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

AI’s Instagram Problem

deeplearning.ai

Post reply on HN
Print view View original

91–99 of 99 posts

Re: AI’s Instagram Problem

#91

Post by beepbooptheory »

pixl97 wrote: post #90
beepbooptheory wrote: post #56

Earlier quoted context omitted.

Maybe, if you have the time, do me a favor: read the (very short) piece I am responding to, its the link at the very top of the page. If you still have a thoughtful or even snarkey comment to make to me after that, then please go ahead. But you're not really giving me anything to work with here, and around here we do strive to have real conversations. I just know you have some good points to make to me! Maybe chopsti…

Ok, so lets split the conversation into fine art versus generic commercial art and design we see everywhere. Because as I see it, this is how the future is splitting out. This goes back to the blacksmith producing unique items. People still buy things like to this day and pay a healthy sum of money for it. But this is not the economy either, the vast majority of consumption is items that are mass produced with as lit…

This doesn't make sense to me as a response. I think it surely goes without saying the models are good for things like game asset design, or ambient music, I say exactly that in a sibling comment. (although the ambient music one is a little silly to me personally. Brian Eno already invented procedures and ideas for generative ambient works without a single GPU like 50 years ago. Not to mention Terry Riley, John Cage, Steve Reich. Consider also the algorithmic nature of classical Indonesian gamelan music. Would be an accomplishment if anything from a statistical model could match the beautiful, simple, almost dumb elegance of such works. It feels like such overkill. But I digress..).

The article which I'm responding to seems to me to be about not being demoralized as an artist in the face of advances in AI, presumably to artists who aren't just trying to get a commission check for some elevator music, but care about what they are making, hence the possibility of being demoralized. And my point is that they don't need to worry, and that it only seems this way right now because of enthusiasm for the idea of AI art itself, not any one work.

In a way by, intentionally or not, eliding this crucial context, you are helping me prove my point! We cant separate the value/worth/goodness of an AI work without reaching outside it to talk about what it means, its reference to a coming future, its bare ability to be "human like", or even just matching the spec, as it were.

People are just so caught up in justifying this stuff right now, they don't have even the ability to consider it in itself. This is all I'm saying.

I think if you read what I'm saying you could agree with me without giving up any of your commitments here, just dont be quick to jump on something that seems wrong without 1. giving charity in what the other is saying, 2. understanding the context in which it is said, in this case the article I was responding to.

Re: AI’s Instagram Problem

#92

Post by softwaredoug »

kanyethegreat wrote: post #62
softwaredoug wrote: post #4

There is a pretty big gap between what’s talked about at conference and blogs vs the actual, not very sexy, hard data science work that goes into doing “machine learning” at scale.

where can i get more content of the flavor hard data science work that goes into doing “machine learning” at scale ? i'm tired of indefensible projects built on OpenAI's APIs

Well some conferences tend to reward more concrete / practical work. I helped found Haystack (http://haystackconf.com), as one such idea. But still, people who propose talks, aren't going to talk about the mundane work. In the same way people who submit talks about software aren't going to talk as much about the mundane, everyday, uninteresting -- but essential -- unit tests they wrote.

Re: AI’s Instagram Problem

#93

Post by malloci »

moomoo11 wrote: post #64
version_five wrote: post #14

This is correct, but there's also some meta comment to be made that research has an "instragram problem". What he's talking about isn't really even AI so much as deep learning: there are lots of other branches that now get virtually no play relative to deep learning. And then there are all sorts of adjacent CS, image processing (when was the last time you saw someone talk about wavelets), language analysis, and other…

The goal of academic funding is to help the military and also/later the economy. To think otherwise is naive. The whole point of research is to have an advantage. Unless that money has no strings attached, you’re beholden to someone. Every historical scientist had a major donor whether it was some kingdom or nation.

Call me naive then. While I agree that a fair amount of academic research funding is defense/military related, it’s certainly not all of it. If said research is funded by organizations like DARPA, then yes, absolutely. There are other sources of funding out there though that isn’t defense backed: NSF, NIH funding comes to mind among others, including corporate backed academic research. When i was in academia, I worked on research funded by grants from NSF, Intel and Microsoft to name a few.

Re: AI’s Instagram Problem

#94

Post by jobigoud »

matsemann wrote: post #79
matwood wrote: post #34

Earlier quoted context omitted.

IDK. Seems like usual social media that it’s all about who you follow.

No, I follow athletes related to cycling, swimming, skiing etc. What content does Instagram promote me? Female models dressed in cycling clothes with their zippers down. Maybe because their algorithm says that it's the "most trending cycle-tagged content at the moment". But it's not related to the professional cyclists I'm following at all. Same goes for almost all content/tags: there will be some half-nude version o…

I follow art-related stuff and nature photography, mainly digital illustration and I do NOT get any soft-erotica stuff. I only get what I'm interested in.

I will say that I prefer going to the Explore tab rather than the Reels tab because this one starts by showing relevant stuff and then as I scroll down it turns into tik-tok skits.

Re: AI’s Instagram Problem

#95

Post by jobigoud »

beepbooptheory wrote: post #67
catach wrote: post #63

Earlier quoted context omitted.

> Human creations are far better still than AI ones. I have seen many examples of AI-generated art where it's beyond my taste level to see how it could be improved on by human hands. And stuff that I genuinely appreciated for what it is, divorced from how it was made. I am reminded that very early on with Midjourney, before this wave of AI gen art was wider known, there was a story of someone winning an art competiti…

I hear you but: > I have seen many examples of AI-generated art where it's beyond my taste level to see how it could be improved on by human hands This is my overriding point, beyond matters of taste. Everyone who talks about this stuff can't seem to talk about its value without reference to the hypothetical human artist or writer or whoever who didn't make it. The awe of just the fact it was made by AI can't be sepa…

> You say you could appreciate it without reference to the origin, but how can any of us claim that?

Easy, you tap the like button without wondering if it was made by/with AI or not. I'm sure many people that don't really care about the tech and just want a endless feed of cool pictures do that.

Re: AI’s Instagram Problem

#96

Post by beepbooptheory »

jobigoud wrote: post #95
beepbooptheory wrote: post #67

Earlier quoted context omitted.

I hear you but: > I have seen many examples of AI-generated art where it's beyond my taste level to see how it could be improved on by human hands This is my overriding point, beyond matters of taste. Everyone who talks about this stuff can't seem to talk about its value without reference to the hypothetical human artist or writer or whoever who didn't make it. The awe of just the fact it was made by AI can't be sepa…

> You say you could appreciate it without reference to the origin, but how can any of us claim that? Easy, you tap the like button without wondering if it was made by/with AI or not. I'm sure many people that don't really care about the tech and just want a endless feed of cool pictures do that.

Yes, and I bet if you were keeping score of someone liking an image or not on their feed, and it was truly blind, and there was even distribution, humans would still win.

Re: AI’s Instagram Problem

#97

Post by atq2119 »

kanyethegreat wrote: post #60
gnramires wrote: post #19

Earlier quoted context omitted.

One advise I'll never forget (from an interview with late mathematician Maryam Mirzakhani) is that you don't have to race faster than everyone else (although if you can, that's great!). You can just run in a direction no one is running toward (and hopefully be persistent). Then you're only racing yourself and speed is largely irrelevant. I myself know I'm very slow, specially developing projects. So I don't try to ra…

“Run in the other direction” is so generic that it isn’t helpful. No one else going in that direction is, at best, unrelated to the quality of the idea and, at worst, a heuristic that the idea is bad.

It's not the other direction, it's an other direction. The distinction is important.

If there are only two directions and everybody runs in one of them, there's probably a good reason. If there are virtually infinite directions, trusting your gut about an unexplored direction that you find interesting is a much better bet. See also: startup advice.

Re: AI’s Instagram Problem

#98

Post by moomoo11 »

malloci wrote: post #93
moomoo11 wrote: post #64

Earlier quoted context omitted.

The goal of academic funding is to help the military and also/later the economy. To think otherwise is naive. The whole point of research is to have an advantage. Unless that money has no strings attached, you’re beholden to someone. Every historical scientist had a major donor whether it was some kingdom or nation.

Call me naive then. While I agree that a fair amount of academic research funding is defense/military related, it’s certainly not all of it. If said research is funded by organizations like DARPA, then yes, absolutely. There are other sources of funding out there though that isn’t defense backed: NSF, NIH funding comes to mind among others, including corporate backed academic research. When i was in academia, I worke…

Im just saying that the end goal of your research is return on investment.

So many folks get no funding, even if their field of work is important to society at large.

Intel and whoever else funded you because they see some long term value in your work.

I wish research was done with “purer” intentions but hey.. I mean we are having this conversation based on a technology invented by the gov/military, Arpa. On my drive home I used GPS. And at home I’m using a microwave to make some popcorn lol.

Re: AI’s Instagram Problem

#99

Post by kanyethegreat »

tysam_and wrote: post #83
kanyethegreat wrote: post #61

I've definitely been feeling deflated. Every angle I've come up with... some project has narrowly shipped before me. And they gain so much popularity so quickly that mindshare per project becomes a power law overnight. It's like the first person to release is at the top of the App Store, and there's no unseating them; the feedback loop is already too powerful. It very much feels like first to release wins. And I swea…

I would not put so much stock in the first mover effect. I can't bring it to mind immediately, but there was an excellent podcast that brought up how the second movers oftentimes do better in the end, at least as a company. Case in point -- once upon a time, there was a big race called Dawnbench for training CIFAR10 to 94% accuracy in the shortest amount of time a little while ago by Stanford University. During that…

Thank you very much for your thought-out response. I really appreciate the effort. I will certainly watch the linked video.
Post reply on HN