Unless you have been living under a rock for the last year or so, I am sure you have heard about the magical powers of Artificial Intelligence / Machine Learning. It’s the elixir that will solve everything from sorting our photos to curing cancer. And there is a good chance that AI actually will solve some of these.
But there have been some really important developments using algorithms for a while now and I found one out today morning. Ever since my commute to office increased by 5 min to 1 hour – I have gotten hooked on to podcasts. Today morning, I was listening to one of favorite ones – Freakonmics Radio by Steven Dubner, the author of Freakonomics. And I have to say that I was quite ashamed to find out about this algorithm so late (it was a re-broadcast) of something that has already happened in 2003-04. It was about a mathematical model which matches kidney transplant patients with kidney donors.
Contrary to common belief, your immediate family may not be a matching donor for you – there are a bunch of factors other than blood type that have to be considered. So this algorithm takes matches people who may want to donate a kidney for someone but may not be a match with some other people in same boat. So if we have say, 2 sets of donors and patients – X (donor) and Y (patient) and A (donor) and B (patient) and X is not a match for Y and A is not a match for B. The algorithm figures out that X is a match to B and A is a match for Y. Of course with 2 or 4 or or even 20 sets, we can do this on paper – but the algorithm does this for hundreds of thousands of people across the US – taking into consideration blood types, antigens/antibodies, and other factors that determine a match. By doing this the algorithm increases the odds of getting a kidney for patients living from dialysis to dialysis.
The person who drove building and implementing of this mathematical model is Alvin Roth and he actually won the Nobel prize (it saves lives!!!).