Theft of what? There was no market for "memes in ghibli style commissioned by the original studio" which would probably cost hundreds if not thousands if hand-drawn. Nobody was going to pay for that. When it became freely available and instantly reproducible, that's the new market.
It's not truly free though, it's a loss leader for the near-trillion dollar AI industry. If we're asking where the stolen value ends up, I think you can answer "in the NVIDIA share price".
It doesn't seem hard to imagine 2-5 years from now when "memes in Ghibli style" turns into "pay us 25 cents and we'll send you a 30 minute cartoon in Ghibli style".
Unless you live in some anarcho-capitalist society, it is theft, in very simple terms. And I wonder, just where are all those highly successful libertarian societies? The ones who don't need to enforce copyright and where every member of society is creating his own creative art content, movies, songs, games etc. Oh, they have all failed miserably to scammers? Poor people, how I pity them (not).
I see the neo-Silicon Valley spirit of "regulatory arbitrage as a service" is unwavering.
It's promotion for OpenAI's product, without any of the appropriate licensing. 3D printing companies don't provide Lego schematics to sell their products. There's also the small matter of their ex-employee turned copyright whistleblower, who ended up dead:
Don't want it to sound harsh but he's 94 years old. Could it be he just doesn't want to go out with a loss and is fine with his final score not fluctuating too much? Even if the markets don't correct/crash like his reserves seems to indicate, nobody can blame him for eventually being too early or not losing money opposed to gaining even more.
Because this strategy seems to be in line with what we've been reading ("buffet amassing cash") for the last 5 years at least?
To me he’s doing the same thing he’s always done. I don’t think this is anything special. The market to gdp ratio is super high. It makes sense to be cautious here if you are managing hundreds of billions of dollars.
Doing it in pure SQL is really impressive but I think the real tell-tale sign of peak "cracked engineer energy" is the maintained, decade-old blogspot site. Can't exactly put my finger on it, but really gives off "niche mastery". I don't even know the authors but I'm sure in the right circles a few dudes maintaining a blogspot site called "database architects" for a decade probably don't need an introduction.
Though I like the novelty of AI generated content, it kind of sucks dead internet theory is becoming more and more prevalent. YouTube (and all of the web) is already being spammed with AI generated slop and "better" video/text/audio models only make this worse. At some point we will cross the threshold of "real" and "generated" content being posted on the web and there's no stopping that.
My hope was that AI would make it easier for people to create new things that haven't been done before, but my fear was that it would just be an endless slop machine. We're living in the endless slop machine timeline and even genuine attempts to make something artistic end up just coming off as more slop.
It might be true for "Creators" etc. but there were things that I always wanted paintings of but I have no talent, time, tools or anything really.
When I first got access to dalle (in '22) the first thing I tried was to get an impressionist style painting of the way I always imagined Bob Dylan's 'Mr. Tambourine Man' I regenerated it multiple times and I got something I was very happy with! I didn't put it on social media, didn't try to make money off it, it's for me .
If you enjoy "art" (nice pictures, paintings, videos now I guess) You can create it yourself! I think people are missing that aspect of it, use it to make yourself happy, make pictures you want to look at!
Even if it's made with AI, it is slop only if you don't add anything original in your prompt, and don't spend much time selecting.
The real competition of any new work is the backlog of decades of content that is instantly accessible. Of course it makes all content less valuable, you can always find something else. Hence the race for attention and the slop machine. It was actually invented by the ad driven revenue model.
We should not project on AI something invented elsewhere. Even if gen AI could make original interesting works, the social network feeds would prioritize slop back again. So the problem is the way we let them control our feeds.
> if you don't add anything original in your prompt
Define "original". You could generate a pregnant Spongebob Squarepants and that would be original, but it would still be noise that doesn't inherently expand the creative space.
> don't spend much time selecting
That's the unexpected issue with the proliferation of generative AI now being accessible to nontechnical people. Most are lazy and go with the first generation that matches the vibe, which is the main reason why we have slop.
The main comparator is Kasane Teto and Suno. Kasane Teto is functionally a piano that uses generative AI for vocal synthesis: https://youtu.be/s3VPKCC9LSs. This is an aid to the creative process. Suno lets you put in a description and completely bypass the creative process by instantly getting to the end: https://youtu.be/UpBVDSJorlU
Kokoro is art. Driveway is content. Art uses the medium and implementation to say something and convey messages. Content is what goes between the ads so the shareholders see a number increase.
I wish there were more things like Kokoro and less things like Driveway.
My hope is that it will be the death of the aggregators and there will be more value in high quality and authentic content. The past 10-15 years has rewarded people who appeal to the aggregation algorithms and get the most views. Hopefully going forward theres going to be more organic, word of mouth recommendations of high quality content.
I felt this same way as image generation was rapidly improving, but I've been caught by surprise and impressed with how resilient we have been in the face of it.
Turns out it's surprisingly, at least for me, to tune out the slop. Some platforms will fall victim to it (Google image search, for one), but new platforms will spring up to take their place.
Put more weight on your subscriptions. I don’t have much AI content in my YouTube suggestions. (Good luck AI generating an interview with Chris Lattner or Stephen Kotkin for example. It won’t work.)
yeah i already have so many AI-generated videos in my feed on all social media it's insane. i spot them from far for now but at some point i'll just be consuming content that took seconds to generate just to get money
Was a pump & dump by Casey Neistat. Lacked true popularity and network effects as it turned out people don't want to share unedited, raw footage. Social media is about looking good. So Casey just used his YouTube/influencer popularity at the time to pump metrics and then managed to sell it to CNN. No idea what CNN did with the tech or people but not much later they shut it down entirely.
>the lack of regulation in crypto has led to any tangible, groundbreaking innovation?
Is creating a ~$1T (and that's just BTC) asset class which went from obscure mailing lists to ETFs not innovation? Of this asset class, only 21M (divisible) units will ever exist[0] and to this date the original asset (again, BTC) has had no compromising (security) incidents deviating from its original mission (P2P ledger).
This doomsaying "doesn't deserve to survive" just seems mean-spirited without any actual arguments as to why it's not deserved for an asset class to exist which is truly deflationary (as in money supply) and shoving everything under "crypto" hoping for some outright ban because it's not regulated. It being unregulatable is a feature, not a bug.
[0] even with forks, the original whitepaper-protocol as we know it today will probably always be "BTC"
Huh? This is a peak performance in capitalism. Free market mechanics at its finest. This guy got roughly 2B$ to step down whilst still billing 50M$ annually as a "consultant".
To me, this seems like engagement clickbait targeting PG to promote an infotainment product (CTO coaching/course):
>there is no cookie banner law
There definitely is. The article explicitly states this:
>you need my consent when you want to track me
"tracking" here means storing data:
>store information in a visitor's browser is only allowed if the user is provided with "clear and comprehensive information", in accordance with the Data Protection Directive, about the purposes of the storage of, or access to, that information; and has given their consent (wikipedia)
The actual directive also explicitly states this
>consent may be given by any appropriate method enabling a freely given specific and informed indication of the user's wishes, including by ticking a box when visiting an Internet website (32002L0058.17)
So write unit tests automatically, change code later and then regenerate the unit tests? Now the code has a bug but the unit tests pass. I'm already seeing this today with devs using ChatGPT to quickly get the "test boilerplate" over and over.
You still need a human in the loop. I doubt Meta are letting it run blindly. More just automating part of the process and having a human decide what is and isn't committed.
Because OpenAI has first mover advantage, an actual product, the household "ChatGPT" name and starting anything from 0 would mean you start with a 3-0 disadvantage. Even if you threw a few billion dollars at it and attracted all of the top minds for at least a year - probably longer - you will be seen as the "ChatGPT alternative". How long before you can capture the momentum they have now with OpenAI? It's also a legal minefield, even when the majority of employees of OpenAI migrate, there will be all kinds of no-competes and conflicts of interest.
Strategically, this is probably a better move. Microsoft doesn't see their investment implode and they probably have some sort of plan to inject or absorb Sam and/or Microsoft back into OpenAI to prevent this in the future. Perhaps replacing the board of directors to prevent further infighting.
