RE: AI is becoming super dangerous?
You are viewing a single comment's thread:
You will never be able to trust a digital file or image ever again. Every frontier for AI is basically done, when it comes for visual media.
You can take an image, and convert it to a video. You can take a script of text, and convert it to a reading of the text, by giving it a voice sample.
Then you can take the image, the sound file, and combine them into a new video, complete with lip-sync, expressions, and changing intonations.
It is wild how quickly the technology has progressed. It will progress even faster as more and more gets thrown at it. However, unlike writing (which is easier to spot when it is AI generated) - more and more people are fooled by AI images because they do not have visual literacy to the same standards.
It isn't yet perfect, all this tech - and it never will be, but manufactured reality is quickly taking over the feeds. We will all need to be stronger sceptics to navigate a future that is superficial and for the most part, fake.
I'm really afraid that one day it will become perfect and we'll literally live in a Black Mirror episode. Even big companies are using AI to advertise... And really, every time I see a company using AI, it makes me distrust them. Although, as I said, there's no going back now. We still have to get used to this and learn to differentiate between reality and AI
I will be writing something in the coming little while (not sure, probably within the next month) about why I left my last job - in short, they were trying to convert everything and as much as possible to work with AI - without doing enough (in my opinion) due diligence as to the outputs and results.