Disclaimer: This post is written in a semi-humorous style. Some characterizations are intentionally exaggerated for comedic effect, but the underlying question is serious.

Recent blocking of claude fable just highlighted the problem we all moving closer to - should we build something that can blow up the world and who we should give the lever?

First, the answer - ofcourse we should, why not, since we can, it is already out, and the best quote I like: “it’s out of our control” (c) Zuckerberg.

The bird is out of the cage, no one can hold it in hands any more. Or actually could we? I’m pretty sure we can, and will explain below.

Second, who should be the chosen one? Here we have another problem, short map on candidates:

  1. Mark Zuckerberg - bit infant, every bad thing he did intentially later excused with spreading hands 🤷‍♂️ like it wasn’t him, and promising he won’t do that any more, but did. Like a kid. Wanna give The Button to kid?

  2. Sam Altman - commis voyageur, will sell your soul to devil, if needed. Should we give it to one without moral?

  3. Dario Amodei - cunning megaloman, have strong moral compas which always pointed on him. Think that AI so dangerous so no one should develop it, except him ofc. What can go wrong when person who wish the power actually comes to power?

  4. Elon Musk - addicted with all possible mental health issues. You don’t even need the proove. Not even an option.

  5. The Celestial Kingdom - Chinese dominance in any advanced tech never ends well to whole other world. If you didn’t hear about Han chauvinism, soon you’ll. So that’s long-term harmful option, like all previous. Who else was left behind?

  6. Maybe google’s duo? No, they will eat whole pie alone without inviting anyone to the party.

  7. Larry Ellison? Maybe, but he will just give the keys to the gov. Not the worst thing, nor the best.

  8. So the only left are Jeff Bezos and Jensen Huang. For both cases it will ends up like in Matrix story - with Jensen we all becomes a batteries for advanced AI, and with Jeff we all will be pedaling the wheel to produce energy for AI.

So, I don’t even know. Are there any solution for this? I think it is.

The solution:

We already have example of effective controlled development of technologies, at least it last 50 years - Taiwan. So technically we could do the same here - chose some piece of land who is equally far from every country (not by distance obviously, and only for democratic countries since we won’t trust others) and establish there some instituion with big and crazy data centers. It should cooperate with every democratic countries, give some usage limits to every institution/laboratory/company, and every such deal should be hard reviewed. For all the rest selling such technologies must be prohibited. Every data center GPU should be tracked, every power plant. Mass storage devices. All engineeres that could build such data centers and PhD who can train such AIs.

AI can be used as a mass weapon and we must count it as one.