20 years is way too far out to make any kind of educated guesses. We used to be able to predict a lot about the relative foundations of education and industry, but not anymore. We will likely have GPT 4 in a year or so. A GPT 5 within a 2 or 3 years after that. By the end of 5 years we may be just about to get GPT 6. There will be many emergent properties as we scale up the size of the datasets and the quantity and quality of the training. There will also be many competitors to GPT. That's just for language models. Education will be profoundly affected 5 years from now, let alone 20.

Then we also have image generation, like Dalle 2, Midjourney, and Stable Diffusion. Those will get much better in 5 years. Look at the progress in the last year.

​

We will also have text to video that is amazing and capable. Different teams have already shown peaks at what they are starting to do now.

​

We will have many specialized and purpose built AIs of immense capability and scale.

​

We will also have the march of hardware improvement in CPUS, GPUS, ASICS, and everything in between.

​

We will have massive improvements in the software engines for 3d worlds, like the next Unreal Engines or their competitors.

​

As we push into the ten year area, even that is incredibly beyond our ability to predict now. Super advanced AIs and robotics and brain computer interfaces along with augmented reality and virtual reality and gene editing and much more. The depth and complexity of those technologies will be so far advanced compared to today's tech.

​

There will be so many emergent technologies and unpredictable apps and combinations of technologies that it's really impossible to say what is that level of advancement will mean for us as a species. That's just 10 years. 20 years we may not even have anything resembling contemporary universities anymore, or they will be so advanced and different as to be unrecognizable.