Came here from this thread about whether ML experts can readily provide examples of barriers between today and AGI. What a blast from the past. Seems like most of the alleged barriers that ML experts cited in this 2020 survey have already been smashed through in 2023. (I’m looking at Table 1, which contains things like “Insufficient ability to construct and decompose sentences according to grammatical rules.”)
Came here from this thread about whether ML experts can readily provide examples of barriers between today and AGI. What a blast from the past. Seems like most of the alleged barriers that ML experts cited in this 2020 survey have already been smashed through in 2023. (I’m looking at Table 1, which contains things like “Insufficient ability to construct and decompose sentences according to grammatical rules.”)