AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Middle of the Road in Algorithmic Science
The one in the extreme is we need to advance algorithmic science. So everything is about making the model architecture better. But you know, there is obviously middle of the road between the kinds of things that I'm saying and what is happening in the world. We're training on data sets which are designed for observational research. And it doesn't reflect variations in code quality and variability across your practices or your providers in real life. In the real world risk of bias from inequitable service provision that means some patients won't even make it to the data. There isn't a central team where there isn't an expertise that's very easy to call upon in the clinical space.