The speaker deliberately pivots the conversation from abstract, long-term concepts like AGI and the singularity to a more grounded discussion about tangible AI capabilities. The focus is on what AI can do now and in the near future, rather than what it might become.
It is presented as a fact that AI has already surpassed human-level performance in numerous domains. This is not a future projection but a current state of affairs that serves as a baseline for future growth.
A strong forecast is made for an explosive growth in the number and diversity of superhuman skills that AI will develop over the next ten years. This suggests a significant acceleration of AI's impact across all industries.
Instead of measuring progress towards a singular, human-like AGI, the speaker proposes evaluating AI based on its acquisition of specific, superhuman skills. This shifts the benchmark from general intelligence to specialized, superior performance.
Keep pulling the thread on Jenson Huang.