At least some state actors will invest the very negligible money of getting to where gpt-4 is now. It does not need to be cost efficient to train or run.
It's total cost is not even near the scope of a space program or even a major military research project.
With 10-100 million dollars you can probably get most of the way there once it gets prioticed.
At least some state actors will invest the very negligible money of getting to where gpt-4 is now. It does not need to be cost efficient to train or run.
It's total cost is not even near the scope of a space program or even a major military research project.
With 10-100 million dollars you can probably get most of the way there once it gets prioticed.