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

This time is different

shkspr.mobi

Post reply on HN
Print view View original

171–180 of 400 posts

Re: This time is different

#171

Post by bigstrat2003 »

madrox wrote: post #158

I got my first tech job in 2001. I've been doing this a while and ridden all the waves. There are two kinds of waves. The ones that don't require collective belief in them to succeed, and those that do. The latter are kinds like crypto and social media. The former is mobile...and AI. If no one else in the world had access to AI except me, I would appear superhuman to everyone in the world. People would see my level o…

> If no one else in the world had access to AI except me, I would appear superhuman to everyone in the world.

You really wouldn't. AI simply isn't that useful because it is so unreliable.

Re: This time is different

#172

Post by bigstrat2003 »

cladopa wrote: post #103

A very cynical article. Actually IT IS different. Actually if they manage to create a viable small nuclear reactors or Quantum computers the world will change like it changed with the Watt thermal engine. Why he is not talking about the Internet, trains, electricity, nuclear bombs, rockets,aviation or engines? Because they worked, like AI works today. All of them were bubbles at the time and they changed the world fo…

> Why he is not talking about the Internet, trains, electricity, nuclear bombs, rockets,aviation or engines? Because they worked, like AI works today.

Except for the minor bit that AI doesn't work today, and it is not yet clear if it ever will.

Re: This time is different

#173

Post by johnfn »

gambiting wrote: post #152
johnfn wrote: post #124

Earlier quoted context omitted.

I'm curious about the statement that hallucinations are "fundamentally unsolvable". I don't think an AI agent has left a hallucination in my code - by which I mean a reference to something which doesn't exist at all - in many months. I have had great luck driving hallucinations to effectively 0% by using a language with static typechecking, telling LLMs to iterate on type errors until there are none left, and of cour…

>> I don't think an AI agent has left a hallucination in my code I literally just went on Gemini, latest and best model and asked it "hey can you give me the best prices for 12TB hard drives available with the British retailer CeX?" and it went "sure, I just checked their live stock and here they are:". Every single one was made up. I pointed it out, it said sorry, I just checked again, here they are, definitely 100%…

My comment was about code, not fact checking - that’s why I said they were a solved problem provided you use static typechecking and tests.

Re: This time is different

#174

Post by johnfn »

bigstrat2003 wrote: post #168
parliament32 wrote: post #123

When I look at LLMs as an interface, I'm reminded of back when speech-to-text first became mainstream. So many promises about how this is the interface for how we'll talk to computers forevermore. Here we are a few decades later, and we don't see business units using Word's built-in dictation feature to write documents, right? Funny how that tech seems to have barely improved in all that time. And, despite dictation…

Yeah this is exactly my view. We've had several years of work on the tech, and LLMs are just as prone to randomly spitting out garbage as they were the first day. They are not a tool which is fit for any serious work, because you need to be able to rely on your tools. A tool which is sometimes good and sometimes bad is worse than having no tool at all.

Do you really think that Opus 4.6 hallucinates to exactly the same degree as GPT-3.5? I am mystified how you can hold this perspective.

Re: This time is different

#175

Post by radlad »

cogman10 wrote: post #100
tomlue wrote: post #90

Earlier quoted context omitted.

"This time is different" has been correct for every major technological shift in history. Electricity was different. Antibiotics were different. Semiconductors were different. Gen AI reached 39% adoption in two years (internet took 5, PCs took 12). Enterprise spend went from $1.7B to $37B since 2023. Hyperscalers are spending $650B this year on AI infra and are supply-constrained, not demand-constrained. There is no…

The problem is in the middle of such a change it's hard to recognize if this is a real change or if this is another Wankel motor. Plenty a visual programming language has tried to toot their own horns as being the next transformative change in everything, and they are mostly just obscure DSLs at this point. The other issue is nobody knows what the future will actually look like and they'll often be wrong with their p…

> Plenty a visual programming language has tried to toot their own horns as being the next transformative change in everything, and they are mostly just obscure DSLs at this point.

But how many of your non-nerdy friends were talking about them, let alone using them daily?

Re: This time is different

#176

Post by selridge »

bigstrat2003 wrote: post #166
tomlue wrote: post #90

Earlier quoted context omitted.

"This time is different" has been correct for every major technological shift in history. Electricity was different. Antibiotics were different. Semiconductors were different. Gen AI reached 39% adoption in two years (internet took 5, PCs took 12). Enterprise spend went from $1.7B to $37B since 2023. Hyperscalers are spending $650B this year on AI infra and are supply-constrained, not demand-constrained. There is no…

> The real debate isn't whether AI is transformative. No, the debate is very much whether AI is transformative. You don't get to smuggle your viewpoint as an assumption as if there was consensus on this point. There isn't consensus at all.

No one is smuggling this in. The debate is over. It's transformative. We're in the midst of transformation.

Re: This time is different

#177

Post by doyougnu »

naravara wrote: post #88
thomassmith65 wrote: post #69

The hype around AI is admittedly annoying - especially from the Wall St crowd who don't know how to pronounce 'Nvidia' correctly, and who haven't managed to internalize the fact that the chatbots they use hallucinate. It really is 'different' , though, in the same way the Internet was. It took about 20 years (ie: since The World ISP) for the Internet to work its way into every facet of life. And the dot com bubble po…

I think a good analogy will be the way word processors changed printing. Suddenly anyone with access to a computer had the ability to do professional level editing and layout. Most of them didn’t have the taste or skills to use the tools to the fullest, but it still opened up a ton of possibilities that weren’t available before because it was never practical to hire an actual professional to do a poster for a dinky c…

I think this is exactly right. I've been thinking of "this time" as similar to the advent of digital spreadsheets. Spreadsheets existed for thousands of years but spreadsheet programs transformed spreadsheet work that took hours or weeks into seconds. You still had to know what you were doing, and if you knew what you were doing you were easily 10x faster than those that didn't.

I think we are in a similar situation with code generation now, then only difference in my mind is that LLMs come with a massive platform risk. Who's to say that one day anthropic decides my company is too much of a competitor to use their tool (like they've already done with openai) or what if they decide that instead of pulling their product from my use they just make it generate worse code, or even insert malicious payloads. A dependence on these tools is wildly more risky than dependency on a word processor or a spreadsheet program. It reminds me of the arguments around net neutrality and I cannot fathom how people building on top of, and with, these tools do not see the mountain of risks around them.

Re: This time is different

#178

Post by enraged_camel »

esperent wrote: post #167

> 3D TV, AMP, Augmented Reality, Beanie Babies, Blockchain, Cartoon Avatars, Curved TVs, Frogans, Hoverboards, iBeacons, Jetpacks, Metaverse, NFTs, Physical Web, Quantum Computing, Quibi, Small and Safe Nuclear Reactors, Smart Glasses, Stadia, WiMAX. Agreed, these things all failed to live up to the hype. But these didn't: Electricity, cheap computing, calculators, photography, the internet, the steam engine, the pri…

The article is trash. The only reason it got voted to the front page is because the author is salty about AI.

Re: This time is different

#179

Post by dwoldrich »

esperent wrote: post #167

> 3D TV, AMP, Augmented Reality, Beanie Babies, Blockchain, Cartoon Avatars, Curved TVs, Frogans, Hoverboards, iBeacons, Jetpacks, Metaverse, NFTs, Physical Web, Quantum Computing, Quibi, Small and Safe Nuclear Reactors, Smart Glasses, Stadia, WiMAX. Agreed, these things all failed to live up to the hype. But these didn't: Electricity, cheap computing, calculators, photography, the internet, the steam engine, the pri…

Exactly my thoughts. Selective whinging indeed.

Also meta-platitude whinging like

> The ideology of "winner takes all" is unsustainable and not supported by reality.

Sometimes the winner deserves to win, AND that's a good thing even at scale. It kindof depends.

Re: This time is different

#180

Post by parliament32 »

johnfn wrote: post #174
bigstrat2003 wrote: post #168

Earlier quoted context omitted.

Yeah this is exactly my view. We've had several years of work on the tech, and LLMs are just as prone to randomly spitting out garbage as they were the first day. They are not a tool which is fit for any serious work, because you need to be able to rely on your tools. A tool which is sometimes good and sometimes bad is worse than having no tool at all.

Do you really think that Opus 4.6 hallucinates to exactly the same degree as GPT-3.5? I am mystified how you can hold this perspective.

If you're not seeing the hallucinations, I'd assert you're either not using it enough, or (more likely) you don't have enough knowledge in the subject matter to notice when it's hallucinating.
Post reply on HN