Skip to content

Precaution, GMOs, Pvalue, Vlad Putin, Bill Easterly, John Grays, Bent Flyvbjerg, MedNexus, Turkey/s

Two young (motivated) friends have started a project to Uberize medicine and medical information, removing noise from medical search. I am running with them a (sort of) symposium here on the idea.

Please comment. Be tough, unrelentlessly inquisitive.
The project is very important *if* it allows the right transparency without interference from the big corporations. As you recall from the New York Times files, we have evidence that Monsanto discussed with a shill (Folta) the manipulation of entries on WebMD. Will access to multiple sources protect us?

So please ask them such questions as 1) how this system ensures independence, 2) whether they have the right filters, 3) what you like or don’t like about it, etc.

Anything that bypasses the middleman is good for us, if done properly.



Recall from Antifragile and earlier discussions here that a doctor’s answer would be different if you put (emotionally speaking) his skin in the game by asking him “what would you do?” instead of “what should I do?”

The opposite works equally well. A trick I did use as a trader: under pressure, to remove the emotional burden and the loss of mental clarity, you imagine that you are someone else in the situation. That someone else should be some precise person, in flesh and blood, say X. What should X do now? buy more? liquidate, etc. It applies to any decision, say “should X buy this house?”

You can use the strategy in a lot of dilemmas. Replace yourself with X, and ask: “should X resign because of ethics?”


Now our examination of GMO problem is taking us in an interesting territory. The initial work was essentially probabilistic — since few people understand both probability and fat tails, it was counterintuitive to many “scientists” (most scientists can’t even get P-values correctly to understand fat tails).

Luckily there is a huge crowd of computer scientists and mathematicians involved or familiar with the so-called P/NP problem and algorithmic complexity in general. They immediately get that:

1) selective breeding is different from insertion of remote genes from a combinatorial standpoint and how nature has to tinker in close, not remote space (genes from vicinity without going very far) to ensure stability

2) understanding the impact on a high dimensional environment of trangenics/GMOs is not possible

also not covered here, that genetics are neat to impress people with science but we will have a hard time understanding how things interact, and so long as P !=NP we will NEVER get things through genetics that we can check via experience and long term testing.

Post a Comment

Your email is never published nor shared. Required fields are marked *