i believe its patently bad when it comes to AI.
1) it could create an economic bubble if the only people interested in LLMs are LLM providers (unlikely scenario though)
the real issue is 2) deepfakes and the end of "authentic video". Let me explain.
Now, bearing in mind that most of us can still recognize a Veo 3 clip instantly, generative AI videos are getting better and better, and soon old people will be seeing clips on facebook of presidential candidates telling them not to vote, or that they're satan worshippers, or something, i don't know.
But here's the key - video AI takes a lot of resources. A text GPT can reasonably be run on a researcher's computer with a good GPU. Could the videos that look 90% real be done without billions of dollars of investment from Google and OpenAI? No.
When there are AI political ads on the TV and Google or Meta or some other AI company is decrying these or lobbying against them and saying "Veo is meant for good! We didn't intend for it to be used this way! Read the TOS!", remember they're the ones who enabled it.
TL;DR Google put billions into Video AI to create a product worth buying, that product is a threat to democracy and rational thought and probably wouldn't exist without their investment. QED, not inevitable.