A fascinating account of how AI is going to shape the future of healthcare. Also does a great job of explaining how AI really works
Rating: 4/5
Deep learning will involve pattern recognition and machine learning for diagnosis but also other applications like at home monitoring
Deep empathy and connection b/w patients and clinicians
Superficial contact w/ doctors and patients promotes incorrect diagnosis and the need for unnecessary tests/treatments
Screening programs over diagnose breast, prostate and thyroid cancer with increased incidence but no change in mortality
Physicians don’t know who will respond to treatment
System 1 thinking is automatic, intuitive, effortless. System 2 is slow, reflective, analytical
Don’t filter the data too early
Developing AI is all about inputs and outputs
New Apple watch 4 is not what it seems to be for ECGs
Computers today can perform specific tasks very well, but when it comes to general tasks, AI can’t compete with a human child
Machines lack of understanding cannot be emphasized enough
AI papers on medicine haven’t been published in top peer reviewed scientific journals and still need to undergo the expert review process (no prospective analysis as well, all retrospective)
Any mismatch in AI, medicine can lead to hundreds of people dying
Blackbox AI is still yet to be understood
Bias is built into every AI algorithm
21st century medicine is increasingly aiming to to upgrade the healthy rather than help the poor
No tech company is tracking its own adherence to ethical guidelines
AI can now predict the timing of a patient’s death with pretty good accuracy
An algorithm could better map more genes related to autism
***
Buy the book here
other links
contact
© Anish Kaushal 2025. All rights reserved.
