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

Coarse is better

borretti.me

Post reply on HN
Print view View original

101–110 of 125 posts

Re: Coarse is better

#101

Post by wombatpm »

JeremyNT wrote: post #76
jamblewamble wrote: post #72

Earlier quoted context omitted.

It is what has underpinned all of human progress towards automation. It isn't a bad thing. Every time we automate something the luddites cry out about the coming mass unemployment. It has never happened.

The promise is to automate the drudge work, freeing people to pursue their passions. Like, you know... creating art.

But most work IS drudge work and the automation causes new different drudgery. Use to be you could dictate a letter and someone from the typing pool would clean it up, proof it, and send it. Now those same people get to write their own crappy email themselves

Re: Coarse is better

#104

Post by tim333 »

coldtea wrote: post #91
jamblewamble wrote: post #72

Earlier quoted context omitted.

It is what has underpinned all of human progress towards automation. It isn't a bad thing. Every time we automate something the luddites cry out about the coming mass unemployment. It has never happened.

> Every time we automate something the luddites cry out about the coming mass unemployment. It has never happened It has happened each and every time, it just haven't affected you personally. Starting of course with the original luddites - they didn't complain out of some philosophical opposition to automation. Each time in changes like this a huge number of people lost their jobs and took big hits in their quality o…

Strictly speaking outsourcing to cheap labour isn't automation.

Re: Coarse is better

#105

Post by dan-robertson »

airstrike wrote: post #37

I'm no image gen expert but these prompts are downright terrible even by my standards. Are you really complaining that ", from the British Museum." leads to it a painting in the actual British Museum? Just remove the sentence, and you'll be fine. Now good luck trying to make Midjourney place the image at the museum! I'm a paying MJ user and am impressed by Nano Banana. They're different models. They each serve their…

These sorts of prompts used to be quite important when DALL-E was new. I do feel like a lot of the article is just that prompts should be written differently though I think there’s some truth in the idea that nanobanana feels less artistic in some ways.

Re: Coarse is better

#106

Post by gwern »

wrsh07 wrote: post #67
airstrike wrote: post #37

I'm no image gen expert but these prompts are downright terrible even by my standards. Are you really complaining that ", from the British Museum." leads to it a painting in the actual British Museum? Just remove the sentence, and you'll be fine. Now good luck trying to make Midjourney place the image at the museum! I'm a paying MJ user and am impressed by Nano Banana. They're different models. They each serve their…

It is interesting what the nbp model takes away from the prompt, though Eg instead of focusing on the artist, it focuses on the location This makes sense! I imagine it was trained in some sort of rlvr like way where you give it a prompt and then interrogate "does this image ..." (where each question examines a different aspect of the prompt) It's obviously an incredible model. I think there's a limit to how useful an…

> I would also welcome someone writing a short takedown where they fix the prompts and get better-than-2022 results from nbp

NBP (and the new ChatGPT generator) are integrated with LLMs to various degrees, so seems like the obvious starting point is a reverse approach: ask them to describe the old images which has the esthetics that Fernando Borretti likes, and start generating from those prompts. If you can recover the old images, then it was just a prompting issue. ("Sampling can show the presence of knowledge but not the absence.") If you can't even with their own 'native' descriptions, then that points to mode-collapse (especially all of the 'esthetic tuning' like DPO everyone does now) as being the biggest problem.

Re: Coarse is better

#108

Post by flexagoon »

speedgoose wrote: post #78
yoan9224 wrote: post #66

The author's prompts are fighting against what Nano Banana was optimized for. Saying "British Museum" to MJv2 worked because it blurred all images tagged with museums into the aesthetic. NBP interprets it literally: show me something IN a museum. This isn't worse - it's different. MJv2 was a happy accident machine. NBP is a precision tool. If you want the coarse aesthetic, prompt for it: "rough brushstrokes, visible…

Thanks ChatGPT. I’m wondering about the motivation to spam HN with LLM generated comments. Not the worst comments though.

I don't think that comment is LLM generated. I would've certainly written it like that myself.

Re: Coarse is better

#109

Post by vlovich123 »

loeg wrote: post #80
vlovich123 wrote: post #74

Earlier quoted context omitted.

Except all the manufacturing jobs got shipped overseas and now those people are Walmart greeters or similar unskilled labor. Having a shit job isn’t unemployment but it’s not a huge step up

That isn't what happened. American jobs are more productive than ever. Americans are richer than ever. The modern luddites dramatically underestimate how bad the past was.

> Americans are richer than ever.

By what metric? One way is to look at the Gini coefficient - that’s worse than ever.

The bottom 20% has 2-3% of total net worth in the US. The middle 40% has seen a decline from 36% in 1989 to 28% in 2020. The top 0.1% has seen their net worth capture double from 7% in 1989 to 14% today.

The subtle thing that net worth ignores of course is inflation from growth in costs, so actually it’s harder for most people than in 1989, unless you’re talking about the ease of buying a TV or phone. Technology is more available and cheaper than ever but food and medicine is more expensive than ever.

Re: Coarse is better

#110

Post by spaqin »

userbinator wrote: post #102
efitz wrote: post #96

Why does anyone serious about art want to make art with AI? A large part of the magic of art is the human choices that go into it.

Prompting an AI and then filtering the results is a "human choice".

Two choices - one of prompt, one of the result, versus hundreds or thousands from the subject and composition, through the medium, to every single brushstroke, where one may have a significant meaning. To be improved upon when your skill improves as well.

This is more akin to going to a supermarket and buying peanut butter (prompt: peanut butter, filter by brand/price/taste). The product may be tasty and enjoyable but I am not impressed by that.

Post reply on HN