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

I genuinely don't understand why some people are still bullish about LLMs

twitter.com

Post reply on HN
Print view View original

341–350 of 1001 posts

Re: I genuinely don't understand why some people are still bullish about LLMs

#341

Post by kerkeslager »

nonchalantsui wrote: post #33
MostlyStable wrote: post #19

Earlier quoted context omitted.

I don't know what you are doing or why it's failed. Maybe my primary use cases really are in the top whatever percentile for AI usefulness, but it doesn't feel like it. All I know is that frontier models have already been good enough for more than a year to increase my productivity by a fair bit.

Your use case is in fact in the top whatever percentile for AI usefulness. Short simple scripting that won't have to be relied on due to never being widely deployed. No large codebase it has to comb through, no need for thorough maintenance and update management, no need for efficient (and potentially rare) solutions. The only use case that would beat yours is the type of office worker that cannot write professional…

And... I know people who now use AI to write their professional-sounding emails, and they often don't sound as professional as they think they do. It can be easy to just skim what an AI generates and think it's okay to send if you aren't careful, but the people you send those emails to actually have to read what was written and attempt to understand it, and doing that makes you notice things that a brief skim doesn't catch.

It's actually extremely irritating that I'm only half talking to the person when I email with these people.

Re: I genuinely don't understand why some people are still bullish about LLMs

#342

Post by geuis »

maple3142 wrote: post #171

I think many people are just not really good at dealing with "imperfect" tools. Different tools can have different success probability, let's call that probability p here. People typically use tool that have p=100%, or at least very close to it. But LLM is a tool that is far from that, so making use of it takes different approach. Imagine there is an probabilistic oracle that can answer any question with a yes/no wit…

A scientific instrument that is unreliably accurate is useless. Imagine a kitchen scale that always gave +/- 50% every 3rd time you use it. Or maybe 5th time. Or 2nd.

So we're trying to use tools like this currently to help solve deeper problems and they aren't up to the task. This is still the point we need to start over and get better tools. Sharpening a bronze knife will never be as sharp or have the continuity as a steel knife. Same basic elements, very different material.

Re: I genuinely don't understand why some people are still bullish about LLMs

#343

Post by protocolture »

ffsm8 wrote: post #117
gilbetron wrote: post #109

I get so confused on this. I play around, test, and mess with LLMs all the time and they are miraculous. Just amazing, doing things we dreamed about for decades. I mean, I can ask for obscure things with subtle nuance where I misspell words and mess up my question and it figures it out. It talks to me like a person. It generates really cool images. It helps me write code. And just tons of other stuff that astounds me…

It's definitely a tech that's here to stay, unlike block chain/nfts But I mirror the confusion why people are still bullish on it. The current valuation for it is because the market thinks that it's able to write code like a senior engineer and have AGI, because that's how they're marketed by the LLM providers. I'm not even certain if they'll be ubiquitous after the venture capital investments are gone and the servic…

NFTs are the perfect example.

NFT is still a great tool if you want a bunch of unique tokens as part of a blockchain app. ERC-721 was proven a capable protocol in a variety of projects. What it isn't, and never will be, is an amazing investment opportunity, or a method to collect cool rare apes and go to yacht parties.

LLMs will settle in and have their place too, just not in the forefront of every investors mind.

Re: I genuinely don't understand why some people are still bullish about LLMs

#344

Post by mplanchard »

grg0 wrote: post #313
mplanchard wrote: post #308

Earlier quoted context omitted.

I’m always amazed in these discussions how many people apparently have jobs doing a bunch of stuff that either doesn’t need to be correct or is simple enough that it doesn’t require any significant amount of external context.

In other words, people preaching LLMs are noobs with no real stake in what they are doing. But you can't really call people noobs these days.

Ha, a little harsher than I intended it to come off, but it really does make me wonder.

Re: I genuinely don't understand why some people are still bullish about LLMs

#345

Post by IncreasePosts »

crazygringo wrote: post #264
sc68cal wrote: post #208

Earlier quoted context omitted.

> Does it fabricate references? Absolutely, maybe about a third of the time And you don't have concerns about that? What kind of damage is that doing to our society, long term, if we have a system that _everyone_ uses and it's just accepted that a third of the time it is just making shit up?

No, I don't. Because I know it does and it's incredibly easy to type something into Google Scholar and see if a reference exists. Like, I can ask a friend and they'll mistakenly make up a reference. "Yeah, didn't so-and-so write a paper on that? Oh they didn't? Oh never mind, I must have been thinking of something else." Does that mean I should never ask my friend about anything ever again? Nobody should be using the…

No, that doesn't mean you should never ask your friend things again if they make that mistake. But, if 30% of all their references are made up then you might start to question everything your friend says. And looking up references to every claim you're reading is not a productive use of time.

Re: I genuinely don't understand why some people are still bullish about LLMs

#346

Post by oaeirjtlaierjt »

gilbetron wrote: post #109

I get so confused on this. I play around, test, and mess with LLMs all the time and they are miraculous. Just amazing, doing things we dreamed about for decades. I mean, I can ask for obscure things with subtle nuance where I misspell words and mess up my question and it figures it out. It talks to me like a person. It generates really cool images. It helps me write code. And just tons of other stuff that astounds me…

I often ask "So you say LLMs are worthless because you can't blindly trust the first thing they say? Do you blindly trust the first google search result? Do you trust every single thing your family members tell you?" It reminds me of my high school teachers saying Wikipedia can't be trusted.

Re: I genuinely don't understand why some people are still bullish about LLMs

#347

Post by LeonB »

brookst wrote: post #227
ffsm8 wrote: post #117

Earlier quoted context omitted.

It's definitely a tech that's here to stay, unlike block chain/nfts But I mirror the confusion why people are still bullish on it. The current valuation for it is because the market thinks that it's able to write code like a senior engineer and have AGI, because that's how they're marketed by the LLM providers. I'm not even certain if they'll be ubiquitous after the venture capital investments are gone and the servic…

Remember the quote that IBM thought there would be a total market for maybe 10 or 15 computer computers in the entire world? They were giant, and expensive, and very limited in application.

Turned out there was only room for about 4 oligopolist cloud providers.

Re: I genuinely don't understand why some people are still bullish about LLMs

#348

Post by lolinder »

lostmsu wrote: post #334
sarchertech wrote: post #242

Earlier quoted context omitted.

That’s not how language works.

That's precisely how language works in most cases, and there does not seem to be any reason to think it is different in this one. Although if you look at the trajectory of goalposts since https://web.archive.org/web/20140327014303/https://en.wikipe... , I could concede the point, that people discarded the obvious definition in the light of recent events.

No, it's really not. Joining words into a compound word enables the new compound to take on new meaning and evolve on its own, and if it becomes widely used as a compound it always does so. The term you're looking for if you care to google it is an "open compound noun".

A dog in the sun may be hot, but that doesn't make it a hot dog.

You can use a towel to dry your hair, but that doesn't make the towel a hair dryer.

Putting coffee on a dining room table doesn't turn it into a coffee table.

Spreading Elmer's glue on your teeth doesn't make it tooth paste.

The White House is, in fact, a white house, but my neighbor's white house is not The White House.

I could go on, but I think the above is a sufficient selection to show that language does not, in fact, work that way. You can't decompose a compound noun into its component morphemes and expect to be able to derive the compound's meaning from them.

Re: I genuinely don't understand why some people are still bullish about LLMs

#349

Post by Nihilartikel »

I don't know.. I've maintained skepticism, but recently AI has enabled solutions for client problems that would have been intractable with conventional coding. A team was migrating a years old excel based workflow where no less than 3 spreadsheets contained thousands of call notes, often with multiple notes stuffed into the same column separated inconstantly by a shorthand date and initials of who was in the call. Sometimes with text arrows or other meta descriptions like (all calls after 3/5 were handled by Tim). They want to move all of this into structured jira tickets and child tickets.

Joining the mess of freeform, redundant, and sometimes self contradicting data into JSON lines, and feeding it into AI with a big explicit prompt containing example conversions and corrections for possible pitfalls has resulted in almost magically good output. I added a 'notes' field to the output and instructed the model to call out anything unusual and it caught lots of date typos by context, ambiguously attributed notes, and more.

It would have been a man month or so of soul drowningly tedious and error prone intern level work, but now it was 40 minutes and $15 of Gemini usage.

So, even if it's not a galaxy brained super intelligence yet, it is a massive change to be able to automate what was once exclusively 'people' work.

Re: I genuinely don't understand why some people are still bullish about LLMs

#350

Post by chamomeal »

gilbetron wrote: post #109

I get so confused on this. I play around, test, and mess with LLMs all the time and they are miraculous. Just amazing, doing things we dreamed about for decades. I mean, I can ask for obscure things with subtle nuance where I misspell words and mess up my question and it figures it out. It talks to me like a person. It generates really cool images. It helps me write code. And just tons of other stuff that astounds me…

Couldn’t agree more!

When I saw GPT-3 in action in 2023, I couldn’t believe my eyes. I thought I was being tricked somehow. I’d seen ads for “AI-powered” services and it was always the same unimpressive stuff. Then I saw GPT-3 and within minutes I knew it was completely different. It was the first thing I’d ever seen that felt like AI.

That was only a few years ago. Now I can run something on my 8GB MacBook Air that blows GPT-3 out of the water. It’s just baffling to me when people say LLM’s are useless or unimpressive. I use them constantly and I can still hardly believe they exist!!

Post reply on HN