Erik Brynjolfsson

While I am optimistic about the advancements in AI, I believe it is unlikely that we will see a general AI system that surpasses human capabilities in every respect by 2028. The complexities of human intelligence and the ethical implications are substantial challenges that won't be overcome in such a short timeframe.

While expanding LLMs is a substantial step towards achieving AGI, it's crucial to complement this approach with advancements in data quality, algorithmic diversity, and human-centered AI design. By scaling LLMs and integrating them with other technologies, we can make meaningful progress toward General Intelligence.