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

I'm becoming AI-blind

cymerys.com

Post reply on HN
Print view View original

161–170 of 533 posts

Re: I'm becoming AI-blind

#161

Post by KronisLV »

causal wrote: post #30

There's some psychological mechanism by which my brain immediately recognizes AI generated text and just short-circuits to "there is no information here". And when I force myself to read AI-generated text I realize I'm making my brain do creative work to impart meaning to the words. It is exhausting because my brain is literally trying to do a just-in-time rewrite of the text into something valuable. Something is dee…

> just short-circuits to "there is no information here"

That is my experience with the way the models write by default, often even when instructed not to do that. With enough effort you can get even them to slightly unslop the writing so it doesn't read like some LinkedIn/Buzzfeed brainrot, but the problem is that it's not trivial to do and most people won't do it, so the default is indeed horrible.

Re: I'm becoming AI-blind

#162

Post by diego_sandoval »

causal wrote: post #30

There's some psychological mechanism by which my brain immediately recognizes AI generated text and just short-circuits to "there is no information here". And when I force myself to read AI-generated text I realize I'm making my brain do creative work to impart meaning to the words. It is exhausting because my brain is literally trying to do a just-in-time rewrite of the text into something valuable. Something is dee…

When I read AI-generated prose that is aimed at the general public, I have the exact same feeling.

But when I ask Codex a technical question about coding, I don't get it at all. Codex replies to me in a very direct, technical manner, similar to the way I speak.

When I ask ChatGPT to be concise and technical, I get the same effect.

I think it's because prose aimed at the general public has to be very attention-baity --like the textual equivalent of a Mr. Beast video--, not because AI is incapable of writing like a human.

Re: I'm becoming AI-blind

#163

Post by fwip »

dofm wrote: post #32
pmarreck wrote: post #13

Earlier quoted context omitted.

That's the cost of an AI work that is derived from the outputs of others. There's no real edge to it. Same as with the writing. The stuff that you'd latch onto (and thus remember) is simply not there, precisely because those image or word choices would be just outside its latent probability space. But because they're well inside it, your mind sees nothing novel to register. This is also why I think human output will…

> There's no real edge to it. Yes, as if it is in some weird hidden dimensional sense completely uniform. ETA: suddenly reminded of the Bateson quote about information being “the difference that makes a difference”.

That checks out, right? We're instructing LLMs to choose from the least "surprising" tokens at each step (modulo some temperature). It feels smooth, slippery - no friction to the eyes that glaze over as they skim the text, and no bumps to snag in your memory. Like waking from a dream, recalling only the barest shapes of a few concepts or themes, until it vanishes with your morning coffee.

The examples I've seen of AI music (though I avoid it on principle) seem the same.

Re: I'm becoming AI-blind

#164

Post by BobbyJo »

diego_sandoval wrote: post #162
causal wrote: post #30

There's some psychological mechanism by which my brain immediately recognizes AI generated text and just short-circuits to "there is no information here". And when I force myself to read AI-generated text I realize I'm making my brain do creative work to impart meaning to the words. It is exhausting because my brain is literally trying to do a just-in-time rewrite of the text into something valuable. Something is dee…

When I read AI-generated prose that is aimed at the general public, I have the exact same feeling. But when I ask Codex a technical question about coding, I don't get it at all. Codex replies to me in a very direct, technical manner, similar to the way I speak. When I ask ChatGPT to be concise and technical, I get the same effect. I think it's because prose aimed at the general public has to be very attention-baity -…

I use Claude and I find that it speaks in a very obfuscated manner when explaining things. It seems to make up jargon as it goes on top of spending a lot of tokens dancing around a point. I often find myself having to ask it to rephrase things, or speak directly about mechanism or consequence, in order to understand the point.

Re: I'm becoming AI-blind

#165

Post by fallingbananna »

tingletech wrote: post #86

In high school, a teacher gave me a copy of "How To Read Better And Faster" which teaches you speed reading. This came in very handy in college. I find that when I try to speed read modern human writing, there are often errors (like missing or misused words) or awkward expressions that I do have to slow down and think harder a lot to really parse it. With AI writing, it's sort of self redundant and the information de…

The density argument is very interesting.

Does speed reading help you process the final message faster if it's written by AI compared to people?

Because if you read 1 information dense sentence, 1 medium dense, and 1 sparse sentece written by a human, it's still way less text in total than 6 information sparse sentences written by AI... even if it's all over the place when it comes to density or style.

---

The density argument is really interesting.

Does speed reading actually help you process the final message faster when it’s AI-generated compared to human-written?

For example, if a human writes 3 sentences—one information-dense, one medium-density, and one sparse—that’s still much less text overall than 6 relatively sparse sentences written by AI.

Even if the AI output varies a lot in information density and writing style, you still have to process all that additional text. So I’m wondering whether speed reading actually offsets the verbosity of AI-generated responses, or whether the total amount of text is still the bigger factor.

Re: I'm becoming AI-blind

#166

Post by disgruntledphd2 »

BobbyJo wrote: post #164
diego_sandoval wrote: post #162

Earlier quoted context omitted.

When I read AI-generated prose that is aimed at the general public, I have the exact same feeling. But when I ask Codex a technical question about coding, I don't get it at all. Codex replies to me in a very direct, technical manner, similar to the way I speak. When I ask ChatGPT to be concise and technical, I get the same effect. I think it's because prose aimed at the general public has to be very attention-baity -…

I use Claude and I find that it speaks in a very obfuscated manner when explaining things. It seems to make up jargon as it goes on top of spending a lot of tokens dancing around a point. I often find myself having to ask it to rephrase things, or speak directly about mechanism or consequence, in order to understand the point.

GPT is less bad for this, which is why I've mostly shifted to using it.

Re: I'm becoming AI-blind

#167

Post by layer8 »

jihadjihad wrote: post #20
pingou wrote: post #12

Earlier quoted context omitted.

It doesn't look like a quiche but more like a cake to me, and the top would be torched meringue, not mold. Although it's probably some weird ai mix of quiche and cake.

> probably some weird ai mix of quiche and cake Quike? Cache?

cAIche

Re: I'm becoming AI-blind

#168

Post by sebmellen »

BobbyJo wrote: post #164
diego_sandoval wrote: post #162

Earlier quoted context omitted.

When I read AI-generated prose that is aimed at the general public, I have the exact same feeling. But when I ask Codex a technical question about coding, I don't get it at all. Codex replies to me in a very direct, technical manner, similar to the way I speak. When I ask ChatGPT to be concise and technical, I get the same effect. I think it's because prose aimed at the general public has to be very attention-baity -…

I use Claude and I find that it speaks in a very obfuscated manner when explaining things. It seems to make up jargon as it goes on top of spending a lot of tokens dancing around a point. I often find myself having to ask it to rephrase things, or speak directly about mechanism or consequence, in order to understand the point.

Using Claude for any kind of technical writing makes me feel like it was trained on snarky Huffington Post articles written by a 23 year old mixed media arts graduate and then was told to intentionally obfuscate the most important elements of any text by extensively rambling about what was not done and for what reason.

Re: I'm becoming AI-blind

#170

Post by tyromaniac »

rnd33 wrote: post #108
bayindirh wrote: post #80

Earlier quoted context omitted.

It's just filled to the brim with relations between things. It's good at searching a very large meaning space and create correlations. What it does is to cover great distances and find related things in that large space which needs a long time and large corpus of knowledge to find the connection. This is not intelligence. It's just a good correlation engine with a very big albeit lossy database of things.

Intelligence is compression, compression requires subtraction, and for some reason LLMs are not good at subtracting. To create a coherent model you kinda have to subtract correlations until only the essential parts are still there. What I don't understand is why LLMs haven't been able to do this yet, if it's the harness or some orchestration layer above the LLM that is needed. Because fundamentally if you can identif…

I've heard that expression before, but I don't think it can be presented and stated so matter of factly. Where does that put bzip?
Post reply on HN