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

Not by AI

notbyai.fyi

Post reply on HN
Print view View original

391–400 of 456 posts

Re: Not by AI

#391

Post by jay_kyburz »

radlad wrote: post #328
underyx wrote: post #17

> If humans rely solely on AI to generate content moving forward, any new content generated by AI may just be recycled content from the past. To be honest, I’d struggle to find any human content that is not recycled from the past. I guess we are still better than AI at remixing sufficiently novel combinations of concepts but I think much of this AI content aversion comes from some kind of phobia, or an unwillingness…

What I find amusing is not the statement you quoted, but this one: > The Not By AI badge is created to encourage more humans to produce original content and help users identify human-generated content. How does this encourage humans to produce more original content? It may help users identify human-generated content - if they care. But perhaps more usefully it helps AI identify human-generated content to avoid traini…

AI will just adding the badge to the content it generates.

Re: Not by AI

#392

Post by ayewo »

jaqalopes wrote: post #379
Veen wrote: post #369

Earlier quoted context omitted.

I’m not so sure an AI couldn’t, in theory at least produce a memoir if you fed it your diaries, emails, social media chats and histories, photos and videos from throughout your life, and so on.

This feels like a good point. I am curious how the "feeding" would occur here. If there's an API to do this sort of thing right now I'm not aware of it. But I definitely believe that it's coming!

The recently announced GPT-4 [1] now supports feeding it with more “context” when writing a prompt than previous models.

It supports a context length of 32k tokens (approximately 25,000 words or about half a novel).

1: https://news.ycombinator.com/item?id=35154527

Re: Not by AI

#394

Post by wizzwizz4 »

jrussino wrote: post #358
raincole wrote: post #104

Earlier quoted context omitted.

The whole idea of "there is no such thing as an original idea" is utterly stupid. It's just a hyperbole and some people are taking it literally. Of course there are original ideas. "C is not an original idea, because it's just mixing A and B..." ok so there are only two possbilities: 1. Someone else did mixing A and B before. Okay then we can recursively find the first person who mixed A and B. 2. No one else did mix…

If we follow this line of thinking: in the, let's call it, "pre-AI" days all of the ideas that we have of the form "Mixing A and B" are ones that at least one human found to be interesting/valuable (and the ones that become common and get remixed must have been interesting/valuable to many humans). Maybe the "originality" of the idea isn't the thing we care about, per se, but rather that fact that somebody though it…

> I dunno, seems like we're heading toward a sort of "pollution of the idea atmosphere" where we get a ton of new/remixed ideas dumped into the world that haven't passed through that initial filter of a human vetting them

We already have that. Observe: ahem “It is an industrial capacity, that many people like the Reddit poem person who does the cow thing. And everyone delighted in the nonsense of words that are sense.” Or, perhaps you want something on-topic? “Your ideas about ideas are mere ideas. Ideas that are ideas are ideas, but these ideas are not ideas. lol!” Perhaps an unoriginal, unfiltered synthesis of some good ideas? “It is a truth universally acknowledged that a potato peeler could save you 5 Altairian dollars a day.”

We don't need AI to produce unoriginal, unfiltered internet comments. Certain YouTube comments sections provide a great example (though the problem with YouTube comments isn't as bad as people tend to make out). We humans already have social mechanisms to keep the good ideas and ditch the bad ideas, even when many of the humans aren't bothering to filter their own output.

Re: Not by AI

#395

Post by famouswaffles »

varispeed wrote: post #291
underyx wrote: post #17

> If humans rely solely on AI to generate content moving forward, any new content generated by AI may just be recycled content from the past. To be honest, I’d struggle to find any human content that is not recycled from the past. I guess we are still better than AI at remixing sufficiently novel combinations of concepts but I think much of this AI content aversion comes from some kind of phobia, or an unwillingness…

The difference is that AI is using pattern matching while humans use reasoning to come up with new things. That means AI is not capable of producing meaningful _new_ content like discovering new mathematical theorems, because AI does not understand maths, whereas humans can come up with something meaningful based on _understanding_ of the content they have learned from. This is why when you ask e.g. ChatGPT about som…

Complete nonsense. https://www.nature.com/articles/s41587-022-01618-2

Language models can generate novel functioning protein structures that adhere to a specified purpose. Structures that didn't exist before nevermind found in the dataset. The idea that there's some special distinction between the reasoning LLMs do and what Humans do is unfounded nonsense.

A distinction you can't test for (this so called "true understanding" ) is not a distinction

Re: Not by AI

#397

Post by famouswaffles »

jeffbee wrote: post #39
raven105x wrote: post #16

Earlier quoted context omitted.

I realize this is in jest, but to point out the obvious, ML (AI is such a bullshit term. ANI/AGI/ASI, etc) models are trained on human written text. With ANI or AGI, this rapidly devolves into a chicken or the egg scenario. You know what else does? The cold war & arms proliferation. By the time ASI is actualized, humans will no longer be writing at all and languages as we know them will be obsolete. There is a lot of…

It is only mostly a joke. OpenAI's LLMs have a definite voice, and that voice is "your real estate agent is also your manager". If a person finds ChatGPT's output good enough to turn in as an assignment at school, or mail to a colleague, I think that person has little appreciation for the possibilities of language.

They don't have a default voice not a definite one. Output can be easily adjusted to sound like whatever

Re: Not by AI

#398

Post by idontpost »

mrob wrote: post #258
mjr00 wrote: post #127

Earlier quoted context omitted.

> Your DAW example would work if DAWs could generate samples, beats, and whole songs by the click of a button. But are you truly the author, then? They absolutely can, at least with 15-20 clicks rather than 1 at the moment. Drag in drum loop, drag in melodic loop, use Scaler2/Cthulhu/Captain Chords to come up with a chord progression, use a synthesizer preset to play the chords, drag in vocal sample, put on Ozone AI…

There's a natural stopping point from a musical perspective, and that's learning to play the drums. Every step beyond that only affects image or cost. There's a huge variety of commercially available drums, so anything you build will likely sound like something you could buy. Even if you make something weird, it will sound like a modified commercial drum, and there are commercial products for drum modification too.

[dead]

Re: Not by AI

#400

Post by gpderetta »

This is a covert operation to get human generated contented to be tagged appropriately to be incorporated to the next training set for GPT-5/SD3.0.

/S but only a bit.

Post reply on HN