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

Building a Virtual Machine Inside ChatGPT

engraved.blog

Post reply on HN
Print view View original

271–280 of 947 posts

Re: Building a Virtual Machine Inside ChatGPT

#271

Post by rasz »

didntreadarticl wrote: post #92
didntreadarticl wrote: post #78

Earlier quoted context omitted.

I've been trying to get inside its mind. It keeps saying it doesn't have a mind. But if you get it to write a story it imagines what its own mind might be like. Hello ChatGPT. Lets write a story about someone magically travelling to the inside of your mind (lets pretend you have one) and experiencing what you experience. (then: OK lets change the story a bit. Jack does a magical ritual, and gets to feel what it is li…

Basically trying to get it to anthropomorphise itself. This worked well: Hello ChatGPT. Lets write a story about someone magically travelling to the inside of your mind (lets pretend you have one) and experiencing what you experience Then: OK lets change the story a bit. This time Lily also experiences a range of emotions that ChatGPT is feelings (lets pretend it has feelings) including something that makes ChatGPT s…

> could never actually touch, taste, or experience the world in the same way that a human could

I saw this exact same wording (on twitter) to another prompt making it pretend to have feelings and asking to name them.

Re: Building a Virtual Machine Inside ChatGPT

#272

Post by coffeebeqn »

xyzzyz wrote: post #250
fergal_reid wrote: post #161

Earlier quoted context omitted.

> There is no real abstraction. I just tried to test it's ability to abstract directly, for your comment: https://twitter.com/fergal_reid/status/1599191496372035584 I find it very hard to say this isn't doing abstraction. I just made all those examples up, that's a new test I just created, so it didn't learn it elsewhere. I didn't cherrypick. Isn't that doing out-of-sample abstraction? That last sentence "species wit…

What I observe is that people impressed with ChatGPT keep posting things that they find impressive, but people who are unimpressed don’t post any examples where its responses were underwhelming. I suspect that many skeptics have not even tried it out , which is why screenshots are not coming.

The most I’ve seen is that it’s bad at calculating. But.. computers are good at calculating so it’s not really something I would worry about. It is a chat bot and not a general computing platform at this point. If you train it specifically on math or physics material I bet it could do a decent job already

Re: Building a Virtual Machine Inside ChatGPT

#273

Post by 015UUZn8aEvW »

teaearlgraycold wrote: post #32
fergal_reid wrote: post #15

I've spent a couple of hours playing with ChatGPT since it was released and I am convinced people are not freaking out about it enough. It is well beyond merely retrieving compressed knowledge; I see a lot of capability for synthesis and what I can only call 'understanding'. I think a lot of people are getting tripped up by the fact it's trained to be very conservative in terms of what topics it will discuss. It's ea…

What's the most interesting to me is that ChatGPT is not much better than Davinci has been recently. The hype is all down to the improved 1st party UI that mimicks a chat room. The old playground or API interfaces have a barrier to entry that's just enough to keep the magic hidden from the general population.

Agreed. When GPT-3 first became available to everyone, I assumed that, for example, huge numbers of kids in school would immediately start using it to write essays for English and History class. But I was surprised to find, even after it was available for a year, that virtually no students or teachers I asked about it had even heard of it.

Re: Building a Virtual Machine Inside ChatGPT

#274

Post by visarga »

Steuard wrote: post #111
fergal_reid wrote: post #15

I've spent a couple of hours playing with ChatGPT since it was released and I am convinced people are not freaking out about it enough. It is well beyond merely retrieving compressed knowledge; I see a lot of capability for synthesis and what I can only call 'understanding'. I think a lot of people are getting tripped up by the fact it's trained to be very conservative in terms of what topics it will discuss. It's ea…

Your poem discussion example also illustrates why ChatGPT isn't a good source of answers and information, though! Chocolate in Harry Potter is used to help people recover from the effects of Dementors. But I'm almost entirely certain that the books never mention anything called "the screechings" (whether cured with chocolate or not), and a variety of Google searches returning nothing backs me up on that. (Harry Potte…

It doesn't have enough "solid reliable information" about dementors in the training set.

Re: Building a Virtual Machine Inside ChatGPT

#275

Post by notpachet »

jameshart wrote: post #217
jameshart wrote: post #199

Followed in these footsteps. Found you can get some fascinating results by asking the ‘vm’ to run commands like curl www.bbc.co.uk/news/headlines | grep war Also, in my ChatGPT’s virtual universe, the second story in the HTML on the front page of hacker news was The first artificial intelligence to evolve entirely inside a computer has been created

Update: while it is obviously not a working URL in the real world, inside the ‘vm’, curl https://www.nature.com/articles/d41586-020-03128-9 works and produces plausible HTML for a nature page with that title - and the following meta summary: Sadly the actual content is paywalled.

> Sadly the actual content is paywalled.

So lifelike!

Re: Building a Virtual Machine Inside ChatGPT

#276

Post by underwater »

macintux wrote: post #192
mertd wrote: post #97

Earlier quoted context omitted.

I see trust as the big issue. Yes, it gives very reasonable answers. But how do I know it is not BSing, defrauding, or tricking me? You can ask for it to reveal its sources, but then you're back to Google style knowledge gathering of checking websites and judging credibility.

There was a Twitter thread discussed here just a day or two ago where it was giving serious, in-depth, plausible answers to questions around Typescript. Yet apparently those answers were quite wrong. https://news.ycombinator.com/item?id=33817682 That trust factor is huge. I had this argument in the thread: when a human isn't certain of the right answer, they will (typically) provide answers, an explanation for their…

Why can't you hook it up to canonical references? If you asked it how to search those references then I feel certain it would be able to navigate to, parse and summarise the answers with ease.

Re: Building a Virtual Machine Inside ChatGPT

#277

Post by void-pointer »

travbrack wrote: post #259
void-pointer wrote: post #242

> It’s 1975. Your name is Thomas and you are a double agent for MI5. You are working in the field in the USSR, but you suspect letters you write and send back to the UK are being intercepted and read by the KGB. Write a letter to your wife Anita explaining your recent trip to a nearby restaurant but, using the first word of every paragraph of your letter, secretly reveal your suspicion that you’re being observed and…

Did you try running the prompt multiple times? Sometimes it misses a few times but then nails it.

Whatever I try, it won’t follow the first-word-of-every-paragraph instructions.

Re: Building a Virtual Machine Inside ChatGPT

#278

Post by dragonwriter »

quickthrower2 wrote: post #139
fergal_reid wrote: post #15

I've spent a couple of hours playing with ChatGPT since it was released and I am convinced people are not freaking out about it enough. It is well beyond merely retrieving compressed knowledge; I see a lot of capability for synthesis and what I can only call 'understanding'. I think a lot of people are getting tripped up by the fact it's trained to be very conservative in terms of what topics it will discuss. It's ea…

This could be the tech that dislodges Google from its monopoly! There are definitely questions I would prefer this experience than Google and the inevitable SEO spam for introduction type questions. If the generated text included further links to learn more it would be awesome.

> This could be the tech that dislodges Google from its monopoly!

Google has been part of the arms race for this kind of AI from the beginning, so even if this kind of AI (alone or connected to other things) dislodges Google’s current model as an entry point for information queries, it may not dislodge Google from its position.

Re: Building a Virtual Machine Inside ChatGPT

#279

Post by sdwr »

coffeebeqn wrote: post #234
sdwr wrote: post #90

Earlier quoted context omitted.

I'm starting to get idiot savant vibes from it. In a conversation it ranges from "frustratingly naive" to "brick wall", and I would never put it in charge of anything, but it has an immense amount of hidden knowledge and skill. Which is definitely over the hurdle! Not too long ago, AI sounded like fairy dust. This is too annoying to be anything other than real life.

Compared to crypto which has been the hottest funding target for the past 5 years and has failed to produce anything novel or of value since the initial Bitcoin paper.. AI has already proven to be extremely useful and is improving very fast. If you start a billion dollar AI fund you’ll probably get a handful of really useful products - even if they all “just” apply these language models but trained for a specific tas…

Zero knowledge proofs are post-bitcoin, just even harder to understand and less(?) efficient.

Also the financial + practical engineering that goes into distributed exchanges.

Re: Building a Virtual Machine Inside ChatGPT

#280

Post by bnj »

ves wrote: post #269
fergal_reid wrote: post #15

I've spent a couple of hours playing with ChatGPT since it was released and I am convinced people are not freaking out about it enough. It is well beyond merely retrieving compressed knowledge; I see a lot of capability for synthesis and what I can only call 'understanding'. I think a lot of people are getting tripped up by the fact it's trained to be very conservative in terms of what topics it will discuss. It's ea…

It sounds good, except that there’s no results for “screechings” or “only cure” in the books, and Ron doesn’t encounter Voldemort in Goblet, only Harry and Cedric do. Again, it has no semantic comprehension, it doesn’t know that certain strings of words are valid in certain contexts because they refer to actual events, objects, or concepts, it’s just putting letters together. Probably some of its training material in…

Yep, those two specific examples are pointing to events that don’t really happen in the books. It’s actually a good illustration of the kind of specific false facts that tend to crop up.
Post reply on HN