LLM Inevitabilism • Hacker News Board

Re: LLM Inevitabilism

#861

Post by dingnuts » Tue, Jul 15, 2025, 8:20 PM UTC

Earw0rm wrote: post #857

strange_quark wrote: post #828

Earlier quoted context omitted.

> There are many technologies that have seemed inevitable and seen retreats under the lack of commensurate business return (the supersonic jetliner) I think this is a great analogy, not just to the current state of AI, but maybe even computers and the internet in general. Supersonic transports must've seemed amazing, inevitable, and maybe even obvious to anyone alive at the time of their debut. But hiding under that…

From a system optimisation perspective, SSTs solved the wrong problem. Want to save people time flying? Solve the grotesque inefficiency pit that is airport transit and check-in. Like, I'm sorry, STILL no high speed, direct to terminal rail at JFK, LAX and a dozen other major international airports? And that's before we get to the absolute joke of "border security" and luggage check-in. Sure, supersonic afterburning…

SST came and went in an era when none of that security theater existed to begin with

Re: LLM Inevitabilism

#862

Post by JohnMakin » Tue, Jul 15, 2025, 8:20 PM UTC

brokencode wrote: post #706

lsy wrote: post #63

I think two things can be true simultaneously: 1. LLMs are a new technology and it's hard to put the genie back in the bottle with that. It's difficult to imagine a future where they don't continue to exist in some form, with all the timesaving benefits and social issues that come with them. 2. Almost three years in, companies investing in LLMs have not yet discovered a business model that justifies the massive expen…

> “most people agree that the output is trite and unpleasant to consume” That is a such a wild claim. People like the output of LLMs so much that ChatGPT is the fastest growing app ever. It and other AI apps like Perplexity are now beginning to challenge Google’s search dominance. Sure, probably not a lot of people would go out and buy a novel or collection of poetry written by ChatGPT. But that doesn’t mean the outp…

> That is a such a wild claim. People like the output of LLMs so much that ChatGPT is the fastest growing app ever.

And this kind of meaningless factoid was immediately usurped by the Threads app release, which IMO is kind of a pointless app. Maybe let's find a more meaningful metric before saying someone else's claim is wild.

Re: LLM Inevitabilism

#863

Post by LeifCarrotson » Tue, Jul 15, 2025, 8:23 PM UTC

underdeserver wrote: post #740

brokencode wrote: post #706

Earlier quoted context omitted.

> “most people agree that the output is trite and unpleasant to consume” That is a such a wild claim. People like the output of LLMs so much that ChatGPT is the fastest growing app ever. It and other AI apps like Perplexity are now beginning to challenge Google’s search dominance. Sure, probably not a lot of people would go out and buy a novel or collection of poetry written by ChatGPT. But that doesn’t mean the outp…

> That is a such a wild claim. People like the output of LLMs so much that ChatGPT is the fastest growing app ever. The people using ChatGPT like its output enough when they're the ones reading it. The people reading ChatGPT output that other people asked for generally don't like it. Especially if it's not disclosed up front.

Especially if it's not disclosed up front, and especially when it supplants higher-value content. I've been shocked how little time it's taken for AI slop SEO optimized blogs to overtake the articles written by genuine human experts, especially in niche product reviews and technical discussions.

However, whether or not people like it is almost irrelevant. The thing that matters is not whether economics likes it.

At least so far, it looks like economics absolutely loves LLMs: Why hire expensive human customer support when you can just offload 90% of the work to a computer? Why pay expensive journalists when you can just have the AI summarize it? Why hire expensive technical writers to document your code when you can just give it to the AI and check the regulatory box with docs that are good enough?

Re: LLM Inevitabilism

#864

Post by bluefirebrand » Tue, Jul 15, 2025, 8:30 PM UTC

kenjackson wrote: post #845

strange_quark wrote: post #828

Earlier quoted context omitted.

> There are many technologies that have seemed inevitable and seen retreats under the lack of commensurate business return (the supersonic jetliner) I think this is a great analogy, not just to the current state of AI, but maybe even computers and the internet in general. Supersonic transports must've seemed amazing, inevitable, and maybe even obvious to anyone alive at the time of their debut. But hiding under that…

I think this is both right and wrong. There was a good book that came out probably 15 years ago about how technology never stops in aggregate, but individual technologies tend to grow quickly and then stall. Airplane jets were one example in the book. The reason why I partially note this as wrong is that even in the 70s people recognized that supersonic travel had real concrete issues with no solution in sight. I don…

> even in the 70s people recognized that supersonic travel had real concrete issues with no solution in sight. I don't think LLMs share that characteristic today

I think they pretty strongly do

The solution seems to be "just lower your standards for acceptable margin of error to whatever the LLM is capable of producing" which should be concerning and absolutely unacceptable to anyone calling themselves an Engineer

Re: LLM Inevitabilism

#865

Post by matthewdgreen » Tue, Jul 15, 2025, 8:33 PM UTC

doctorpangloss wrote: post #808

hattmall wrote: post #718

Earlier quoted context omitted.

I'm not really countering that ChatGPT is popular, it certainly is, but it's also sort of like "fastest growing tire brand" that came along with the adoption of vehicles. The amount of smartphone users is also growing at the fastest rate ever so whatever the new most popular app is has a good chance of being the fastest growing app ever.

No… dude… it’s a new household name. We haven’t had those in software for a long time, maybe since TikTok and Fortnite.

Lots of things had household recognition. Do you fondly remember the Snuggie? The question is whether it'll be durable. The lack of network effects is one reason to be skeptical.

Re: LLM Inevitabilism

#866

Post by anthonypasq » Tue, Jul 15, 2025, 8:52 PM UTC

ramoz wrote: post #759

magic_hamster wrote: post #754

Earlier quoted context omitted.

There are pretty hidden assumption in this comment. First of all, not every business in the AI space is _training_ models, and the difference between training and inference is massive - i.e. most businesses can easily afford inference, perhaps depending on model, but they definitely can. Another several unfounded claims were made here, but I just wanted to say LLMs with MCP are definitely good enough for almost every…

LLMs with tools* MCP isn’t inherently special. A Claude Code with Bash() tool can do nearly anything a MCP server will give you - much more efficiently. Computer Use agents are here and are only going to get better. The conversation shouldn’t be about LLMs any longer. Providers will be providing agents.

correct and companies will be exposing their data via mcp instead of standard rest apis.

Re: LLM Inevitabilism

#867

Post by artninja1988 » Tue, Jul 15, 2025, 8:58 PM UTC

smeeger wrote: post #801

sircastor wrote: post #172

The hardest part about inevitablism here is that the people who are making the argument this is inevitable are the same people who are the people who are shoveling hundreds of millions of dollars into it. Into the development, the use, the advertisement. The foxes are building doors into the hen houses and saying there’s nothing to be done, foxes are going to get in so we might as well make it something that works fo…

agree. we should be shooting foxes rather than enabling them.

Somebody ban this guy

Re: LLM Inevitabilism

#868

Post by da_chicken » Tue, Jul 15, 2025, 9:00 PM UTC

kenjackson wrote: post #845

strange_quark wrote: post #828

Earlier quoted context omitted.

> There are many technologies that have seemed inevitable and seen retreats under the lack of commensurate business return (the supersonic jetliner) I think this is a great analogy, not just to the current state of AI, but maybe even computers and the internet in general. Supersonic transports must've seemed amazing, inevitable, and maybe even obvious to anyone alive at the time of their debut. But hiding under that…

I think this is both right and wrong. There was a good book that came out probably 15 years ago about how technology never stops in aggregate, but individual technologies tend to grow quickly and then stall. Airplane jets were one example in the book. The reason why I partially note this as wrong is that even in the 70s people recognized that supersonic travel had real concrete issues with no solution in sight. I don…

> The reason why I partially note this as wrong is that even in the 70s people recognized that supersonic travel had real concrete issues with no solution in sight. I don't think LLMs share that characteristic today.

The fundamental problem has already been mentioned: Nobody can figure out how to SELL it. Because few people are buying it.

It's useful for aggregation and summarization of large amounts of text, but it's not trustworthy. A good summary decreases noise and amplifies signal. LLMs don't do that. Without the capability to validate the output, it's not really generating output of lasting value. It's just a slightly better search engine.

It feels like, fundamentally, the primary invention here is teaching computers that it's okay to be wrong as long as you're convincing. That's very useful for propaganda or less savory aspects of business, but it's less useful for actual communication.

Re: LLM Inevitabilism

#869

Post by mxschumacher » Tue, Jul 15, 2025, 9:03 PM UTC

Earw0rm wrote: post #857

strange_quark wrote: post #828

Earlier quoted context omitted.

> There are many technologies that have seemed inevitable and seen retreats under the lack of commensurate business return (the supersonic jetliner) I think this is a great analogy, not just to the current state of AI, but maybe even computers and the internet in general. Supersonic transports must've seemed amazing, inevitable, and maybe even obvious to anyone alive at the time of their debut. But hiding under that…

From a system optimisation perspective, SSTs solved the wrong problem. Want to save people time flying? Solve the grotesque inefficiency pit that is airport transit and check-in. Like, I'm sorry, STILL no high speed, direct to terminal rail at JFK, LAX and a dozen other major international airports? And that's before we get to the absolute joke of "border security" and luggage check-in. Sure, supersonic afterburning…

that's not a technology problem, many airports are super efficient, e.g. Singapore. Public transport in the US is held back by other forces.

Re: LLM Inevitabilism

#870

Post by Earw0rm » Tue, Jul 15, 2025, 9:03 PM UTC

dingnuts wrote: post #861

Earw0rm wrote: post #857

Earlier quoted context omitted.

From a system optimisation perspective, SSTs solved the wrong problem. Want to save people time flying? Solve the grotesque inefficiency pit that is airport transit and check-in. Like, I'm sorry, STILL no high speed, direct to terminal rail at JFK, LAX and a dozen other major international airports? And that's before we get to the absolute joke of "border security" and luggage check-in. Sure, supersonic afterburning…

SST came and went in an era when none of that security theater existed to begin with

It did for international, maybe not at the dawn of SSTs but after a string of hijackings in the 70s/80s they brought it in. Not for US internal flights, it's true.