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

Ask HN: Are people in tech inside an AI echo chamber?

news.ycombinator.com

Post reply on HN
Print view View original

411–420 of 755 posts

Re: Ask HN: Are people in tech inside an AI echo chamber?

#411

Post by tester457 »

bigger_inside wrote: post #300
sensanaty wrote: post #138

Definitely. The tech is impressive but anyone I've spoken to thinks of it as Cleverbot 2.0, and among the more technically minded I've found that people mostly are indifferent. Hell, IRL most people I know don't think much of it, though on HN and elsewhere online I see a lot of people praising it as the next coming of Christ (this thread included) which puts it in a similar tier as crypto and other Web3 hypetrains as…

I teach college, and in the beginning days, everyone was screaming about "the students will have chatGPT write papers now". Well, apart from the fact that chatGPT is really incapable of developing a thought, and also apart from the fact that half will fail to delete sentences like "I'm a language model, so I can't..." (insert gist of question here), it's painfully obvious if something is LLM generated. The moment a s…

> "it's crucial to remember"

> "I'm a language model, so I can't..."

You won't catch the clever students who programmically remove these (e.g. using Langchain).

Re: Ask HN: Are people in tech inside an AI echo chamber?

#412

Post by bondarchuk »

Obscurity4340 wrote: post #66
samuellevy wrote: post #31

There's definitely a few echo chambers around AI, but it's definitely not something that "just techies" are onto. ChatGPT made some waves at the end of last year. My in-laws were wanting to talk to (at) me about it at Christmas. There's plenty of awareness outside of the tech circles, but most of the discussion (both out and in of the tech world) seems to miss what LLMs actually _are_. The reason why ChatGPT was impr…

Not to be difficult but wouldn't "confabulating" be a preferable description for this behaviour? Hallucinating doesn't quite feel right but I can't exactly articulate why confabulate is superior in this context

"Hallucinating" (normally) means having a subjective experience of the same type as a sensory perception, without the presence of a stimulus that would normally cause such a perception. I agree it's weird to apply this term to an LLM because it doesn't really have sensory perception at all.

Of course it has text input, but if you consider that to be equivalent to sensory perception (which I'd be open to) then a hallucination would mean to act as if something is in the text input when it really isn't, which is not how people use the term.

You could also consider all the input it got during training as its sensory perception (also arguable IMHO), but then a proper hallucination would entail some mistaken classification of the input resulting in incorrect training, which is also not really what's going on I think.

Confabulation is a much more accurate term indeed, going by the first paragraph of wikipedia.

Re: Ask HN: Are people in tech inside an AI echo chamber?

#413

Post by lhl »

I think it's too easy to fall into the anecdote trap, and I like the comments with study data, like the Pew Research and Line Research (sequential surveys, even better)...

I also think adoption will depend on "industry." ChatGPT in education for example:

* BestColleges [1] (n=1000) found 43% of college students have used ChatGPT, and 22% have said they used it to help complete assignments or exams.

* Study.com [2] (n=1100) had some crazier numbers. "Over 89% of students have used ChatGPT to help with a homework assignment. ... 48% of students admitted to using ChatGPT for an at-home test or quiz, 53% had it write an essay, and 22% had it write an outline for a paper."

* Interestingly, in K-12, adoption appears to be higher by teachers than students [3]: "Within two months of its introduction, a 51% majority of teachers reported using ChatGPT, with 40% using it at least once a week, and 53% expecting to use it more this year. Just 22% of students said they use the technology on a weekly basis or more."

* In Japan, a recent survey [4] (n=4000) of undergraduate students conducted from Tohoku University showed 32.4% have used ChatGPT. This is compared to about 7% office workers in Japan using ChatGPT on the job [5] (n=13814) in a recent poll by MM Research Institute.

Of course, education isn't the only industry with an outsized impact (although it's interesting in the sense that it's a good temperature check for the upcoming generation, especially if it's something that is so prevalent at colleges/universities).

But there are other industries as well. a16z Games has been doing a game development survey on Generative AI use in games, and their preliminary results [6] are in-line with my personal experience/view into the game industry - that it has already been completely re-aligning/disrupting the production pipeline: "We heard from 243 game studios - large and small - the results were astonishing ... 87% of studios use an AI tool or model in their studio TODAY. 99% of studios PLAN to use the technology in the future."

I think it's worth noting, that while the sources are dodgy, that even if the 100M user number is accurate for ChatGPT, that's only ~2% of global internet users, and do think this is going to change very rapidly however, but no matter how fast it goes, it won't be overnight.

[1] https://www.bestcolleges.com/research/college-students-ai-to...

[2] https://study.com/resources/perceptions-of-chatgpt-in-school...

[3] https://www.waltonfamilyfoundation.org/chatgpt-used-by-teach...

[4] https://www.asahi.com/ajw/articles/14927968

[5] https://asia.nikkei.com/Business/Technology/Use-ChatGPT-at-w...

[6] https://www.linkedin.com/posts/troykirwin_ai-x-game-developm...

Re: Ask HN: Are people in tech inside an AI echo chamber?

#414

Post by ulfw »

AI is a profound change for how we will live. It's more a socioeconomic change (loss of millions of jobs) than a technological advancement.

The worst part is the current state though, where all the crypto hype bros have come out of the woodwork and are rebranding as AI hype bros now. My Linkedin feed is sadly now full of idiots calling themselves "AI Expert" and similar such terms with made up resumes trying to sell their 'expertise'.

Re: Ask HN: Are people in tech inside an AI echo chamber?

#415

Post by bobthepanda »

DonHopkins wrote: post #357
runlaszlorun wrote: post #233

Earlier quoted context omitted.

Everywhere?? That’s exactly the hype talk that’s going to burst this bubble. Here’s tech’s dirty little secret. Despite all the screams about automation and universal basic income… the exact numbers where job replacement would show are in the labor productivity numbers. If GDP stays flat or grows while the number of jobs is reduced… bingo… you’d see that number climb. Productivity has actually stayed flat or gone dow…

Maybe you live in a monolinguistic bubble, or speak every language fluently, but when is the last time you hired a human being to translate a language you don't understand instead of using AI like Google Translate? AI is already so ubiquitous and useful that you blindly take it for granted without even thinking.

Most companies will at least hire a human to check translation work, or at least contract one. Bad translations are something that can quickly destroy a brand in a market, and translation tools are not so fine grained to take into account dialect.

Spanish, for example, has many examples of words that are innocuous in one dialect and profane in another.

Re: Ask HN: Are people in tech inside an AI echo chamber?

#416

Post by pksebben »

janoc wrote: post #326
andsoitis wrote: post #291

Earlier quoted context omitted.

> The era of having a personal voice assistant that is capable, adaptable, and intuitive is VERY close and that is something that's exciting. Intuitive to use? Or has intuition?

And that actually anyone wants? Google and Amazon have tried to sell theirs for a long time. And none were actually selling much. Amazon admitted to be selling theirs at a loss. Facebook has tried their own - and quickly cancelled them. Google's is in every Android device - and yet pretty much nobody uses them. Even Apple's Siri is more annoyance than help. That something can be built doesn't mean it will sell or tha…

> Hoping for one of these with an open API that anyone could integrate whatever they want with, I am not holding my breath here.

OpenAI is already on it. The latest gen of GPT-3 and -4 are finetuned to respond to "do this thing" commands with JSON structured to:

- provide the name of a given function call

- provide arguments to that function call

it's "early stage", which in this case probably means "good enough to be useful within a month or two", given the rate at which these things have been developing.

Anecdotally, I've been playing with giving the models instructions like:

"When asked to perform a task that you need a tool to accomplish, you will call the tool according to its documentation by this format:

TOOL_NAME(*args)

Below you will find the documentation for your tools."

...and I've gotten it working pretty damn well (not even with the JSON-finetuned models, mind you). All you really need is python-style docstrings and a minimal parser and you're off to the races. I recommend anyone interested play with it a bit.

Re: Ask HN: Are people in tech inside an AI echo chamber?

#417

Post by Quarrelsome »

krageon wrote: post #181
Quarrelsome wrote: post #100

Earlier quoted context omitted.

> In your opinion, how does the "hallucination" issue differ from the same behaviour we see in humans? I feel like if you have any belief in philosophy then LLMs can only be interpreted as a parlour trick (on steroids). Perhaps we are fanciful in believing we are something greater than LLMs but there is the idea that we respond using rhetoric based on trying to find reason within in what we have learned and observed.…

> if you have any belief in philosophy [...] You're going to have to explain which part of philosophy you mean, because what came after this doesn't follow from that premise at all. It's like saying a Chinese Room is fundamentally different from a "real" solution even though nobody can tell the difference. That's not a "belief in philosophy", that's human exceptionalism and perhaps a belief in the soul.

The belief that your thoughts are constructed based on an understanding of principles such as logic, rationality, ethics. That your interactions are built from a solid understanding of these ideas. As opposed to every train of thought just being glued together from pertinent fragments you can recall from your knowledge in response to a prompt provided by the circumstances of reality.

> that's human exceptionalism and perhaps a belief in the soul.

I would also argue that LLMs are not proven to be equivalent to what's going on in our minds. Is it really "human exceptionalism" to state that LLMs are not yet and perhaps never will be what we are? I feel like from their construction it is somewhat evident that there are differences, since we don't raise humans the same way we raise LLMs. In terms of CPU years babies require significantly less time to train.

Re: Ask HN: Are people in tech inside an AI echo chamber?

#418

Post by Roark66 »

sensanaty wrote: post #138

Definitely. The tech is impressive but anyone I've spoken to thinks of it as Cleverbot 2.0, and among the more technically minded I've found that people mostly are indifferent. Hell, IRL most people I know don't think much of it, though on HN and elsewhere online I see a lot of people praising it as the next coming of Christ (this thread included) which puts it in a similar tier as crypto and other Web3 hypetrains as…

To be fair, commenting on your customer support anecdote, you can get very good quality answers from chatgpt on common knowledge items. You just have to craft the prompt correctly.

I don't even start talking to it without the first instruction being "answer all following questions in the shortest form possible"

This cuts out 90% of the useless output such models generate.

Re: Ask HN: Are people in tech inside an AI echo chamber?

#419

Post by bondarchuk »

samuellevy wrote: post #105
Obscurity4340 wrote: post #66

Earlier quoted context omitted.

Not to be difficult but wouldn't "confabulating" be a preferable description for this behaviour? Hallucinating doesn't quite feel right but I can't exactly articulate why confabulate is superior in this context

Nah, my issue with both terms is that they imply that when the answer is "correct" that's because the LLM "knows" the correct answer, and when it's wrong it's just a brain fart. It doesn't matter if the output is correct or not, the process for producing it is identical, and the model has the exact same amount of knowledge about what it's saying... which is to say "none". This isn't a case of "it's intelligent, but i…

>It doesn't matter if the output is correct or not, the process for producing it is identical

I don't see how this differs from a human earnestly holding a mistaken belief.

Re: Ask HN: Are people in tech inside an AI echo chamber?

#420

Post by FrustratedMonky »

There is a large contingent of the doomsayer on AI, where one of the doom scenarios, is that people don't realize they are being manipulated by AI.

So this case where you are talking to people outside the bubble, and they don't realize what is going on, IS THE PROBLEM.

Outside the bubble, people will be seeing images, video, text, voice, that is all fake and manipulated, and THEY ARE NOT PAYING ATTENTION, they don't know it is fake. Inside the bubble we are noticing it, outside the bubble, large chunks of population don't know it is happening.

Post reply on HN