Over the past few days I’ve been thinking about my future employment. Will robots soon make me redundant as well?
There are already systems which can write articles, reports, and digest information much faster than I. From a quantitative approach, an artificial intelligent (AI) stock picker also has me beat.
AlphaGo has already beaten the world’s best Go players. IBM’s Deep Blue has done the same on a chess board.
But Go and chess aren’t the equivalent of investing. In chess you can see your opponent’s position. The same is true in Go. You’re making calculated decisions on complete information.
You might also end up playing the man rather than the game. If I do this, then he can do that. So if he’ll do that I’m going to do this. But what if he’s expecting this and not that?
This isn’t how you invest.
You don’t buy CSL Ltd [ASX:CSL] because you believe 2018 will be the year all investors flock to biotech stocks. You also don’t buy CSL because the stock price is rising and you think more people will jump on.
Your decision to buy CSL, or any stock, should be because you can buy it for less than its worth, whilst also handicapping for any foreseeable risks.
It pays to sit and think
When it comes to risk, there are known knowns, known unknowns and unknown unknowns.
Thinking about the latter is a useless exercise. If an unknown unknown causes investment losses, then we call that bad luck.
The former two, however, should preoccupy your mind whenever considering an investment.
What are the known risks that could result in a permanent loss? What are the unforeseeable risks, which could also result in a permanent loss?
Just like a punter at the race track, you need to handicap for these events.
A stock might have a huge potential payoff, but the odds of success could be tiny. By considering the odds and potential payoff together, you’ll end up making the best possible decision.
Of course that doesn’t mean you always make money. But with the odds in your favour, you’ll have the best chance to succeed over time.
A machine might be able to learn how to handicap risk. Yet there are so many more subjective qualities when it comes to investing.
Take business valuation for example. I’ll bet if two investors try to come up with a value for CSL, those two answers won’t be the same.
There are so many assumptions and inputs baked into those valuations. Among them are interest rates, growth expectations, future capital needs and margin consistency.
I’ve probably made it sound harder than it actually is. Pausing to think about the qualitative factors of a business is relatively simple to us. For a machine, I’d assume it would be much harder.
When asked about AI picking robots, Warren Buffett said:
‘We are always trying to get something that is worth buying. People want to find some formula – I call this ‘physics envy’. People want a formula, but the world isn’t like physics outside of physics and false precision does nothing but get you into trouble.
‘You have to master the general ideas and work to improve your judgement. I do not think most people have a lot to gain from machine intelligence.
‘I don’t think that they bring much to the able in terms of capital allocation or investing. I may be missing something entirely.’
To which Buffett’s right-hand man Charlie Munger said:
‘I am not sure the AI will create an economic revolution. I can see that AI is working in the marketing arrangements of Facebook and Google, and so it’s working well. I don’t know what exactly it will be. I’ve done so well in life by using organized common sense that I never wanted to get into fields like AI. I can walk along shores picking up boulders of gold. Why go sift and pan in the mines?’
To repeat a line from Buffett, I may be missing something entirely.
Maybe AI stock pickers will put me out of the job. Maybe humans and their emotions have no place in the market. Instead it should be made up of machines, efficiently buying on complete information.
But before that happens, we’ll probably say goodbye to many more doctors.
How AI could save millions
‘A woman with late-stage breast cancer came to a city hospital, fluids already flooding her lungs. She saw two doctors and got a radiology scan. The hospital’s computers read her vital signs and estimated a 9.3 per cent chance she would die during her stay.
‘Then came Google’s turn. A new type of algorithm created by the company read up on the woman — 175,639 data points — and rendered its assessment of her death risk: 19.9 per cent. She passed away in a matter of days.
‘The harrowing account of the unidentified woman’s death was published by Google in May in research highlighting the health-care potential of neural networks, a form of artificial intelligence software that’s particularly good at using data to automatically learn and improve.
‘Google had created a tool that could forecast a host of patient outcomes, including how long people may stay in hospitals, their odds of re-admission and chances they will soon die.’
This is where AI can really do some good. Unlike the stock market, a diagnosis or medical course of action is dependent on past data. Just because a stock grew sales by 30% last year doesn’t mean they’ll see the same growth again this year.
‘What impressed medical experts most was Google’s ability to sift through data previously out of reach: notes buried in PDFs or scribbled on old charts.’
When it comes to medicine or medical procedures, proven means it works almost all of the time. Yet nothing is proven in the market. Management consultant McKinsey & Co. wrote:
‘The role of big data in medicine is one where we can build better health profiles and better predictive models around individual patients so that we can better diagnose and treat disease.
‘One of the main limitations with medicine today and in the pharmaceutical industry is our understanding of the biology of disease. Big data comes into play around aggregating more and more information around multiple scales for what constitutes a disease—from the DNA, proteins, and metabolites to cells, tissues, organs, organisms, and ecosystems. Those are the scales of the biology that we need to be modelling by integrating big data. If we do that, the models will evolve, the models will build, and they will be more predictive for given individuals.’
And with any luck, I’ll still be around to write about the big data medical revolution.
Editor, Money Morning