... in a standardized test that has nothing do to with real enemy encounters - the complete title.
We've seen this before, even for a lot more limited and controlled environment like the game Dota the "AI" can beat the humans a few times before the humans learns to exploit it's many weaknesses.
Some program that self learned a (admittedly impressive) number of reactions based on seeing/playing a huge number of simulations is not intelligence so not AI. It cannot reason on the spot and will fall for the most ridiculous of traps. For example the AI that beat the professional Dota players fell for running in circles around a tower forever while getting slowly damaged to death. Even the most simple of mammals (which we do not consider intelligent) would react to the pain at some point and bolt.
My theory is that AI will not exist until we reach AGI. Because with specialized AI you can always fall outside it's area of "expertise" and behave like a stupid bot.
We've seen this before, even for a lot more limited and controlled environment like the game Dota the "AI" can beat the humans a few times before the humans learns to exploit it's many weaknesses.
Some program that self learned a (admittedly impressive) number of reactions based on seeing/playing a huge number of simulations is not intelligence so not AI. It cannot reason on the spot and will fall for the most ridiculous of traps. For example the AI that beat the professional Dota players fell for running in circles around a tower forever while getting slowly damaged to death. Even the most simple of mammals (which we do not consider intelligent) would react to the pain at some point and bolt.
My theory is that AI will not exist until we reach AGI. Because with specialized AI you can always fall outside it's area of "expertise" and behave like a stupid bot.