do AI tools understand such a license text and evaluate if they can or cannot use the material?
So, this is the fun part: AI tools don’t auto-ingest material to process it. The developers choose the materials to feed into the models.
And while the tech bros can understand your licenses, they don’t give a flying fuck, because they think they’ll be billionaires beyond consequences by the time anyone discovers that their work in particular has been ripped off.
Well the companies and developers don’t decide for every single material. In example what I expect is, that they program the scraper with rules to respect licenses of individual projects (such as on Github probably). And I assume those scraper tools are AI tools themselves, programmed with AI tool assist on top of it. There are multiple AI layers!
At this point, I don’t think that any developer knows exactly what the AI tools are fed with, if they use automatically scraped public sources from the internet.
So, this is the fun part: AI tools don’t auto-ingest material to process it. The developers choose the materials to feed into the models.
And while the tech bros can understand your licenses, they don’t give a flying fuck, because they think they’ll be billionaires beyond consequences by the time anyone discovers that their work in particular has been ripped off.
Well the companies and developers don’t decide for every single material. In example what I expect is, that they program the scraper with rules to respect licenses of individual projects (such as on Github probably). And I assume those scraper tools are AI tools themselves, programmed with AI tool assist on top of it. There are multiple AI layers!
At this point, I don’t think that any developer knows exactly what the AI tools are fed with, if they use automatically scraped public sources from the internet.