Will future language models converge on "what Einstein would have thought of Many-Worlds?" before 2036?
Plus
17
Ṁ5352035
46%
chance
1D
1W
1M
ALL
Get
1,000
and1.00
Sort by:
The idea is: Imagine that best methods for integrating all text data were as well-understood as using odds ratios to aggregate bits of evidence. An LM properly trained just for predictive ability should do that in the limit.
Technically such convergent answers can remain uncertain, though this one shouldn't, for a reasonable context and manner in which he were asked.
Related questions
Related questions
By the end of 2026, will we have transparency into any useful internal pattern within a Large Language Model whose semantics would have been unfamiliar to AI and cognitive science in 2006?
38% chance
Will AIs be widely recognized as having developed a new, innovative, foundational mathematical theory before 2030?
32% chance
Conditional on no existential catastrophe, will there be a superintelligence by 2030?
38% chance
What opinion will future language models converge to, on what Einstein would have thought of Many-Worlds?
Will AIs be widely recognized as having developed a new, innovative, foundational mathematical theory before 2035?
43% chance
Will models be able to do the work of an AI researcher/engineer before 2027?
40% chance
Will a sentient AI system have existed before 2030? [Resolves to 2100 expert consensus]
38% chance
Will any 10 trillion+ parameter language model that follows instructions be released to the public before 2026?
51% chance
By 2030, will large language models still be at the peak of AI? [DRAFT]
30% chance
Conditional on no existential catastrophe, will there be a superintelligence by 2100?
75% chance