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

ChatGPT turned generative AI into an “anything tool”

arstechnica.com

Post reply on HN
Print view View original

71–80 of 245 posts

Re: ChatGPT turned generative AI into an “anything tool”

#71

Post by consumer451 »

Alupis wrote: post #70
consumer451 wrote: post #55

Earlier quoted context omitted.

> I'd like to have LawGPT I tried ChatGPT 4 for a specific TX county's laws. This was after I had spent hours doing traditional search. ChatGPT got everything correct, and even gave me some more direction.

> ChatGPT got everything correct My anecdote with technical information is exactly the opposite, unfortunately. I routinely use it to help guide my research - generally about topics I know something about but want to know more. It often will provide some response that is entirely wrong but looks very good. When called out, it apologizes, then proceeds to produce more information that's also not entirely correct. Ther…

I have had similar experiences in other areas, that's why I didn't try an LLM before spending hours researching. But I must say that I was shocked at the accuracy in that one particular area... Harris County estate law.

What is the difference here? No hallucinations, everything correct… is it just random chance?

Re: ChatGPT turned generative AI into an “anything tool”

#73

Post by verve_rat »

spott wrote: post #43
assistantist wrote: post #35

Earlier quoted context omitted.

If it was just a matter of flops then AI would already be driving cars and making door to door food deliveries. The problem is neither the energy expenditure nor the compute capacity. No one knows what the proper architecture should be for general purpose intelligence assistants so everyone has to fine-tune the existing models on their specialized use cases and even then it's a hit or miss depending on how much train…

I think you are missing my point. The OP said Siri et al. got worse because their creators optimized them to reduce compute costs. I’m saying that if that were true, then they should be at least as good as they were because you can get tons more compute for cheaper these days. So it isn’t compute that makes the various assistants “worse” than when they were introduced. I made no statements about their quality in gene…

That assumes that holding perceived performance steady while letting compute consumption get cheaper saves enough money, quick enough.

If voice assistants were losing $10 billion a year (made up number) in 2013 and bringing in zero revenue, then it is easy to see why a decade later they would have less compute allocation now than they did back then, even in absolute terms.

If the revenue is near zero, then the spend on compute will be near zero too, after everyone figured out that it wasn't going to win some AI battle for all the consumers. If they are not willing to set money on fire for no measurable benefits anymore, then compute will tend to zero over time.

Re: ChatGPT turned generative AI into an “anything tool”

#74

Post by mensetmanusman »

dandellion wrote: post #34
JimtheCoder wrote: post #16

Earlier quoted context omitted.

"I look forward to hooking all my at-home cameras up to a voice recognition, image recognition, and LLM detection suite and just speak aloud and have the AI-suite "do stuff". It sounds really, really fun to me." Alternatively, this sounds like the starting plot to a crossover sci-fi/horror movie...

The door refused to open. It said, “Five cents, please.” He searched his pockets. No more coins; nothing. “I’ll pay you tomorrow,” he told the door. Again he tried the knob. Again it remained locked tight. “What I pay you,” he informed it, “is in the nature of a gratuity; I don’t have to pay you.” “I think otherwise,” the door said. “Look in the purchase contract you signed when you bought this conapt.” In his desk d…

The 5 cent reference has a history in Boston, it’s why the mbta card is called the Charlie card:

https://youtu.be/MbtkL5_f6-4?si=IQZrPJhDWhNVyhx_

Re: ChatGPT turned generative AI into an “anything tool”

#75

Post by version_five »

keyle wrote: post #67
Someone1234 wrote: post #2

Although let's see if it is sustainable. I pay for ChatGPT Plus but most people aren't, and from what I've read they're losing money like crazy. We may wind up seeing this as the technical high point before the various competitors cut back on expensive hardware/energy usage for the free tier, and the quality of the responses takes a dive. This might sound a bit conspiratorial so apologies for that but back when the c…

I'm actually worried of the inevitable enshitification of Open AI. Right now, GPT 4 is great and I use it every day. I rely less and less on Google and save a ton of time. Google has gone downhill hard for a few years, thanks to all these silly blog engines like medium, substack and the like. I'm really worried and would like to 'freeze' GPT 4 as is, at least with the data up to Sep 2021, and for them to not change o…

This is a good point and I'm surprised it hasn't had more attention. We've seen history repeat itself enough times that it's inevitable. Probably the best you can do is download a few of the full precision 70B or bigger models (LLaMA etc) to at least have some state of the art weights around you could use yourself if you wanted if/when openAI goes to shit.

The nice thing about the language models is as least they're relatively compact, compared to google search which you'd never be able to host locally.

Though thinking about it, in a way LLMs are local copies of the "good" internet. The big ones mostly seem to know what information there is to know and have none of the crap and link/content farms that is now most of the internet.

Re: ChatGPT turned generative AI into an “anything tool”

#76

Post by satvikpendem »

keyle wrote: post #67
Someone1234 wrote: post #2

Although let's see if it is sustainable. I pay for ChatGPT Plus but most people aren't, and from what I've read they're losing money like crazy. We may wind up seeing this as the technical high point before the various competitors cut back on expensive hardware/energy usage for the free tier, and the quality of the responses takes a dive. This might sound a bit conspiratorial so apologies for that but back when the c…

I'm actually worried of the inevitable enshitification of Open AI. Right now, GPT 4 is great and I use it every day. I rely less and less on Google and save a ton of time. Google has gone downhill hard for a few years, thanks to all these silly blog engines like medium, substack and the like. I'm really worried and would like to 'freeze' GPT 4 as is, at least with the data up to Sep 2021, and for them to not change o…

That's why I think open source LLMs are the ones that will actually persist throughout the commercial enshittification and reign supreme at the end.

Re: ChatGPT turned generative AI into an “anything tool”

#77

Post by keyle »

satvikpendem wrote: post #76
keyle wrote: post #67

Earlier quoted context omitted.

I'm actually worried of the inevitable enshitification of Open AI. Right now, GPT 4 is great and I use it every day. I rely less and less on Google and save a ton of time. Google has gone downhill hard for a few years, thanks to all these silly blog engines like medium, substack and the like. I'm really worried and would like to 'freeze' GPT 4 as is, at least with the data up to Sep 2021, and for them to not change o…

That's why I think open source LLMs are the ones that will actually persist throughout the commercial enshittification and reign supreme at the end.

I'm with you but I'm yet to find a solution that isn't both rocket science and requiring rocket hardware.

Re: ChatGPT turned generative AI into an “anything tool”

#78

Post by Turskarama »

NoLinkToMe wrote: post #64
Someone1234 wrote: post #2

Although let's see if it is sustainable. I pay for ChatGPT Plus but most people aren't, and from what I've read they're losing money like crazy. We may wind up seeing this as the technical high point before the various competitors cut back on expensive hardware/energy usage for the free tier, and the quality of the responses takes a dive. This might sound a bit conspiratorial so apologies for that but back when the c…

> the responses to the various voice assistance was almost eerie in that it could infer what you wanted/needed. Then as things cooled off, they got gradually dumber/worse every year since then. They're legitimately bad today (Siri in particular It's a compelling argument but I find that difficult to believe given the timespan. Siri launched in 2011. The FLOP/s per dollar rates for GPUs involved in ML doubled every 2…

The hardware being cheaper only matters if they truly want to improve the product, which it's not clear that they do. Lowering the number of FLOPs needed will always reduce the cost, as the hardware gets cheaper the cost is just reduced faster.

The question is, would improving Siri sell more IPhones? If not, then why bother?

Re: ChatGPT turned generative AI into an “anything tool”

#79

Post by satvikpendem »

keyle wrote: post #77
satvikpendem wrote: post #76

Earlier quoted context omitted.

That's why I think open source LLMs are the ones that will actually persist throughout the commercial enshittification and reign supreme at the end.

I'm with you but I'm yet to find a solution that isn't both rocket science and requiring rocket hardware.

What do you mean?

Re: ChatGPT turned generative AI into an “anything tool”

#80

Post by Turskarama »

consumer451 wrote: post #55
lakomen wrote: post #29

I'd like to have LawGPT, where it reads all laws of a certain country, and I can ask it things like, "what are my options when the neighbors tree crosses over into my property?"

> I'd like to have LawGPT I tried ChatGPT 4 for a specific TX county's laws. This was after I had spent hours doing traditional search. ChatGPT got everything correct, and even gave me some more direction.

The issue is that you cannot _guarantee_ it will get everything correct, the fact that it managed it in your case is largely a matter of luck.

In law, being correct 99% of the time is not really good enough.

Post reply on HN