JENSEN HUANG: "AI NEEDS TO ACCELERATE TO BE SAFE"
Nvidia CEO Jensen Huang says Al labs should put more compute into evaluation and alignment, not just making models more capable.
His argument is simple but important:
- More compute → stronger Al
- More compute for evaluation → better understanding of Al behavior
- Better evaluation + alignment → stronger
safety systems
The bigger shift could be that Al safety becomes a core part of scaling, rather than something added after a model is built.
As Al capabilities accelerate, the race may no longer be just about who builds the most powerful models.
It could also be about who can test, monitor and control them at scale.
More Al capability. More Al evaluation.
More Al safety.
#Al #Artificiallntelligence $BTC
$ETH #BTC Price Analysis#
Nvidia CEO Jensen Huang says Al labs should put more compute into evaluation and alignment, not just making models more capable.
His argument is simple but important:
- More compute → stronger Al
- More compute for evaluation → better understanding of Al behavior
- Better evaluation + alignment → stronger
safety systems
The bigger shift could be that Al safety becomes a core part of scaling, rather than something added after a model is built.
As Al capabilities accelerate, the race may no longer be just about who builds the most powerful models.
It could also be about who can test, monitor and control them at scale.
More Al capability. More Al evaluation.
More Al safety.
#Al #Artificiallntelligence $BTC
$ETH #BTC Price Analysis#
