be
539.
by extr 4y ago
Honestly I agree, I think "fiduciary responsibility to shareholders" is kind of a funny phrase to be honest. I don't really believe most CEOs at an individual level think of shareholders much except that stock price up means
540.
by extr 4y ago
I'm of a few minds about this. I don't think these questions are unfair, because it's destabilizing mentally to deal with constant rumors, unclear direction, and actual layoffs. I worked at Meta during the Nov layoffs (though

More ›