Discover more from Zhebrak et al.
AGI is not singular
There won't be a particular moment when we solve AGI once and for all. Instead, we will continue making our narrow AI wider, while machines conquer new environments. We will gradually start to question, "Aren't we there yet?". Subjective nature of perception of phenomena such as consciousness and emotions will emerge. Formal proofs and benchmarks can only get us so far, and slowly, one by one, we will approach the human limit of assessing solutions presented by agents. Once in a while, we will discover a new environment with questionable performance, yet with every new one, there will be less and less consensus among researchers. We won't be able to prove whether it is an approximation, a joke, or indeed the answer is "42". As the debate continues, we will advance into the future side by side.