Watching IBM’s artificial intelligence ‘Watson’ beat champion players in the US quiz show Jeopardy! is fascinating.
Watson’s speed, its cognitive powers, ability to learn and to get answers right before the human champions had even understood what was being asked is deeply impressive.
On the day of the test it had access to 200m pages of structured and unstructured content consuming four terabytes of disk storage including the full text of Wikipedia.
In training Watson IBM engineers played over 100 games against past winners in order to improve its chance of winning, significantly developing its artificial intelligence since Deep Blue beat grandmaster Garry Kasparov at chess.
Not only did Watson have to figure out the answers to questions but it actually had to figure out what the questions meant in the first place.
Watson sits at the top of the tree in terms of the application of massive computing power and machine learning. Machine learning is where computers have the ability to learn without having every single step being explicitly programmed for them.
Think about how Google finds information for you, Amazon suggests products you might want to buy or LinkedIn figures out who your connections might be.
These algorithms have learned how to do this through data rather than being pre-programmed.
It’s also why Google’s driverless cars had a million miles under their belt without any accidents caused by them (until recently) because they have learned how to tell the difference between a human and a tree and what to do about it.
Implications for financial planning
So as we apply digital to financial planning the right starting point is to ask what role do we expect digital to play and how far can it really go?
Digital financial planning solutions can be categorised into a hierarchy, from those that can be more simply automated, to those which are more complex:
The hierarchy of automation
Level 1: Needs analysis:
• Individual needs analysis or gap analysis (for each particular need, what is the financial gap that needs to be closed?)
• Cash-flow planning (what will my income and expenditure look like now and in the future under different scenarios, ideally post tax?)
• Balance sheet planning (what will my assets and liabilities look like now and in the future under different scenarios, ideally with non-financial products e.g. my home(s), cars and other assets taken into account?)
Level 2: Strategy and planning:
• Prioritisation across multiple needs (given my stage in life, dependents and situation what need areas are important to me in which order?).
• Household-based needs planning including partner and dependents.
• Current portfolio risk assessment (how is my portfolio currently allocated in which products, which asset classes and which costs and features?) What risk does this represent?