In our most recent edition of our BrightInsight Digital Health C-Suite Series, we spoke with Amar Goel, founder and CEO of Bito, about GenAI’s potential in healthcare and other regulated industries that life sciences can learn from. Here’s a recap of our discussion. Catch our 10-minute interview here, along with other interviews in the 2024 JPM Digital Health C-suite series here.
Amar brings a unique perspective to the discussion of GenAI in Healthcare, since Bito works with a range of Fortune 100 companies in many different industries, helping software developers by providing what they think are the "next generation of tooling and capabilities to use GenAI to help developers become dramatically more productive and creative and efficient."
From Amar’s perspective, "what we think about as AI today is certain kinds of models. They're typically deep learning neural network-based models. The term comes from trying to simulate the neurons in your brain. So you think about those electric signals go from one node to another node to another node.”
He then explains that GenAI, or generative AI, are models that predict or "output words, text, images, audio…they generate content. They're a prediction model. So you give it a phrase or you give it a hundred pages and it can then out predict the next word and then the word after that."
A great example is the potential of GenAI in education. For students studying online with a virtual tutor, each lesson can be personalized to the way that student learns. As a student solves a problem, the GenAI tutor can customize the next problem or provide clues to help the student solve the problem. It can also customize how that information is presented, so some students may see a video while another may read text or listen to an audio file.
We had some fun with this question, as Amar asked ChatGPT which industries have the potential to be revolutionized by GenAI and which can serve as models for healthcare and life sciences companies.
According to ChatGPT, "number one, the automotive industry, the use of generative AI and automotive design and manufacturing can be a model for healthcare. AI driven simulations and generative design algorithms allow for the creation of more efficient and safer vehicles. Similarly, healthcare can use these methods for drug design, creating medical equipment, or optimizing hospital layouts for better patient care." There were several industries identified including finance and banking. The financial industry uses AI for fraud detection, risk assessment, and personalized customer services. In healthcare, similar AI algorithms can be applied for fraud detection and billing, risk assessment in patient care, like predicting disease outbreaks or patient deterioration and personalizing patient care plans.
Amar makes a great analogy to the adoption of GenAI to the adoption of the microprocessor. As he notes, while the transistor was invented in 1947, it wasn’t until 11 years later in 1958, when Jack Kilby invented the integrated circuit that the full potential and impact of the technology was realized. "Those two inventions really power all of our computing, all of the technology that we have today."
"And so if you kind of look at GenAI today…I think you have to look at the journey that we're on." The technology will continue to improve as we use it more. "I think in four years you'll have models that are 10 or 15 times more powerful than they are today. So maybe that's GPT-6 or something will be 10 or 15 times more powerful. And if you go out maybe 10 years, you might be looking at models that are maybe like 500 times more powerful than they are today. It's pretty exponential, the growth."
"I wouldn't just be thinking about where we are today from a science or health care perspective, but where are we going to be in 3 years, 5 years, 10 years? Because I think that is where organizations can really benefit in a significant way."