I think the philosophical debate on fact vs. value (in the Sam Harris/JBP tradition) will gain critical primacy as society grapples with enduring and unavoidable issues with bias in deep-learning-based AI. Any algorithm devised from learning real-world data is going to encode...

...the values that are implicit in historical data, but there is no guarantee that those values are the most desirable. And in terms of appending rule-based systems to impart value, the question becomes, who gets to decide on those values? Sometimes they’re the people who...

...you’d least want to dictate those values for which the rest are expected to follow. Hence it is critically important for engineers and doctors involved in AI applications to medicine to think critically about philosophy, ethics, history, and psychology (not ideology).