15 ms·

I started off feeling hopeful/excited by OpenAI, GPT-3, DALL-E, etc, as a creator/writer. As time has progressed and I've used these tools and other machine lea

I started off feeling hopeful/excited by OpenAI, GPT-3, DALL-E, etc, as a creator/writer. As time has progressed and I've used these tools and other machine learning tools to create media, I've become disillusioned. It feels as if this is about to explode an existing problem, namely, there's too much content in the world already. Lowering the effort bar for content creation and generation is, in some ways, going to cause a deluge of low-quality/mid-quality content to be generated. I wonder if we'll see a new generation of "editors" showing up to help curate and provide some guidance in terms of what to play if you're limited on time. I often go through the new releases on Steam and I'm amazed at how many games are released that have zero reviews, or less than 5 reviews, yet people are still constantly releasing more and more. I'm sure some of those games are masterpieces, but will never see the light of day.

pithon 4y ago
> It feels as if this is about to explode an existing problem, namely, there's too much content in the world already Imagine applying that same perspective to software, pre-github.
- emmo 4y ago
  I'm not sure I see the relevance. Could you explain further, please?
  - pithon 4y ago
    In many ways, sites/tools like sourceforge and github made it much easier to publish and obtain open source software, leading to much more software being created and shared. I'm suggesting that "too much content" isn't a problem, and leads to good things.
- kixiQu 4y ago
  Did people feel that way about software?
jillesvangurp 4y ago
We're at the beginning of a rapidly evolving situation where out of the blue it is now possible to direct an AI to produce imagery that is interesting using just language. From there to moving animations is a relatively small step and already being worked on. From there to photo realistic scenes directed by just words not that big of a leap either and also something that is being worked on. Are there constraints, glitches, and limitations. Of course. Has that ever stopped artists producing interesting art? Of course not. People have been telling stories via art ever since people figured out that they could scratch things into rock, clay, and whatever. Perhaps the depressing thing for some artists is that they need to adapt to this and up their game. We've seen the same with photography. Every idiot can now have a decent phone camera and there are plenty of people on Instagram doing not so interesting things with those. Does that invalidate what really good photographers do? No. But it does make their work less special when an AI guided camera in the hands of an absolute amateur can produce photos that are decent. And now with this, you could produce realistic looking photos of things that don't exist by simply asking for them in the right way. More content means the quality standards are raised. There are not a lot of artists left that make a living scratching things in rocks. But there are more artists than ever. The only real limitation is their imagination.
- ido 4y ago
  I don’t want it “whatabout” this, but…Is art really the thing we need to automate right now? There’s so much labor that we more directly need more of and that we could all benefit a lot more from automating it. To give a mundane example off the top of my head- street cleaning. I live in a big city (Berlin) which is often filthy with litter. Obviously there are not enough people cleaning it (or it wouldn’t be). Can’t we automate that instead? How about automating construction? Part of the reason housing is expensive are lot prices but surely a lot of it is also labor cost. It just feels like such a frivolous and relatively useless thing to automate, and a misallocation of resources.
  - luismmolina 4y ago
    It's not about what we need. Scientific research in this case has been the result of looking around and seen what can be done with the technology at hand without caring if it was what we need or don't. If you feel like "we" should be automating other stuff, you are free to make your own contribution, it's not like OpenAI owns the keys to the field.
riffraff 4y ago
> Lowering the effort bar for content creation and generation is, in some ways, going to cause a deluge of low-quality/mid-quality content to be generated. Isn't this the case already? For example, I have recently discovered science fiction poetry exists. The amount of stuff that gets produced in this niche of a niche is baffling already, and growing almost exponentially. I feel human curation and editing will just have to be important again.
hnick 4y ago
> I wonder if we'll see a new generation of "editors" showing up to help curate and provide some guidance in terms of what to play if you're limited on time. Iron Pineapple has an entertaining series he calls "Steam Dumpster Diving" which is similar to this, but it's very focused on Souls-Like games. https://www.youtube.com/watch?v=9KARY5ocvKo&list=PLuY9odN8x9pvIoTDtNOZxzo5m6EsBspAi https://www.youtube.com/watch?v=9KARY5ocvKo&list=PLuY9odN8x9... It does sometimes include games many of us have heard of but there are a lot of unknown games, and plain bad ones (he has a soft spot for student projects too). I think he sees the value of new ideas hidden in jank, as do I, so I like the series. I think the shift will be towards matching personalities, you get to know a reviewer and see what they like and how compatible they are with your own opinions. It doesn't have to be a one-to-one match as long as you're aware how they diverge on certain genres. The curator section on Steam seems to be an attempt at this, but any social media that allows following/subscription could serve too. It's still a significant time investment to keep up to date and to even find that matching personality. I wonder if I actually bothered leaving reviews on Steam, would it suggest people with similar tastes to me?
shreyshnaccount 4y ago
there's also an interesting take that as these generative models become better they will be used for search instead of regular search engines, and social apps would automatically generate content for you, trained on the vast dataset they already have. i think the hardest thing in this trend is going to be assessing 'truthfulness' automatically
simion314 4y ago
> Lowering the effort bar for content creation and generation is, in some ways, going to cause a deluge of low-quality/mid-quality content to be generated. This is not a problem, the issue is with Steam of whoever curates this content for us, with some good filters you can avoid the problem. My first program and game was garbage but it was not on GitHub or Steam to bother anyone, so the issue is not that people have access to good tools (I was using some Visual Studio for students edition) the issue might be that maybe some services have bad filters. For new indie devs, please use whatever tools make your vision , publish them, ignore the haters, you will probably not make a living from Patreon or donations but if you love what you doa nd makes you happy that is the important thing.
LoveMortuus 4y ago
I understand your fear, especially because a very similar, if not the same, thing happened with Unity, where our allowed the creation of just what you're talking about a massive amount of low to mid quality games. I even remember reading articles about how the Unity splash screen has become recognised as a mark for low quality games...
garblegarble 4y ago
I'm still hopeful that this technology will enable writers who have a good idea for a story to build much more on their own without having to be in a position to hire voice actors and artists: whether that's pixel graphic backgrounds for adventure games with decent text-to-speech for characters, or 3D worlds built in Unreal with Metahuman-generated NPCs playing out a curated story within a procedurally generated city. There being lots of low/mid quality games is, I think, a problem that already exists today on app stores -- on Steam especially between Greenlight & vast numbers of games 'releasing' in Early Access... and the more that a single writer with some dev skills can do on their own the more opportunity they have to develop their games as a brand, and get feedback to fire their creativity.
- cmsefton 4y ago
  Yes, I also like the prospect that this tool would allow rapid prototyping and storyboarding for games, graphic novels, films etc. Being able to generate a rough idea of what you want or envisage for a scene and show that to an artist/director of photography/whatever would be very useful.
- j_mo 4y ago
  > writers who have a good idea for a story Or programmers who are bad at art, so, programmers in general ;)
  - thenickdude 4y ago
    I'm a programmer that's bad at art, and DALL-E is amazing for me. It generated me a logo I love for my open source app within a couple of batches. I would have gone logoless rather than slog through trying to find an artist on Fiverr.
dfxm12 4y ago
It feels as if this is about to explode an existing problem, namely, there's too much content in the world already. Lowering the effort bar for content creation and generation is, in some ways, going to cause a deluge of low-quality/mid-quality content to be generated. Like you say, this is an existing situation (I hesitate to call anything that lowers the barrier to entry a "problem"). The printing press opened up printing for more people. As did desktop publishing. Similarly, garage band and bandcamp make it so much easier to produce and publish music. Yeah, there's a ton of books and music out there, including a bunch of junk, but the cream still tends to rise to the top. I'd rather take my chances on having to find a masterpiece in a haystack than have it never get published at all, either because of lack of funding/lack of some specific skill that AI can handle/no willing publisher, etc.
MattRix 4y ago
It doesn’t matter how many bad Steam games there are, because their algorithms prevent you from ever seeing the vast majority of games on the platform. The same is true for videos on youtube and songs on spotify, you won’t find the really low effort/quality content unless you actively seek it out.
irrational 4y ago
I remember back when I was in college the WWW was barely released and search was in its infancy. If you were doing a research paper, trying to use the WWW to search for papers on the topic was laughably useless. Instead you would go to the university library (which still used card catalogs!) and make an appointment with these people who had access to some search system (I can’t remember what it was called). You would sit down with them and explain what you were researching and they would craft search queries in the query language this system used and read over the results with you to see if they sounded promising. They would tweak the query based on your feedback until you got the results you wanted. Then they would print out the results and you could go elsewhere in the library to lookup the papers. It cost money per search, but the university covered the cost. I wonder if we might see something similar with these ai generation tools. People who become experts at crafting the queries to get back the results you are looking for act as intermediaries.

pithon 4y ago

> It feels as if this is about to explode an existing problem, namely, there's too much content in the world already Imagine applying that same perspective to software, pre-github.

emmo 4y ago
I'm not sure I see the relevance. Could you explain further, please?
- pithon 4y ago
  In many ways, sites/tools like sourceforge and github made it much easier to publish and obtain open source software, leading to much more software being created and shared. I'm suggesting that "too much content" isn't a problem, and leads to good things.
kixiQu 4y ago
Did people feel that way about software?

jillesvangurp 4y ago

We're at the beginning of a rapidly evolving situation where out of the blue it is now possible to direct an AI to produce imagery that is interesting using just language. From there to moving animations is a relatively small step and already being worked on. From there to photo realistic scenes directed by just words not that big of a leap either and also something that is being worked on. Are there constraints, glitches, and limitations. Of course. Has that ever stopped artists producing interesting art? Of course not. People have been telling stories via art ever since people figured out that they could scratch things into rock, clay, and whatever. Perhaps the depressing thing for some artists is that they need to adapt to this and up their game. We've seen the same with photography. Every idiot can now have a decent phone camera and there are plenty of people on Instagram doing not so interesting things with those. Does that invalidate what really good photographers do? No. But it does make their work less special when an AI guided camera in the hands of an absolute amateur can produce photos that are decent. And now with this, you could produce realistic looking photos of things that don't exist by simply asking for them in the right way. More content means the quality standards are raised. There are not a lot of artists left that make a living scratching things in rocks. But there are more artists than ever. The only real limitation is their imagination.

ido 4y ago
I don’t want it “whatabout” this, but…Is art really the thing we need to automate right now? There’s so much labor that we more directly need more of and that we could all benefit a lot more from automating it. To give a mundane example off the top of my head- street cleaning. I live in a big city (Berlin) which is often filthy with litter. Obviously there are not enough people cleaning it (or it wouldn’t be). Can’t we automate that instead? How about automating construction? Part of the reason housing is expensive are lot prices but surely a lot of it is also labor cost. It just feels like such a frivolous and relatively useless thing to automate, and a misallocation of resources.
- luismmolina 4y ago
  It's not about what we need. Scientific research in this case has been the result of looking around and seen what can be done with the technology at hand without caring if it was what we need or don't. If you feel like "we" should be automating other stuff, you are free to make your own contribution, it's not like OpenAI owns the keys to the field.

riffraff 4y ago

> Lowering the effort bar for content creation and generation is, in some ways, going to cause a deluge of low-quality/mid-quality content to be generated. Isn't this the case already? For example, I have recently discovered science fiction poetry exists. The amount of stuff that gets produced in this niche of a niche is baffling already, and growing almost exponentially. I feel human curation and editing will just have to be important again.

hnick 4y ago

> I wonder if we'll see a new generation of "editors" showing up to help curate and provide some guidance in terms of what to play if you're limited on time. Iron Pineapple has an entertaining series he calls "Steam Dumpster Diving" which is similar to this, but it's very focused on Souls-Like games. https://www.youtube.com/watch?v=9KARY5ocvKo&list=PLuY9odN8x9pvIoTDtNOZxzo5m6EsBspAi https://www.youtube.com/watch?v=9KARY5ocvKo&list=PLuY9odN8x9... It does sometimes include games many of us have heard of but there are a lot of unknown games, and plain bad ones (he has a soft spot for student projects too). I think he sees the value of new ideas hidden in jank, as do I, so I like the series. I think the shift will be towards matching personalities, you get to know a reviewer and see what they like and how compatible they are with your own opinions. It doesn't have to be a one-to-one match as long as you're aware how they diverge on certain genres. The curator section on Steam seems to be an attempt at this, but any social media that allows following/subscription could serve too. It's still a significant time investment to keep up to date and to even find that matching personality. I wonder if I actually bothered leaving reviews on Steam, would it suggest people with similar tastes to me?

shreyshnaccount 4y ago

there's also an interesting take that as these generative models become better they will be used for search instead of regular search engines, and social apps would automatically generate content for you, trained on the vast dataset they already have. i think the hardest thing in this trend is going to be assessing 'truthfulness' automatically

simion314 4y ago

> Lowering the effort bar for content creation and generation is, in some ways, going to cause a deluge of low-quality/mid-quality content to be generated. This is not a problem, the issue is with Steam of whoever curates this content for us, with some good filters you can avoid the problem. My first program and game was garbage but it was not on GitHub or Steam to bother anyone, so the issue is not that people have access to good tools (I was using some Visual Studio for students edition) the issue might be that maybe some services have bad filters. For new indie devs, please use whatever tools make your vision , publish them, ignore the haters, you will probably not make a living from Patreon or donations but if you love what you doa nd makes you happy that is the important thing.

LoveMortuus 4y ago

I understand your fear, especially because a very similar, if not the same, thing happened with Unity, where our allowed the creation of just what you're talking about a massive amount of low to mid quality games. I even remember reading articles about how the Unity splash screen has become recognised as a mark for low quality games...

garblegarble 4y ago

I'm still hopeful that this technology will enable writers who have a good idea for a story to build much more on their own without having to be in a position to hire voice actors and artists: whether that's pixel graphic backgrounds for adventure games with decent text-to-speech for characters, or 3D worlds built in Unreal with Metahuman-generated NPCs playing out a curated story within a procedurally generated city. There being lots of low/mid quality games is, I think, a problem that already exists today on app stores -- on Steam especially between Greenlight & vast numbers of games 'releasing' in Early Access... and the more that a single writer with some dev skills can do on their own the more opportunity they have to develop their games as a brand, and get feedback to fire their creativity.

cmsefton 4y ago
Yes, I also like the prospect that this tool would allow rapid prototyping and storyboarding for games, graphic novels, films etc. Being able to generate a rough idea of what you want or envisage for a scene and show that to an artist/director of photography/whatever would be very useful.
j_mo 4y ago
> writers who have a good idea for a story Or programmers who are bad at art, so, programmers in general ;)
- thenickdude 4y ago
  I'm a programmer that's bad at art, and DALL-E is amazing for me. It generated me a logo I love for my open source app within a couple of batches. I would have gone logoless rather than slog through trying to find an artist on Fiverr.

dfxm12 4y ago

It feels as if this is about to explode an existing problem, namely, there's too much content in the world already. Lowering the effort bar for content creation and generation is, in some ways, going to cause a deluge of low-quality/mid-quality content to be generated. Like you say, this is an existing situation (I hesitate to call anything that lowers the barrier to entry a "problem"). The printing press opened up printing for more people. As did desktop publishing. Similarly, garage band and bandcamp make it so much easier to produce and publish music. Yeah, there's a ton of books and music out there, including a bunch of junk, but the cream still tends to rise to the top. I'd rather take my chances on having to find a masterpiece in a haystack than have it never get published at all, either because of lack of funding/lack of some specific skill that AI can handle/no willing publisher, etc.

MattRix 4y ago

It doesn’t matter how many bad Steam games there are, because their algorithms prevent you from ever seeing the vast majority of games on the platform. The same is true for videos on youtube and songs on spotify, you won’t find the really low effort/quality content unless you actively seek it out.

irrational 4y ago

I remember back when I was in college the WWW was barely released and search was in its infancy. If you were doing a research paper, trying to use the WWW to search for papers on the topic was laughably useless. Instead you would go to the university library (which still used card catalogs!) and make an appointment with these people who had access to some search system (I can’t remember what it was called). You would sit down with them and explain what you were researching and they would craft search queries in the query language this system used and read over the results with you to see if they sounded promising. They would tweak the query based on your feedback until you got the results you wanted. Then they would print out the results and you could go elsewhere in the library to lookup the papers. It cost money per search, but the university covered the cost. I wonder if we might see something similar with these ai generation tools. People who become experts at crafting the queries to get back the results you are looking for act as intermediaries.