This is the Guardian.

What if there was a machine that could see disease years before any doctor.

We ran our model on that mammogram and they found like very early cancer.

Listen to Black Box, a new podcast series from The Guardian, seven stories about AI and us.

Coming soon. Sholte David didn't plan on becoming a scientific sleuth. I generally think of myself as something of a scientist,

although not a very good one.

I did my PhD and sell a molecular biology at Newcastle University.

But an obscure hobby has taken him down an unexpected rabbit hole.

As well as my full-time jobs in industry, I've been writing about errors in the scientific literature.

For years, Sholto has quietly... about errors in the scientific literature.

For years, Sholto has quietly and thanklessly

been flagging thousands of flawed research papers.

But recently, he found himself center stage

and the subject of profiles in nature, the Guardian, and even the New York Times.

I'm a bit uncomfortable about the attention because I don't want to limit what I say.

I don't want to stop sending rude emails. I don't want to stop writing rude blog posts.

While searching for mistakes in published papers

from his home in Wales,

Sholto discovered worrying errors in dozens of studies

from one of the world's most prestigious cancer centers,

the Harvard Affiliated Dana-Farber Cancer Institute in Boston.

Many were conducted by top executives at the Institute, so he wrote about it in a blog post.

...