Wednesday, November 16, 2022

Theranos had a great idea

Theranos promoted the notion of a 'Lab On a Chip' for blood tests. It promised to quickly do multiple blood tests from a single drop of blood. It was better in every way, except for one hitch: It didn't work. Still, the idea itself is a great one. It is feasible to at least do blood testing better. According to the Washington Post, companies are still working on improvements and still raising funds: https://www.washingtonpost.com/technology/2021/11/16/blood-startups-theranos/

Here is something that might be a promising research protocol and ultimately an operational lab on a chip:

With a traditional testing lab, gain samples of blood classified as positive or negative based on traditional testing. 

Do spectral analysis of the samples and use that data for Machine Learning to train recognition of the condition based on spectral analysis. 

Create a 'lab on a chip' that can do a spectral analysis and match it to the condition. 



No comments:

AI Alignment and Security Now!

As far as I know, people charged with 'Alignment' at AI companies are not convinced things are safe enough and some have quit becaus...