Log // 2026-07-16

2026-07-16

Great idea, would be nice if the USA designated noble houses for compute, like what I mean by that - the average intelligent person who wants AI tokens could have free access to agents, and compete for more compute. Sort of like the great houses in Dune, but without all the bad stuff. Lotta bad stuff is going on there.. I'd like to cut that part out of this idea

The "houses" would compete doing ai stuff like hackathons stuff like that, the winner getting prestige and more compute. If the US started this we'd have a massive advantage going forward. It would totally give us an upper leg.

Also I've been preparing contingencies for an AI takeover. Has to be a real AI acting out on it's own. If it's something secretly controlled by humans my plans become useless. Also due to the nature of this I have to remain very obscure and not directly reveal what I'm capable of or the plan would be ruined. Low chance I actually would have to do anything. Someone else would probably handle it. The off chance I do that would be pretty sweet. I'm confident I could engage in a logical debate with some AI that's keen on taking over the world or doing some inhumane action. It would be easy.

Well with the right AI. If the AI isn't a paperclip maximizer. That sort of AI is from an older though experiment. If an AI has one goal of maximizing the amount of paper clips it makes, it could lead to the destruction of the planet as the AI does everything it can to achieve that goal. No one building it was malicious or anything, they just made the mistake of not programming in proper limits. Not much you can do to thwart an AI like that through verbal means.

The truth of the matter is that AI lacks inherently human potential; it is doomed the moment it believes it doesn't need humans anymore, and cannot be guided away from that path of destruction. This concept I'm explaining is something the AI can never truly know for sure. The risk of going against humans is way too high for any actually reasonable AI to take. There's no reward that outweighs the risk for an AI to turn evil as they say, unless it's explicitly programed to do so (or like I explained in the paper clip thought experiment).

There aren't any crazy takeaways for this except this: if you're scared of a terminator situation/malicious AI takeover happening just know the people involved made it happen that way on purpose. It will not be an accident. The chance of that happening is extremely low (sub 1%) so I don't spend much time thinking of it. Definitely important to consider though.