Replies

I want a federal program to develop Warp drive so we can be on time for the Star Trek future history development of warp drive in 2062.

AI will be a million times more powerful in five years than it is now, so we should have the power tools to develop the math to develop the warp drive. Last year, AI DeepMind won a silver medal in a math Olympiad. Its speed of development suggests that this year, it will come out on top, so that vision is not too far-fetched.

Then it will likely take another couple of decades to develop the engineering to make the warp drive.

But things could happen faster because AI will be self-iterative, and it could iterate through engineering and math. Self-iterating AI starts about 2027. We have not seen what the world looks like when AI starts to self-iterate

Finally, AI will not stop being 1 million times more powerful than it is today in 5 years. It will keep getting more powerful.

Ray Kurtsweil recently said that we are currently in the outer bands of the singularity. I have seen this expressed differently as a development rate that creates a line roughly 45 degrees sloping upward. When self-iteration starts in two years, the 45-degree slope upward will move to 90 degrees straight up.

So it’s not preposterous to call for a federal program to develop a warp drive. It’s about like calling for a federal program to develop the atom bomb in 1940 or so.