“Research is to see what everybody else has seen, and to think what nobody else has thought.” 

Albert Szent-Gyorgyi, Hungarian researcher and Nobel biochemistry laureate

I am preparing a workshop for clinical researchers at the Pediatric Academic Societies meeting this week and I took the opportunity to think of major takeaways for the current researcher and everyone else as we navigate an exciting era in artificial intelligence (AI).

Intelligence is a behavior that includes: problem-solving, reasoning, planning, decision-making, making inferences, and learning. While artificial intelligence may be able to perform some of these tasks, it is not yet able to execute all of these elements that are relatively easy for humans (Moravec’s paradox). The paradox states that certain tasks are easy for humans (reasoning, perception, mobility, etc), but these are sometimes very difficult for artificial intelligence to execute.

AI is machines performing tasks that require intelligence. While machine learning (ML) is a subset of AI techniques that uses statistical methods that enable machines to improve with experience, deep learning (DL) is a subset of machine learning that uses neural networks, which mimic the human brain. Data science is a discipline that captures, maintains, processes, analyzes, and communicates data.

Technology like AI advances in a zig-zag fashion up an exponential curve, rather than a smooth curve as one may imagine. In addition, a technology may arrive too early or  unexpectedly, and society may have difficulty to adopt the technology. The ancient Greek Antikythera Mechanism, the first analog computer known in history, was an example of an advanced technology that was not widely adopted by its potential users due to its sophistication. 

For a technology to be adopted widely in a domain, the local domain experts may need some representatives to be its champions. One strategy is to have domain experts that become knowledgeable in both the domain knowledge as well as the new technology. These dually-educated champions are “bilingual” and can become the new technology’s trusted advocates, especially during the early adoption stage when trust and transparency are essential challenges.

While conventional statistics focus on models with instructions with the purpose of determining inferences by estimation, machine learning focuses on systems that learn from data with the aim for optimization (prediction accuracy) with perpetual improvement. In addition, machine learning usually is involved with high-dimensional data whereas statistics is usually applied to lower-dimensional data.

None of the aforementioned techniques would be possible without data, the collection of raw and unorganized numbers, facts, figures, etc. Once data are processed and organized as well as interpreted and contextualized, data become information that can then be used for decision-making. Knowledge is the deeper and sometimes tacit and personal understanding of the information based on learning or reasoning as well as personal values or experiences.

The importance of these observations of artificial intelligence in healthcare will be part of the topics of discussion at the in-person AIMed Global Summit 2023 scheduled for June 4-7th 2023 in San Diego. The remainder of the week will be other exciting AI in medicine events like the Stanford AIMI Symposium on June 8th. Book your place now!

We at AIMed believe in changing healthcare one connection at a time. If you are interested in discussing the contents of this article or connecting, please drop me a line – [email protected]