[caditinpiscinam]

We've all heard the phrase "the sum of all human knowledge".

I've been feeling more and more that generative AI represents the average of all human knowledge. Which has its place. But a future in which all thought and creativity is averaged away is a bleak one. It's the heat death of thought.

[dang]

Thought and creativity won't be averaged away because human beings have a drive for these things. This just raises the bar for it. And why not? We get complacent when not pushed.

Dostoevsky said that if all human knowledge could ever be reduced to 2 + 2 = 4, man would stick out his tongue and insist that 2 + 2 = 5. That was a 19th century formulation—he was a contemporary of Boole. I wonder what the equivalent would be for the LLM era.

[frm88]

Thought and creativity won't be averaged away because human beings have a drive for these things.

That may or may not be true, but the expression of thought and creativity matters to transfer meaning. If you average that out, it loses momentum. Example: https://news.ycombinator.com/item?id=47346935. Compare the posters first and second, LLM assisted, paragraph. The second one is just bleak. If I had to read several pages like that, my eyes would glaze over. It cannot hold attention.

[palmotea]

> Thought and creativity won't be averaged away because human beings have a drive for these things. This just raises the bar for it. And why not? We get complacent when not pushed.

The why not is: human beings are valuable in and of themselves, not just because of what they can do. If you raise the bar too high, you kick people out. And our society just isn't setup for that, and is unlikely to ever be in our lifetimes.

And I'm talking about a radical shift in the concept of ownership, where shareholding is radically democratized. Basically every random Joe needs the option to live comfortably on passive income generated by things he owns.

[kruffalon]

But it's a weird kind of average... Not the 3 from 1, 2, 3, 4 & 5 but rather like the bland tv-dinner which tastes non-upsetting for most people.

[tovej]

It's more like a blur filter and a thousand layers of jpeg compression.

[jacamera]

Great read: https://www.newyorker.com/tech/annals-of-technology/chatgpt-...

[EarlKing]

An intellectual Mode rather than a Mean or a Median?

[kruffalon]

I don't understand what you mean by "intellectual mode".

I mean that it's a kind of lowest common denominator average where it's more important to seem reasonable and to not upset anyone rather than be really good in some ways and bad in others.

[papyrus9244]

> I don't understand what you mean by "intellectual mode".

https://en.wikipedia.org/wiki/Mode_(statistics)

If human knowledge were a pyramid, LLMs just make the pyramid flatter, i.e. shorter, wider at the bottom, and narrower at the tip. It makes Humans dumber.

[kruffalon]

Thank you!

The capital M had meaning that I didnt grasp since I hadn't heard of Mode in that way before.

Today's learning!

[jibal]

https://stats.stackexchange.com/questions/200282/explaining-...

[kruffalon]

What a great resource, thank you <3

The comment by Joseph Greenpie[0] is just marvellous, what a gem!

-----

[0] https://stats.stackexchange.com/a/204558

[jibal]

The comment is actually by https://stats.stackexchange.com/users/107126/vishal ... Joseph Greenpie made the last edit to it.

[ModernMech]

The soft gaussian blur of all human knowledge.

[thirtygeo]

Racing towards average!

[larodi]

Mediocre is the word perhaps :D

[altairprime]

Perhaps closer to “the mean vector point such that all outbound vectors to different training tests are in sum the smallest”? I assume that’s a property of neural networks anyways, though I’m out of date on current math for them.

[ludicrousdispla]

If you want a more accurate measure then you should subtract "the sum of all human ignorance" before taking the average.

[pessimizer]

> I've been feeling more and more that generative AI represents the average of all human knowledge.

No, it's far worse. It's the mode of all human knowledge. The amount of effort you have to put into an LLM to get it to choose an option that isn't the most salient example of anything that could fit as a response is monumental. They skip exact matches for most common matches; it's basically a continuity from when search engines stopped listening to your queries and just decided what query they wanted to respond to - and it suddenly became nearly impossible to search for people who had the same first name as anyone who was famous or in the news.

I've tried a dozen times to get LLMs to find authors for me, or papers, where I describe what I remember about them fairly exactly. They deliver me a bunch of bestsellers and popular things, over and over again, who don't even match at all large numbers of the criteria I've laid out.

It's why they're dumb and can't accomplish anything original. It's structural. They're inherently biased to deliver lowest common denominator work. If you're trying to deliver something original or unusual, what bubbles up is samplings of the slop that surrounds us every day. They're fed everything, meaning everything in proportion to its presence in the world. The vast majority of things are shit, or better said, repetitions of the same shit that isn't productive. The things that are most readily available are already tapped out. The things that are productive are obscure.

You can't even get LLMs to say some words by asking them to "say word X." They just will always find a word that will fill that slot "better." As I said, this is just google saying "did you mean Y?" But it's not asking anymore, it's telling.

edit: It's also why asking it to solve obscure math problems is a dumb test. If the math problem is obscure enough, and there's only one way to possibly solve it, and somebody did it once, somewhere, or referred to the possibility of solving it that way, once, somewhere, you're going to have a single salient example. It's not a greenfield, it's not a white sheet of paper: it's a green field with one yellow flower on it, or a piece of white paper with one black sentence on it, and you're asking it to find the flower or explain the sentence.

edit: https://news.ycombinator.com/item?id=47346901 - I'm late and long-winded.

[red_hare]

I feel the same about Claude Code. It's a fast but average developer at just about everything and there are some things that average developers are just consistently bad at and therefore Claude is consistently bad at.

[Cthulhu_]

I'm not sure, I think you overestimate the average developer. But then, the average code doesn't end up in public repositories, it spends decades in enterprise codebases rotting.

At this point I'd rather review LLM generated code than a poor developer's.

[baxtr]

Yes, it’s the "sum" of which you extract an average.

[oblio]

> I've been feeling more and more that generative AI represents the average of all human knowledge.

It's literally what it is. Fairly sure that mathematically it's a fancier regression/prediction so it's a form of average.

[permo-w]

You're falsely conflating knowledge with intelligence

[ninjagoo]

> I've been feeling more and more that generative AI represents the average of all human knowledge.

Have you tried the paid versions of frontier models? They certainly do not feel like they spew the average of all human knowledge. It's not uncommon for them to find and interpret the cutting edge of papers in any of the domains that I've asked them questions about.

[fuzzer371]

Yup. And they all sound like slop. Read the papers, comprehend the papers, don't make someone else's computer do it for you.

[Otterly99]

Every scientist I ever met (and myself included) has a backlog of papers to read that never seems to shrink. It really is not trivial to stay up to date on research, even in niche fields, considering the huge volume of research that is being produced.

It is not uncommon for me to read a recently published review and find 2-3 interesting papers in the lot. Plus the daily Google scholar alerts. It can definitely be beneficial to have a LLM summarize a paper. Of course, at this point, one should definitely decide "is this worth reading more carefully?" and actually read at least some parts if needed.

[codemog]

Anti-tech contrarian sentiment happens with every new technology. Someone older than you probably said the same thing about the internet.

[BuddyPickett]

Yep. Even windows, the most widely used OS on the planet has a fringe group of contrarians still today. Amazing.

[Xfx7028]

I grew up using windows and was a fan of it, but now I am a contrarian because of how shitty it has become. The fact that it is widely used is not an argument that it is good. It is widely used because of existing market share and reluctance of change by people.

[xigoi]

Even tobacco, the second most widely used drug, has a group of contrarians still today. Amazing.

[array_key_first]

Being widely used means almost nothing. There's a lot of reasons things can be widely used. Inertia, history, coercion, cheating, lying, addiction.

Windows is the most widely used because of history and inertia. It is not the best option in just about any metric, and a lot of this is objective. It's slow, the software is poorly designed and obtuse to use, and it lacks functionality. However, people already kind of know it, so it remains.

If Microsoft didn't get the IBM gig early on, we would not be using Windows.

[jibal]

What's sad is that there's so much of that at this site. This page in particular is a disaster, and what we're actually seeing a lot of at HN is claims that real humans are bots. And the people who make these accusations are certain of their validity.

[toraway]

Have you considered that this suspicion is because the number of obvious bots has exploded in the last half year or so, particularly after OpenClaw became the latest fad?

Start going to the profiles of every comment from a green account you see for a week and you’ll see how bad it is.

There will be friendly fire but unfortunately that’s to be expected when you click the top comment in a thread and realize an account has been posting 100% slop for months.

[jibal]

What I see is massive intellectual dishonesty, like this comment that doesn't engage with my actual points and instead attacks strawmen.

I won't comment further.

[account42]

That's not sad, it what makes this place still worth visiting.

[streetfighter64]

And they were right, the internet does make us dumber and less human.

[selcuka]

True, and they were right about it when they said that. They wouldn't be right anymore, because the Internet has evolved. The same might happen to LLMs, but currently one would be right to call LLM output "slop".

[darkwater]

Depending on the criticism at the time, they were probably wrong at the time and are correct now. There were always trolls and bad people but at least there were no mega-corp playing with people's minds.

[ninjagoo]

> Read the papers, comprehend the papers, don't make someone else's computer do it for you

Why not?

Personally, I don't have the specialized knowledge, nor the time needed, to read and understand papers outside my own 2-3 domains. LLMs do. And I appreciate what they can do for me. They do it better, faster, and more accurately than most 'popular science', provide better coverage and also provide the ability to interact with the material to any degree or depth that I care to, better than any article.

It would be silly to pass up this capability to make my life better simply because random folks on the Internet disparage the quality of the output (contrary to my own experience) and make hand-wavy points about 'someone else's computer) while offering no credible or useful alternative :)

[framapotari]

How do you evaluate the quality of a summary of a paper you do not have the knowledge to read and understand?

[ninjagoo]

> How do you evaluate the quality of a summary of a paper you do not have the knowledge to read and understand?

Tough question. I think the straightforward answer is that you can't.

That said, there is some confidence gained in an LLM's abilities based on its performance on papers in domains that I do understand. Yes, it's not going to be the same across all domains, but the frontier labs do publish capability scores across different domains, and that helps scrutinize the answers it provides, and how much salt to take with those.

[kruffalon]

I wonder if you have asked the same LLMs to explain or summarize a paper in one of your fields and see if it still makes sense.

It could be that the LLMs are good at stringing words together in a way that seems reasonable when you are not an expert yourself, much like people from other fields seem very knowledgeable until you compare many of them or hear/see them talk with each other.

[ninjagoo]

> I wonder if you have asked the same LLMs to explain or summarize a paper in one of your fields and see if it still makes sense.

I have, and it does, hence my confidence in its ability to do the same in other domains. Depending on what you're using it for, it is advisable to maintain some level of quality control (spot checks, sampling, deep dives, more rigorous continuous review) as in any process control.

[kruffalon]

Nice, that's good to hear and from the Zeitgeist that I get kind of new if I understand it correctly.

[larodi]

pooling as it is called, is, well the same as averaging. has nothing to do with swimming really. it happens all the time in latent space. it is a tool, not a side effect.