I don’t particularly use generative IAs that create Alt Text from images, but I’ve seen a lot of people who do use them and I think that’s the only use of AI from which I can’t draw anything negative.
The fact is that there are many people who do not know how to make a good Alt Text, and therefore either do not put them or put one that is only the word “photo/image”, which is much worse.
I think if AI had started there, as a tool to help accessibility, it wouldn’t have the stigma it has today.
But of course, there are not enough blind people in the world from whom to get absurdly, ridiculously, vulgarly obscene amounts of profit…
If you’re only looking at the tools everybody can get a hold of, I agree. I think if you look a bit further, you will find medical diagnostics that can hopefully top human detection scores and that’s worth pursuing as well.
I don’t see any good reason why the general public needs to have access to most of the models today. Most people just play around with it - and I don’t see the value there. When we get the final tally, we will have made the climate crisis worse and caused droughts with all the thirsty data center consumption. All so Alexa can remember what you said two queries ago and you can animate your childhood teddy in the Ghibli style.