- AI
- Molecular Imaging
- CT
- X-Ray
- Ultrasound
- MRI
- Facility Management
- Mammography
Is Elon Musk’s Foray into Radiology AI an Opportunity to Reexamine the Embarrassment of Learning?
Self-imposed fears of embarrassment may lead to an inadvertent defensive stance that prevents us from learning.
Elon Musk recently invited (some might say challenged) folks to submit radiological imaging studies for his AI, Grok, to interpret. He admitted it was “early stage, but ... already quite accurate.”
I have expressed confidence that AI is nowhere near ready to replace radiologists, and probably won’t be for decades. His announcement momentarily shook my confidence. Who could blame me? Have you noticed the stuff of which this guy is capable? If not, do yourself a favor and watch footage of how he recently got his rockets to land vertically.
Fortunately for my paychecks, if nothing else, the confidence rattling was short-lived. Folks came out of the woodwork to report that Grok wasn’t up to the task. One thread, in particular, was from a rad who found that the software couldn’t even tell it was looking at an image of a breast. After being corrected on that front, it failed to identify a big honkin’ malignancy within the breast.
Some might wonder why someone like Musk, with so much to lose, would stick his neck out in the public square and risk such embarrassment. A favorite podcaster of mine has addressed the issue, albeit not this particular instance of it: Immunity to embarrassment is kind of a superpower.
Most people go to considerable lengths to avoid embarrassment. It prevents a lot of risk-taking, which is an adaptive thing for social animals such as us. That can, however, result in lost opportunities where taking a chance could have led to great success, both for the individual and the society which might have benefited from his or her achievement. Some (probably most) really big breakthroughs have happened when someone like Musk stuck his neck way out and risked massive embarrassment, not to mention financial loss.
To make that happen, the individual has to be able and willing to ignore, or somehow embrace, the embarrassment that most of us shy away from. “What if I fail and prove my critics were right? What if I make a fool out of myself?” That’s an even bigger feat when you think about how much we knuckle under on a far smaller scale, both in terms of risk and reward.
I forget whether it was this line of thought or some other recent circumstances that had me dwelling on my medical education and postgrad training. One doesn’t get very far into the process before knowledge stops being spoon fed in lecture halls. Pretty soon, it’s expected that one has devoured textbooks and otherwise learned enough to withstand more of a Socratic method of teaching.
This results in endless opportunities for embarrassment as the student/intern/resident is challenged to showcase what he or she learned. It could be just the individual and his or her attending, the small team of other house staff, or an entire classroom. Failing to know what it seems you should have known is commonplace, but that doesn’t make it feel good, even if your attending doesn’t use it as an excuse to browbeat or otherwise haze.
A lot of this is self-imposed. Historically, one doesn’t even get into med school, let alone do well in it, if he or she isn’t capable and competitive. It’s very easy to become an unrealistic perfectionist, excessively punishing oneself for anything but a flawless performance. If one experiences “impostor syndrome” (covered in this blog 4.5 years ago), add the fear of being “found out” to perfectionism and the lower-grade avoidance of embarrassment folks generally endure.
It has been a couple of decades since I was in a residency reading room or standing on interminable “rounds” in hospital hallways, but I can still clearly recall the mental and verbal gymnastics I performed to avoid giving wrong answers. Even phrasing a correct statement in a way that might betray a gap in my knowledge could be bad news. An observant attending might know to probe at that weakness.
It might have been good practice for the “oral boards” we used to have in radiology. A bunch of you younger rads were spared the experience. (Congratulations on making it through this relatively brief window. The orals are coming back for future classes.) However, it was tried and true wisdom that anything you said could in front of your examiners could and would hurt you.
Meanwhile, when you’re supposed to be learning things — as a med student or house staff — this defensive behavior really isn’t to your long-term benefit. Sure, if you manage to keep up an illusion of being smarter than you actually are, you might jockey for better ranking in your class, score better positions down the line, etc., but that creates its own impostor syndrome.
On the other hand, if you can manage a micro-version of Musk’s superpower-ish robustness against embarrassment, it can be more adaptive. Being completely immune to it would be going too far, but there is a “sweet spot.”
Imagine, for instance, you are in your residency reading room and some point of radiological wisdom comes up. You don’t know it properly, and it stings. Nobody makes fun of you or accuses you of not keeping up with your studies — indeed, you might be the only one feeling any discomfort — but that little sting punctuates the factoid you had not mastered, and now it is cemented in your memory.
If you had just bluffed or otherwise finessed your way through, that knowledge wouldn’t be yours. Lather, rinse, and repeat that scenario over a few years of medical training. For the price of a few moments of ego bruising, you have racked up a lot more capability.
Better yet, carry that approach forward into your subsequent career as a practicing radiologist. Sure, you want your reads to be flawless for multiple reasons. When they aren’t flawless though or when you are staring at a case you could really use some help with, the threat of embarrassment can be your reminder. Here is a chance at self-improvement, one that can upgrade you and benefit many patients in your future.
Mammography Study Suggests DBT-Based AI May Help Reduce Disparities with Breast Cancer Screening
December 13th 2024New research suggests that AI-powered assessment of digital breast tomosynthesis (DBT) for short-term breast cancer risk may help address racial disparities with detection and shortcomings of traditional mammography in women with dense breasts.
The Reading Room Podcast: Emerging Trends in the Radiology Workforce
February 11th 2022Richard Duszak, MD, and Mina Makary, MD, discuss a number of issues, ranging from demographic trends and NPRPs to physician burnout and medical student recruitment, that figure to impact the radiology workforce now and in the near future.
Mammography News: FDA Grants Expanded 510(k) Clearance for AI-Powered SmartMammo Dx DBT Software
November 29th 2024Originally cleared by the FDA in 2021, the SmartMammo Dx software for digital breast tomosynthesis (DBT) can now be utilized with the Senographe Pristina mammography systems from GE HealthCare.
