20/11/17: Your Family Doc, Called AI...

In a recent post, I wrote about the AI breaching the key dimension of 'intelligence' - the ability to self-acquire information and self-replicate knowledge (see http://trueeconomics.blogspot.com/2017/10/221017-robot-builders-future-its-all.html).  And now, Chinese AI developers have created a robot that is capable of excelling at (not just passing) a medical certification exams: https://futurism.com/first-time-robot-passed-medical-licensing-exam/.

Years ago, working for IBM's think tank, IBV, I recall discussions about the future potential applications for Watson. Aside from the obvious analytics involved in finance (my area), we considered the most feasible application for AI and language-based software in... err... that's right: medicine. More precisely, as family doctors replacement. For now, Watson is toiling primarily in the family doctors' support function, but truth is, there is absolutely no reason why AI cannot currently replace 90% of the family doctors' practices.

And, while we are on the subject of AI, here is an interesting article on how China is beating the U.S. (and by extension the rest of the world) in the AI R&D game: https://futurism.com/china-could-soon-overtake-the-us-in-ai-development-former-google-ceo-says/ and https://futurism.com/china-has-overtaken-the-u-s-in-ai-research/.

Still scratching your heads, Stanford folks?.. 

17/10/17: Intel Opens the Era of Unemployed Insurance Brokers...

If you have a job structuring and selling, marketing and monitoring/managing car insurance contracts, you should stop reading this now... because, Intel has developed the first set of algorithmic standards for self-driving vehicles that aim to ensure that any accident involving a self-driving vehicle cannot be blamed on the software that operates that vehicle.

How? Read some scant details here: https://www.bloomberg.com/news/articles/2017-10-18/intel-proposes-system-to-make-self-driving-cars-blameless.

What does this mean? If successful, regulating algorithmic standards, most likely more advanced than the one developed by Intel, will mean that self-driving vehicles collision will be by system definition blamed only on human drivers, bicyclists and pedestrians. Which will, de facto, perfectly standardise all insurance contracts covering vehicles other than those operated by people. The result will be rapid collapse in demand for car insurance as we know it.

Instead of writing singular (albeit standardised) contracts to cover individual drivers (or vehicles driven by them), using actuarial risk models that attempt to identify risk profiles of these drivers, insurance industry will be simply writing a single contract to cover software running millions of vehicles, plus a standard contract to cover the vehicle (hardware). There will be no room left for profit margins or for service / contract differentiation or for pricing variation or for bundling of offers. In other words, there will be no need for all the numerous marketing, sales, investigative, enforcement, actuarial etc jobs currently populating the insurance industry. Car insurance sector will simply shrink to a duopoly (or proximate) providing cash management service to autonomous vehicles owners.

There will be lots of armchair-surfing for currently employable insurance industry specialists in the near future...