ML vs Wet Lab: The Great Impedance Mismatch
You’ve seen the articles about how Machine Learning and AI are revolutionizing drug discovery. You’ve listened to the podcasts. You’ve read the blog posts. It’s such an obvious match made in heaven, it’s surprising there aren’t more people doing it.
But the reality is, there’s still quite a lot of friction. Even though biology research has seen a huge surge in the amount of data that labs can produce, this data often isn’t quite the right fit for the tools of machine learning. Instead, we often need approaches that live somewhere between ML and more traditional scientific/mathematical models.
In my blog post this week, I explore how the nature of the data in each of these fields has pushed them towards different shared mental models for defining mathematical models, and how neither shared mental model quite fits the needs of today’s tech biotechs. Read it here: