Because when you're a medical professional you can lose your licence for experimenting on patients.

There's a big difference between "reading research papers" and "doing research." The former is expected. The latter generally is not. Most professional computer scientists do not write peer-reviewed papers in top academic journals during the course of normal professional development. (There's a big difference!)

Your code does not need to go through the FDA nor cost millions of dollars to run a valid trial.

He could work for NASA in which case his code would go through millions of dollars of analysis before it's run. ;)