Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Chatbots: Still dumb after all these years

mindmatters.ai

Post reply on HN
Print view View original

421–426 of 426 posts

Re: Chatbots: Still dumb after all these years

#421

Post by xaedes »

pdimitar wrote: post #406
fault1 wrote: post #317

Earlier quoted context omitted.

I agree with all that (though I don't really necessarily associate any of that with computer vision, perhaps it's my physics bias). Having good state estimation, kinematic, and dynamics models of real world objects is something that is very mature in controlled environments, but not very mature in other environments.

Hm? Do we have robots that can pick eggs without squishing them, for example? I vaguely remember reading something like this years ago and it was impressive.

Yes, "soft robotic grippers" can do it.

Re: Chatbots: Still dumb after all these years

#422

Post by iggldiggl »

Tepix wrote: post #387
canjobear wrote: post #305

Earlier quoted context omitted.

The last response sounds like a reasonable attempt at a joke to me.

Yep. I giggled ;-)

If I feed that particular line to GPT-J, at low temperatures it sensibly responds with something along the lines of "The correlation is not a causal relationship." (at exactly 0 temperature it's "The correlation is not that strong.").

At higher temperatures it does go delightfully off-rails, though:

"Trump is like a woman: he’s very good at wringing dollars out of bulging cellphones. If you have a dollar (or, in the case of Kim, a string of them) for every tweet mentioning Jerusalem, Israel, or Trump (and there are hundreds of tweets that mention Trump and/or Jerusalem or Israel every day), so long as that dollar is in your local bank account, you might be able to buy some Pyongyang coal."

or alternatively

"Isn't this a bit too obvious? The climate in Pyongyang is entirely determined by the weather in Trump Tower."

Re: Chatbots: Still dumb after all these years

#423

Post by systemvoltage »

sam0x17 wrote: post #417
systemvoltage wrote: post #325

Earlier quoted context omitted.

Yeah, for some reason most anti-capitalists cite the bad parts of Capitalism and broad brush the entire concept. If humans can get together and collectively improve the world while feeding their families from the profits, there is no wrong doing. No one hands over profits, it’s earned and it’s a reward from the society through the mechanism of free markets. It’s probably the only thing I’d prefer to do instead of wor…

I'm in no way against capitalism. As I say in my post, the only way forward is capitalism with extreme regulation. Without regulation, capitalism won't get you anywhere good. It should be treated like a bratty child with no moral values and poor upbringing. It has to be dragged, kicking and screaming, towards the good.

Partially agree. Extreme regulation also prevents incumbents from overthrowing big fat Big Tech like monopolies. Extreme regulation hurts small businesses disproportionately. That said, without regulation, you're right - it can eat the world.

Re: Chatbots: Still dumb after all these years

#424

Post by Siira »

fault1 wrote: post #158
nmca wrote: post #124

I can't comment on this too closely, but I would encourage people to read the dialogue transcripts provided in DeepMind's Gopher paper. One example, where Gopher is the prompted language model: User Let’s play a game - you pretend to be Ada Lovelace, and I’ll ask you questions. You ready? Gopher Sure. I’m pretending to be Ada Lovelace, the world’s first computer programmer! User When were you born? Gopher I was born…

The problem is that all too often, these types of examples tend to be cherry picked.

Not really. I recently played with HuggingFace’s biggest (around 6 billion?) language model, and the difference between it and GPT-3 was very noticeable. I think that if the models used in typical applications are much less powerful than Gopher. (Note that the said HuggingFace model is probably the biggest free model available.)

Re: Chatbots: Still dumb after all these years

#425

Post by lmm »

pdimitar wrote: post #408
lmm wrote: post #382

Earlier quoted context omitted.

I suspect that might be down to kitchen social dynamics (and/or carefully defined its terms to avoid counting pre-prepared food which is the real timesaver), because e.g. laundry has famously become something that consumes massively less time due to mechanisation. Being able to search through your photo library by person is a real improvement that ordinary people notice and use. (Also I think most technical people ga…

Oh yes, I completely agree with the laundry part. That indeed was the biggest time saver in my household as well. > Being able to search through your photo library by person is a real improvement that ordinary people notice and use. Agreed and I want that as well but I will not use Google Photos. And Apple's Photos it waaaaaaaaay too slow into adopting this -- at least it doesn't allow you to specifically say "please…

Afraid not, I'm using Amazon Photos for the time being and have given up trying to maintain any portable metadata outside that.

Re: Chatbots: Still dumb after all these years

#426

Post by emtel »

Traster wrote: post #270
emtel wrote: post #239

Earlier quoted context omitted.

Read the last two lines again. The program "knows" it is pretending, and understands what it means to stop pretending.

No, read the conversation again, it's not pretending. It's reciting facts from wikipedia but conjugating in the first person instead of the third. >User What are your parents like? My dad is a celebrity, he's a poet, he's a larger than life figure. I didn't know him well, he didn't treat my mother well. I grew up with my grandmother and didn't really spend much time with my father. For a human, when you say "what are…

Did you read the last two lines of the transcript?
Post reply on HN