DeepMind’s Latest A.I. Health Breakthrough Has Some Problems
The Google machine learning company trumpeted its success in predicting a deadly kidney condition, but its results raise questions around…
I’ve been following
’s health work since it launched. And if it’s taught me anything, it’s to say what you see, especially when no one else wants to. Some cautionary notes on the firm’s biggest healthcare breakthrough to date
Good analysis of the less publicized, but serious, shortcomings of Google DeepMind's latest high-profile publications:
"Beyond the headlines and the hope in the DeepMind papers, however, are three sober facts"
🔥 1: Nothing has been predicted- work is on historic data 2: Predictive pattern only works 55% of the time 3: Study population was 93.6% male
The always sharp
w/an excellent close read of DeepMind's latest health research. Behind the bold claims lurk major limitations, which deserve way more airtime. These include a reliance on data taken almost exclusively from male patients (94%!)
this part isn't highlighted as much as other issues here, but it's important : "the system generates two false positives for every accurate prediction"
"Streams, then, seems typical of DeepMind’s way of working. It offers few overall gains in clinical outcomes, creates anxiety and additional workload for physicians, and was built on the back of deeply controversial access to patients’ data."
Really interesting analysis of a recent DeepMind paper from
DeepMind’s Latest A.I. Health Breakthrough Has Some Problems The Google machine learning company trumpeted its success in predicting a deadly kidney condition, but its results raise questions around data rights and patient #diversity