Microsoft’s 3 billion bet on Copilot is assessed to be a disaster

[Mnementh]

SvennoJ said:

Mnementh said:

"Rationalizing" already implies it is not rational. You use that word multiple times. The definition you give is correct: "Rationality is the quality of basing beliefs and decisions on reason, logic, and empirical evidence rather than emotion or intuition." People claiming their decision being rational doesn't make it so and excluding factors that are known is indeed not rational. So yes, people lie a lot to define their decisions as rational, but I cannot see war being a rational decision, if you take everything into account. People use their emotions (fear, superiority) or personal gains to decide pro war, and then construct something to claim their decision is rational. It is not. In science starting from the result is not something you do. Rationalization is not the same as being rational. Because you already start with the conclusion and then cherry-pick your evidence to support it. So you are not arguing against rationality at all, just a facade over emotionally based decision.

Exactly. That's why AI is dangerous because it is used to rationalize atrocities.

Expecting AI to be actually rational while being trained on human history is a fallacy. AI is not rational, it's a collection of human mess. Humans (at least neuro typicals) use top down thinking, start with the conclusion and defend it at all cost, now using AI to rationalize and advance their top down decisions. 

AI is providing the facade, another layer to hide behind, deflect accountability. That's why it's so dangerous. AI only makes it easier to cherry pick, it's not going to make anyone more rational.

What I said in the initial post of me in this convo: "Don't say AI should steer us, but I don't fear it irrationally either. The rational stance is to fear humanity." So I don't say AI is perfect or rational. I just said I don't put too much trust in humanity either. And humanity is ultimately in control here, the AI is basically a digital slave. And yes, humanity makes not too seldom very bad decisions, based on emotions like fear, greed or hubris.

I didn't defend AI from you, I defended rationality.

Last edited by Mnementh - on 23 May 2026