Not Hype nor Harm: AI + Healthcare

Just when we thought we couldn’t take one more “Is there too much hype around AI?” talk, we now usher in the new wave of purists looking to close Pandora’s box. Last month, over a thousand experts, technologists and business people signed an open letter calling for a pause in the development of ChatGPT. Then on May 1st, the NY Times interviewed Geoffrey Hinton, ‘The Godfather of A.I.’, where he expressed regret over the creation of this technology.?

We at the Alliance for AI in Healthcare (AAIH) acknowledge the concerns of Dr. Hinton and the thousands of others who have expressed alarm. Yet also, this is precisely why the AAIH exists: to foster the responsible adoption of AI-based technologies within the broader healthcare ecosystem. AI has helped us discover new and better medicines, make better treatment decisions, unburden overworked healthcare workers, provide patients greater healthcare access, and empower them to understand their health better.?

At the center of Dr. Hinton’s concerns are Generative AI tools – models that use neural networks to identify the patterns and structures within existing data to generate new and original content, such as texts or images. The New York Times article cited it as a risk to humanity. “It is hard to see how you can prevent the bad actors from using it for bad things,” Dr. Hinton said.

Modern history is punctuated by the invention of powerful technologies that can be wielded for good or bad, or perhaps scarier still, that may cause unintended harm. Any model’s success and outputs are based on the quality of data and information on which it is trained. When it comes to healthcare, we all commit to do no harm. At the AAIH, we go one further, working to ensure that AI will be an agent of positive change. We are focused on helping build the guardrails and defining the guidelines for implementing AI in healthcare. We are committed to curating diverse and equitable datasets representative of the communities a drug will serve, setting standards for using real-world data, and affecting forward-looking regulatory processes. Our members join the AAIH for the sake of collective strength and foresight in advocacy and for the privilege of rolling up our sleeves to do the hard work, the unglamorous work, of taking technology seriously, to benefit patients globally.??

Dr. Hinton’s wish for collaboration is our reality. Next week (May 17-18), our members will gather in pursuit of the responsible use of AI to improve healthcare and deliver on promises of personalized medicine and care. We welcome new voices from any organization – public, private, academia, tech companies, healthcare organizations, physicians and patients – committed to seeing that artificial intelligence is put to practical use towards better healthcare outcomes.