Of course we can predict what effect AI will have, it's super obvious.

First it reduces the quality of writing as a lot of people already mindlessly trust AI to write for them. This leads to people believing more falsehoods, becoming lazier and thinking for themselves less. Simultaneously it's used for social media manipulation with fake users and fake photographs. Again people believe more falsehoods and begin to doubt if they can know anything. Meanwhile it's also used for censorship. We are here. Then it becomes the dominant software creator and there are a lot more software bugs and hacks causing supply chain issues because people don't know how to fix the AI-written code and AI isn't logically smart enough yet. Then it's used in robots to take even more jobs, radically increasing unemployment, leading to more socialism and discussion of UBI. The value of education and skill depreciates as people see less reason to learn things, be creative or try to get a job. Then it's used in robots not only for taking jobs but also for police and private security.

Then AI gets better at scientific thinking and is used to devise new weapons, viruses, and chemicals. Leading up to this there are discussions about restricting access to AI or freedom to develop AI because of the dangers. This allows people to be more easily propagandized by a handful of corporate AIs. However, all the major AI discoveries are easy to access and there are plenty of people making open source models, so anyone with the motivation can get uncensored AI. Then someone uses AI with robots to make a planet-destroying nuke or a slow-killing virus more infectious than the flu, or something else, and the vast majority of people die, if not everyone.

But that's only if 90% of people don't turn against AI early on to stop its development. Or divine intervention. So what's it going to be?