The CEO says AI is more likely to impact healthcare than any other area of life.
Taipei, Taiwan – An executive at semiconductor giant Nvidia said generative artificial intelligence (AI) is already powering a “revolution in healthcare” and is poised to transform everything from drug research to patient diagnosis and post-surgical care.
Kimberly Powell, Nvidia's vice president of healthcare, said Wednesday that while it's still “early days,” healthcare will be more impacted by AI than any other area of life.
“Healthcare will probably be the most impactful application of generative AI,” Powell said at Nvidia's AI summit on the sidelines of the Computex trade show in Taipei.
Powell said AI is already making progress in drug development and testing, but that under current timelines this could take up to 15 years and cost up to $2 billion.
“We value speed in our industry, and speed means doing more, and we recognize that drug discovery is essentially an infinite problem. We're looking at chemical space and 10 to the power of 60 potential compounds,” Powell said.
“This is essentially infinite computation. Probably the only way to intelligently search that space is with generative data.”
Powell said AI can be used for modeling to help researchers understand how new compounds interact with the human body, which could help reduce the typical 90 percent failure rate for most drugs in clinical trials.
“Generative AI not only enables us to generate more ideas and make more accurate predictions, but also enables us to model biology in new and exciting ways, thereby increasing the success rate of bringing new chemical entities into the clinic,” she said.
Nvidia, the world's third-largest company by stock price, has made healthcare a key focus of its business as it seeks to capitalize on many future applications of AI.
The California-based startup has developed a suite of platforms, software and medical devices to assist healthcare professionals in areas including digital imaging, diagnostic scanning and robotic-assisted surgery.
In March, the company announced it had signed agreements with Johnson & Johnson and GE Healthcare to use AI in surgery and medical imaging, respectively.
Powell said similar technology is already being used in self-driving cars to provide live information for real-world decision-making.
“If you think about ultrasound, robotic surgery or self-driving cars, they're not that different. There's a ton of sensor data coming in and decisions are made in real time,” she said.
Powell said generative AI will also be invaluable during post-operative and subsequent medical treatment phases, using patient data to compile post-treatment reports and reviewing previous surgeries to gauge their success.
“There are six to 12 people in the room operating and making decisions in real time, and just like many athletes do at the end of a match, the surgeon will look back on the operation and understand what could have been done better,” she said.
“It is not difficult to imagine how generative AI could become of major utility at every stage of surgery.”
Nvidia's graphics processing units have driven a boom in AI investment, turning a once-little-known startup into a company with a market capitalization of nearly $3 trillion in just a few years.
Generative AI went mainstream last year with the release of OpenAI's groundbreaking app ChatGPT, creating a flurry of excitement and concern about the technology's potential applications.
