Experimenting with a new regular thread. Wanna chat about something non-infosec amongst those of us who frequent /c/cybersecurity? Here’s your chance! (Keep things civil & respectful please).
Have a great weekend!
Been playing around with Image generation for a little while now - so much fun. I’d never want to monetize off my work, mainly because I know that a lot of the image training is stolen (specific verbiage ‘stolen’ is my opinion, not legal damnation)
With that being said, does anyone know of any specific projects/models which are solely trained on consensual training/sharing? I know if my own artistic abilities were better, I’d want to share with the community and would have interest in training my own models to generate new, creative stuff with my unique stroke added.
fart_pickle@lemmy.world 5 months ago
So, here’s my input. So-called AI is today’s bubble that will burst just like the NFT did. The vast majority of AI startups will go under, and the few that survive will be bought or destroyed by big corporations. What’s left will be extremely invasive in terms of privacy and will be a new source of data breaches. The fact that people responsible for AI startups have no idea what an AI/ML is, will the process even worse.
stevedidwhat_infosec@infosec.pub 5 months ago
Interesting take - you mention AI pretty broadly here, is there a specific sector/use-case you have in mind specifically? Or do you truly mean the technology itself?
fart_pickle@lemmy.world 5 months ago
I was talking about today’s AI - virtual assistants, bots and the good old regular complex algorithms rebranded as AI. The “real AI” will benefit from the buzz. AI/ML was around for few decades now but it lacked media attention. Since LLMs got into mainstream, I think the science will continue to grow but the “public” projects will mostly die when the bubble bursts.