How about a nice game of Chess?
“We have it! Let’s use it!”
This morning, preferably.
In my observations, GPT utterly lacks the sensitivity to fine distinctions and subtleties that are needed to drive timely and critical decision making. Sor of like the wargames that Tom Schelling ran during the Berlin crisis which demonstrated that [almost] no-one was willing to make the decision to torch civilization just to save Berlin.
AI models like GPT don’t actually “think” or “decide” anything—they are merely advanced predictive engines that generate output based on the training data they’ve been fed with. The results often feel like a statistical slot machine, with countless layers of complexity foiling any attempts by researchers to determine what made the model arrive at a particular determination.
I have no doubt the “experts” will screw everything up, either on purpose or because of their arrogance.
Oh great are we now gonna use AI to decide when it’s time to let the nukes fly? What could go wrong?
AI has the brain of a democrat
“Would you like to play a game?”
This is BS. No one in their right mind would take a human finger off the button.
AI can certainly tell the odds…but attaching that to the system is not going to happen.
Now…once the button is pushed, I can see AI determine the targets.
AI is really nothing more than a very big, very fast database system that calculates probabilities based on past experience. It is not sentient. It’s not going to “take over the world.”
Unless AI is capable of forcing several humans to physically launch nukes I don’t see how that would be possible. From what I understand you can’t digitally launch nukes because the final launch sequence is analog.
Exactly what Skynet did…
From the AI perspective, this makes sense. There will be peace if there are no humans around.