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

Princeton ‘AI Snake Oil’ authors say GenAI hype has ‘spiraled out of control’

venturebeat.com

Post reply on HN
Print view View original

81–90 of 112 posts

Re: Princeton ‘AI Snake Oil’ authors say GenAI hype has ‘spiraled out of control’

#81

Post by Turskarama »

brendamn wrote: post #71
kromem wrote: post #44

Earlier quoted context omitted.

Really? You think people who write for a living might be biased towards confirming that the thing already replacing many people who write for a living is a fad that's going to blow over soon? My favorite part with crap predicting doom is the sheer unawareness of the rate of change of the rate of change. If you had asked me 18 months ago how long after GPT-3 until something existed with the capabilities of GPT-4, I'd…

> already replacing many people who write for a living Is this actually the case? Are there examples of professional writers/academics/etc who have lost their jobs to LLMs?

Yes, mostly low quality newspapers though. It remains to be seen if those companies regret their decision.

Re: Princeton ‘AI Snake Oil’ authors say GenAI hype has ‘spiraled out of control’

#83

Post by pepa65 »

pdntspa wrote: post #70
TuringNYC wrote: post #23

Earlier quoted context omitted.

Point taken, but how about these: - My mom cant figure out Blockchain either - My sister (doctor) and brother (3 degrees) cant figure out Blockchain either, because they are too afraid of scanning the wrong QR code and having their balance drained My mom, sister, and brother all use ChatGPT. My brother uses it actively for work.

You can dig deeper.... what exactly are the real-world applications of an immutable transaction ledger? I can't think of very many. Let alone a bunch of kids.

Immutable decentralized public ledgers could have many usages, but the biggest problem is: who is authorized to add new entries? There are different answers to this question depending on use case, but it is really the crux of the matter.

Re: Princeton ‘AI Snake Oil’ authors say GenAI hype has ‘spiraled out of control’

#84

Post by pepa65 »

pharmakom wrote: post #77
Turskarama wrote: post #72

Earlier quoted context omitted.

Midjourney and ChatGPT are both free to try, if someone gave me free bitcoin I might have made a wallet.

They had this in the early days. Bitcoin Faucet

Sigh... I "parked" my free bitcoin in the wrong place. Didn't know yet about just holding on to the private key and let it sit on the chain.

Re: Princeton ‘AI Snake Oil’ authors say GenAI hype has ‘spiraled out of control’

#85

Post by md_ »

Turskarama wrote: post #73
lIIllIIllIIllII wrote: post #10

Earlier quoted context omitted.

I'm routinely using LLMs in place of search engines. Would you rather type in a cocktail and instantly get a recipe? or do you want to open google, scroll past 50 ads, click through a bunch of blogspam that gives you the history of cocktails, the prohibition, the chemical formula for ethanol and their life story, followed by a recipe that was itself generated by a substandard AI model anyway edit: I'm prepared for th…

It depends on how important it is that I get a precisely correct answer. For a cocktail it's not a big deal if I get one that's slightly different, in other cases an incorrect answer might be catastrophic.

Amusingly, Kevin Roose and Casey Newton made a ChatGPT-authored cocktail recipe on "Hard Fork" a while back: https://www.nytimes.com/2023/06/30/podcasts/hard-fork-ai-poi....

I think the review was basically, "I’ve had worse cocktails. I’ll be honest."

So...meh?

Re: Princeton ‘AI Snake Oil’ authors say GenAI hype has ‘spiraled out of control’

#86

Post by theonlybutlet »

Turskarama wrote: post #73
lIIllIIllIIllII wrote: post #10

Earlier quoted context omitted.

I'm routinely using LLMs in place of search engines. Would you rather type in a cocktail and instantly get a recipe? or do you want to open google, scroll past 50 ads, click through a bunch of blogspam that gives you the history of cocktails, the prohibition, the chemical formula for ethanol and their life story, followed by a recipe that was itself generated by a substandard AI model anyway edit: I'm prepared for th…

It depends on how important it is that I get a precisely correct answer. For a cocktail it's not a big deal if I get one that's slightly different, in other cases an incorrect answer might be catastrophic.

Using bing chat, you always just check the source of its answer if it's crucial. Saves time searching.

Re: Princeton ‘AI Snake Oil’ authors say GenAI hype has ‘spiraled out of control’

#87

Post by theonlybutlet »

yeahwhatever10 wrote: post #58
lIIllIIllIIllII wrote: post #10

Earlier quoted context omitted.

I'm routinely using LLMs in place of search engines. Would you rather type in a cocktail and instantly get a recipe? or do you want to open google, scroll past 50 ads, click through a bunch of blogspam that gives you the history of cocktails, the prohibition, the chemical formula for ethanol and their life story, followed by a recipe that was itself generated by a substandard AI model anyway edit: I'm prepared for th…

I prefer search engines because I can look through multiple sources. LLMs just give an answer without context or what alternatives might be. They make it difficult to judge to validity of the response.

Not all questions need a variety of answers. If its good enough for your needs go for it. Saves some time searching.

Re: Princeton ‘AI Snake Oil’ authors say GenAI hype has ‘spiraled out of control’

#88

Post by md_ »

piloto_ciego wrote: post #21
version_five wrote: post #3

Not that I disagree, but as I've said before, it's easy to be a theoretical contrarian, as in claim something is BS but not stake anything on it or profit from it (other than book deals). We're in a housing bubble, at the peak of a multi-century economic mega-cycle, we're overdue for earthquakes, tsunamis, asteroids, the caldera under Yellowstone, flu pandemics, etc etc. It's super easy to say all that and demand peo…

Being a pessimist is academically safe and intellectually lazy. If you’re wrong in your predictions nobody cares because it’s better than if you’re right. On the other hand, if you’re an optimist and you are wrong you look like an idiot and the social penalty is higher. So people continue to predict doomsday and most of the time it doesn’t happen.

> So people continue to predict doomsday and most of the time it doesn’t happen.

Well, the same is true for any prediction of anything unlikely. People continue to predict world-changing innovations, and most of the time they are wrong. That's the reality of any low-probability/high-impact prediction, in the positive or negative direction.

As a specific case of that: most startups fail, so predicting a startup will fail is safe.

I'm pretty prone to exactly that prediction. But I like this quote from Erik Davis (https://www.google.ch/books/edition/High_Weirdness/Rcq2DwAAQ...): "In the court of the mind, skepticism makes a great grand vizier, but a lousy lord."

Re: Princeton ‘AI Snake Oil’ authors say GenAI hype has ‘spiraled out of control’

#89

Post by visarga »

ggm wrote: post #78
m00x wrote: post #69

Earlier quoted context omitted.

Crazy thing is that intelligence doesn't mean you get the right answer. Humans get the answer right fewer times than AI. It's absolutely not AGI, but it's not "not the road to AGI". Every step forward could be the road to AGI.

Could is doing a lot of heavy lifting here. People have been hypothesising about the nature of mind and intelligence for a long time. There is no compelling connectionist, moar-will-work, it-can-learn-to-think model, which infers the nature of intelligence and how that relates to GAN or GPT or any of what people are doing now. What people are doing now is amazing. I love looking at pictures of the pope in puffer jack…

The future will probably have amazing computers and models, but what we already have has the capacity to profoundly change the world.

I don't think you are focusing on what matters most. It's really the language itself that contains the intelligence, not the AI models. The models are just vessels for absorbing all that knowledge encoded in text. Just like human brains - we can have different neural wiring but learn the same things through education.

So the huge datasets these AI systems are trained on are key. That's how models like GPT-4 gain such language understanding. The architecture matters less than all that linguistic data it ingests. And language has been evolving for millennia, long before AI. It replicates through culture, speech, writing. Now with AI it has a whole new medium.

Fascinating question - how will AI affect language evolution going forward? As models produce more human-like text, which can further train better models, it's like a Lamarckian evolution. Acquired linguistic intelligence gets absorbed by AI then improved and propagated back into the corpus.

So while AI tech will keep advancing, it's language evolution that's most profound. By creating this new way for language to evolve, AI could really reshape cultural evolution. Since language enables intelligence, it'll have big impacts on where AI is headed next.

Re: Princeton ‘AI Snake Oil’ authors say GenAI hype has ‘spiraled out of control’

#90

Post by theonlybutlet »

gen220 wrote: post #22
Someone1234 wrote: post #8

The key way I try to distinguish "Snake Oil" is: Is anyone actually using this routinely , and are the problems it solves self-evident to new non-zealot users after a demo? Take blockchain for example; people spent years and millions of dollars looking for additional usages for the technology beside coins/coin-contracts; and aside from the forementioned usages it didn't really gain routine adoption. That was why the…

Are people durably achieving a competitive advantage through the GPT? I think we've seen some people striving for a competitive advantage, and willing to see one exist, but I don't know if we've seen any "AI company" turn a meaningful and durable profit yet (open to being wrong on this). To me (so far), it seems like GPT use cases that apply it as a table stakes feature of some greater application (rather than an eco…

I agree with your last paragraph.

In my opinion, Microsoft's proposed 365 Co-pilot pricetag of $30 per month per user will probably make a profit. High usage users being offset by low usage users when corporate 365 group memberships come into the mix. Many corporates will take a chance on it, willing to throw plenty of money at anything that is perceived as having a chance of improving hard to define competitive edge, creativity and productivity.

Post reply on HN