I understand the importance of alt text for images. But as I was drafting my image description just now a question occurred to me – in adding alt text to images am I also inadvertently making it easier for the AI bots to scrape my artwork?
@beach yes. IIRC that's why @harriorrihar doesn't do it. Ultimately it's a matter of personal preference on the balance between accessibility and this risk (for example, @davidrevoy provides extremely detailed ones instead), with a seasoning of visibility (images with alt text get boosted more and thus reach wider in the Fediverse, reaching more humans but also raising the chance it will be scraped).
I haven't uploaded new images for 2 years either... except for some ScavengersReign concepts, since they were nominated for the Annie, and I thought it would be good to show it...
@alien Glaze's goal is noble, and they have good press and communicate brilliantly about it. The problem is that there are no studies showing the effectiveness of the system against generative images AI. Worst of all, you can find AI generative enthusiasts on Reddit who prove how they can train on Glaze artworks without any problems. (Could be a communication war, here too).
The project is also closed source, only for Windows and Mac. So def' not an option for me.
@beach You are not at all naive. These things change so quickly, the AI companies try everything like in the Wild West. It's nearly impossible to know the actual situation.
I only know for sure: blind people and people with sight problems will appreciate your ALT texts! 😊
@NatureMC@beach@FediTips@feditips
Why shouldn't there be scraping bots in the Fediverse? It has an open structure that is especially well suited for scraping. And the robots.txt file is not mandatory, it's just that most search engines abide by it, but they can simply ignore it of they want.
@ditol Therefore I asked the specialists. The decentralized structure can be used very easily to block and isolate these bots by admins. We have instances which block many bots already. So it's possible to isolate them.
With public posts appearing in Google they can scrape there, of course.
@NatureMC Yes, but most posts are public and can be accessed via http(s), even if your instance blocks all bots. This is what I meant without making it clear enough.
Add comment