The Verge: Users can prompt Bard to generate photos using Google’s Imagen 2 text-to-image model. Bard, now powered by Google’s Gemini Pro large language model, was always going to have image generation. It was assumed the more powerful Gemini Ultra model would power it; however, that model remains in development.
:format(webp)/cdn.vox-cdn.com/uploads/chorus_asset/file/24931893/google_bard_extensions.png)
3 comments:
Wow, the weekly AI article! Somebody’s got a copyright claim or nude photos made of them. The latter in this case. Artificial intelligence is progressing so fast, I find it really scary. It seems that for every big advancement, there’s another big scandal or leak or some kind of thing that isn’t technically illegal but definitely problematic. In my opinion, the government should for now regulate how AI is being used, at least until there can be laws and guidelines and requirements for AI services to follow in order to be safe. Right now, it seems too powerful and experimental to be safe, as is evidenced by all the shit that comes up with AI all the time. I’m also not thrilled that Google has an AI image service now – I’ve heard that google skims personal files for training material, and while it’s probably a conspiracy theory it wouldn’t be that surprising if it was revealed, would it?
Going to be so honest with this one - the use of “finally” in the title of this article is genuinely kind of scary to me. It’s now been a literal semester and a half of writing comments on weekly articles about AI and the amount of technological progression that has occurred in that amount of time is mind blowing. Especially right after the AI deep fake image generation scandals have been increasing in commonality and severity, celebrating another major chatbot developing these features seems at best hollow and at worst downright depressing. It kind of feels like they're saying, “Hey it’s great that a major competitor to the largest AI platform can now make up its own photos! Maybe it will be more ethical!” The speed at which videos, photos and text is generated is incredible and I know this technology will continue to progress at a dizzying pace to places I can’t imagine today. I understand and acknowledge the importance of AI but man it’s a scary part of our reality sometimes.
This may be a controversial take, but I think generative AI is inherently unethical. There's practically no way for it to become anything without standing on the backs of people who probably haven't consented to using their works in this manner, and it's nearly impossible to regulate the output of such a thing. We've seen this with things like celebrity deep fakes, plagiarized artwork, and AI is getting caught using material from owners that had never given permission to use it. And yet we keep making more AIs. yay. as if that's going to solve the problem. I understand that people in the technological sector just want to keep going, keep innovating, and keep improving, but I really wish they'd take a step back, and look at what they are doing, and truly think about what it means to all of us, and to our futures.
Post a Comment