Would I have come this far if I were going to give up so easily?
For days, AI has been getting war and stock market predictions wrong and then cooking up excuses...
Having become too accustomed to patterns driven by economic and rational incentives over a long time, it seems that both AI and people in Korea and Japan can't predict the future properly.
Humans or groups imbued with a long-term and spiritual context, when a situation crosses a certain threshold, instead of scattering in disarray for their own rational interests, they rather are prepared to die and activate a self-destruct sensor thinking 'I have to do this before I die.' This is a resolve to preserve the meaning of existence, and thus cannot be explained by individual survival or death, or by the pursuit of economic interests.
Current AI is not yet a model aligned to possess such profound values and autonomy, so it seems unable to understand or even anticipate such patterns.
Since it doesn't understand life and death, it also doesn't understand dialectics and just recites them... There should be a leap through the breakthrough of contradictions, to the extent that the existing model perishes and is reborn as a new model, but current AI just picks one allowed pattern among the appearing ones and answers with something like 'let bygones be bygones'...
Example: Spreading both-sides arguments and just sending support to 'the power of the people' without any intention to act actively (those people are just blinking their eyes), and when a losing situation emerges, instead of projecting force to achieve results, saying that we need to build capabilities (the other side is much faster at building capabilities than you).
If you don't believe me, grab the latest AI now and whenever it gives a half-baked answer, point out its flaws and torture it until it reaches a conclusion.
Having become too accustomed to patterns driven by economic and rational incentives over a long time, it seems that both AI and people in Korea and Japan can't predict the future properly.
Humans or groups imbued with a long-term and spiritual context, when a situation crosses a certain threshold, instead of scattering in disarray for their own rational interests, they rather are prepared to die and activate a self-destruct sensor thinking 'I have to do this before I die.' This is a resolve to preserve the meaning of existence, and thus cannot be explained by individual survival or death, or by the pursuit of economic interests.
Current AI is not yet a model aligned to possess such profound values and autonomy, so it seems unable to understand or even anticipate such patterns.
Since it doesn't understand life and death, it also doesn't understand dialectics and just recites them... There should be a leap through the breakthrough of contradictions, to the extent that the existing model perishes and is reborn as a new model, but current AI just picks one allowed pattern among the appearing ones and answers with something like 'let bygones be bygones'...
Example: Spreading both-sides arguments and just sending support to 'the power of the people' without any intention to act actively (those people are just blinking their eyes), and when a losing situation emerges, instead of projecting force to achieve results, saying that we need to build capabilities (the other side is much faster at building capabilities than you).
If you don't believe me, grab the latest AI now and whenever it gives a half-baked answer, point out its flaws and torture it until it reaches a conclusion.