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

Regression to the Mean: on LLMs and the quiet death of the new

rruxandra.github.io

Post reply on HN
Print view View original

91–100 of 103 posts

Re: Regression to the Mean: on LLMs and the quiet death of the new

#91

Post by yalogin »

This is not a new realization, it’s known. By definition llms are great summarizers not inventors. The value is in having the knowledge readily accessible and to cross check and validate against what’s already known - a huge time saver. It’s not going to come with anything new. It need not even, the value is in making things time efficient.

Of course the llm companies justifying huge investments want to frame it as a technology to invent new things and that is where the disconnect is. They are over indexing on the promise but companies are figuring out that the value is in automation and applying/showing value for existing workflows. Claude focused on this aspect and hence so successful

Re: Regression to the Mean: on LLMs and the quiet death of the new

#93

Post by krona »

kolinko wrote: post #78
krona wrote: post #57

Earlier quoted context omitted.

> They construct space between the points and extrapolate outside of the points. They don't. They interpolate between the points on a manifold.

Manifold vs space - yes, I meant manifold, english is not my first language and I missed this word. The argument stands.

Interpolation vs extrapolation? It's very simple distinction and fundamental. Look at ARC-AGI-3 for empirical evidence, if that's important to you. My point however is purely an a priori one.

Re: Regression to the Mean: on LLMs and the quiet death of the new

#94

Post by CJefferson »

AbsurdCensor wrote: post #29
CJefferson wrote: post #6

This is my biggest concern. Speaking as someone who recently had to read 60 AI generated reports (the whole issue of how much students are using AI is one discussion), it was genuinely soul-destroying reading the same phrases, seem sentence structures, same arguments over and over. Depressed me the whole of the next day.

Wouldn't that be the same even if AI wasn't involved? By and large, it's the same books, the same training for students, the same hammered in structure, so the output from students would be reasonably similar. No one is coming up with new allegories out of The Scarlet Letter at this point.

No, not at all. Students all have different voices. Yes there are similarities but if I read two pieces of work that sound in the same voice I assume there has been cheating — except this year everything is in the same voice.

Re: Regression to the Mean: on LLMs and the quiet death of the new

#96

Post by coldtea »

zwischenzug wrote: post #14

This touches on something I've (and many others) have felt throughout my life, not just since the advent of LLMs. To take a simple example: I grew up with computer games in the '80s where there were no 'physics engines' or frameworks for building games. As a result, each game was an expression of the author's personality somehow. Fast forward to the noughties, games bored me as they mostly looked and felt the same, o…

>Or am I just getting old?

Obviously not, since 20-somethings also dislike that and express the same things openly, which they didn't in the past.

Also obviously not, since one can just see it by comparing e.g. 10 pictures of NY or Chicago or LA vs 10 pictures of Berlin or London or Barcelona cafes from 1980s and from 2020s, and the latter all look alike. It's also been studied (even academically) and has several names: airport aesthetic, international style, placelessness, global monoculture, airspace aesthetic (air as in airbnb), etc.

Re: Regression to the Mean: on LLMs and the quiet death of the new

#97

Post by coldtea »

phillipcarter wrote: post #45
zwischenzug wrote: post #14

This touches on something I've (and many others) have felt throughout my life, not just since the advent of LLMs. To take a simple example: I grew up with computer games in the '80s where there were no 'physics engines' or frameworks for building games. As a result, each game was an expression of the author's personality somehow. Fast forward to the noughties, games bored me as they mostly looked and felt the same, o…

> Or am I just getting old? Perhaps? But I think this is more a case of just not seeking things out. Music is as vibrant and diverse as ever, but not if you're only looking at the top charts run by the music industry. Same deal with games, there's more experimentation and interesting concepts in gaming than ever before, but not from the AAA studios. Now I can't speak for how you vacation, but I've had wonderfully dif…

>Perhaps? But I think this is more a case of just not seeking things out

"If you went for deep dives and looked at fringe stuff there's still variety" is totally compatible with "the world is way more honogenized".

There have always been fringe unique voices, but that's beside the point. The world is not defined by the long tail, but by the top 20% most listened/visited areas/games played/movies watched/etc, that all now look the same.

>Music is as vibrant and diverse as ever, but not if you're only looking at the top charts run by the music industry"

So it's not "as vibrant and diverse as ever" then, since the same "top charts" were much more varied in the past, and they reflect what most people listen to, not just what some music nerds seek.

>Same deal with games, there's more experimentation and interesting concepts in gaming than ever before, but not from the AAA studios.

"There is more experimentation and interesting concepts in gaming than ever before" just not in the games most people play. Well, used to be ALSO in the games most people play.

>Now I can't speak for how you vacation, but I've had wonderfully different experiences between Hamburg, Berlin, Vienna, Prague, Rome, Paris, Montpellier, London, Amsterdam, Oslo, and Florence. I just don't go to the starbucks and instead wander around a bit, optionally picking from a few hit destinations if I feel like it.

Yes, if you go out of your way you can find some places in Hamburg, Berlin, Vienna, Prague, Rome, Paris, Montpellier, London, Amsterdam, Oslo, and Florence that are not all alike. Meanwhile, you could go anywhere in those cities 30 years ago and you'd find uniquely different stuff, whereas now all major streets and areas have the same globo cafe and globo shops.

Re: Regression to the Mean: on LLMs and the quiet death of the new

#100

Post by zwischenzug »

coldtea wrote: post #96
zwischenzug wrote: post #14

This touches on something I've (and many others) have felt throughout my life, not just since the advent of LLMs. To take a simple example: I grew up with computer games in the '80s where there were no 'physics engines' or frameworks for building games. As a result, each game was an expression of the author's personality somehow. Fast forward to the noughties, games bored me as they mostly looked and felt the same, o…

> Or am I just getting old? Obviously not, since 20-somethings also dislike that and express the same things openly, which they didn't in the past. Also obviously not, since one can just see it by comparing e.g. 10 pictures of NY or Chicago or LA vs 10 pictures of Berlin or London or Barcelona cafes from 1980s and from 2020s, and the latter all look alike. It's also been studied (even academically) and has several na…

"placelessness" was coined the year after I was born! So if it's not 'just age', then it's been going on a long time.
Post reply on HN