Artificial General Intelligence (AGI) may be closer than expected, according to Sam Altman, CEO of OpenAI. In a recent interview with YC chief Gary Tan, Altman shared his insights on the accelerated progress made by OpenAI towards AGI.
Altman expressed confidence in achieving human-level reasoning and advancing to the next phase of AGI, “level 3.” He attributed this success to OpenAI’s strategic focus on scaling laws and deep conviction. The company has also faced criticism from experts like Gary Marcus, who claims that Microsoft is using a clause in their contract to potentially trigger an ‘AGI Achieved’ loophole.
However, Altman and OpenAI’s CFO Sarah Friar disagree with Marcus, stating that they have a clear plan in place to achieve AGI. Their progress has been backed by recent advancements, including the launch of their latest model, o1, which marks a significant leap towards AGI.
The competition for AGI is intensifying, with other major players like Google, Meta, and Microsoft also racing to ship AI advancements. Google’s rapid rollouts, such as Gemini 1.5 and Gemma 3, demonstrate the company’s focus on achieving this milestone. Meanwhile, NVIDIA’s Jensen Huang emphasizes the importance of scaling laws and massive computational advances in reaching AGI.
As the pursuit of AGI continues, it remains to be seen whether OpenAI will achieve its goal by 2025, as predicted by Altman.
Source: https://analyticsindiamag.com/ai-insights-analysis/agi-is-coming-in-2025