Data or diatribes

I went this week to the launch of the Royal Society's report on machine learning. 
It is a very interesting look at all the issues around the subject.
One questioner asked when we would reach the stage when algorithms will be able to do anything a human can do. None of the panel could put a date on it.
But by one yardstick, we are already there. 
Computer algorithms learn from the data they find to produce their outcomes. Just like scientists. 
An awful lot of humans base their outcomes on what Boris Johnson or Donald Trump says. 
Which do you think is better?

Good health

I am learning about the Health Service. The organisation is chaotic. The people who work in it are brilliant.
The people who have seen me at Bermondsey Spa, my GP, at Guys and St Thomas' Hospitals and at University College London have all been very good at the medical stuff and very good with a patient.
The organisation behind them is less good. 
It is not a normal business. Unlimited demand for a free service which divides political opinion is not the ideal model for an efficient, cost-effective business.
But it must be possible to have one businesslike bunch dealing with the allocation of beds, the booking and managing of appointments, the peaks and troughs, leaving the brilliant medical bunch to do what they are good at.
And they must have the means to raise the money it takes to meet this unlimited demand. If the government won't pay the bill, the Health Service needs to be able to raise it some other way.
I would invest in NHS bonds, or NHS crowdfunding.