Nvidia’s CEO sees next-gen AI with ‘reasoning’ ability, but it depends on computing cost – report

Nvidia’s CEO (NASDAQ:NVDA) Jensen Huang said that the future of AI would be services which can “reason” however the cost of computing needs to reduce to reach that phase, Bloomberg News reported.

Next-generation AI tools will be able to answer questions by going through hundreds or thousands of steps and reflecting on their own conclusions, Huang noted during a podcast hosted by Arm (ARM) CEO Rene Haas, according to the report.

This would enable the future software to reason, making it different from current systems such as OpenAI’s ChatGPT. Huang added that he uses ChatGPT every day.

Huang added that Nvidia will set the stage for these advancements by strengthening chip performance every year by two to three times, consuming the same level of energy and at the same cost. This would change how AI systems handle inference, the report noted.

Inference is the process of an AI model being in action, wherein it produces predictions or conclusions.

“We’re able to drive incredible cost reduction for intelligence,” Huang said. “We all realize the value of this. If we can drive down the cost tremendously, we could do things at inference time like reasoning.”

On Tuesday, Nvidia unveiled the next phase of AI at its AI Summit in Washington, D.C., with applications spanning across advanced AI agents, factories run by robots, weather prediction, curing cancer and contacting extraterrestrial lifeforms among other things.