Antimicrobial resistance (AMR) is one of those nasty looming threats on the horizon that it's difficult to even think about, especially given how many other problems we have as a species. In the most extreme scenarios, we end up back where we were in the 1800s, where a simple bacterial infection could kill you.

Over the last few weeks, I've been helping to edit a report that looks at how we could use artificial intelligence (AI) to tackle AMR. The report is a joint effort by Google DeepMind and the Fleming Initiative and is out today. It's free to read:

Harnessing Artificial Intelligence to Tackle Antimicrobial Resistance

There's also a press release on the Fleming Initiative website, which summarises the key points.

The report starts with the need to figure out which aspects of AMR are most "learnable" by AIs. From there, it proceeds through the need for much more data and for that data to be widely shared. I also want to highlight the importance of equity: the AIs need to be accessible to low- and middle-income countries, as they're the ones at most risk from AMR.

I don't know how much of this will come to pass, but I hope it does: it's in problems like AMR, where there's a huge amount of complex data to sift through, that AI can be genuinely useful.