13 Dec Cyborgs, Moons and Frogs: How AI will impact Industry
Introduction to AI
When many of us think of Artificial intelligence, we have images of cyborgs overtaking the world and enslaving mankind. Whilst there is always the possibility of this, we are many moons away from that realization. But AI has many useful and profound applications that are being implemented today and if deployed correctly, could help developing economies leapfrog up the economic ladder. But what is AI exactly. Let’s de-mystify the sector and provide some clarity.
Broadly speaking, there are 2 types of AI. ANI or Artificial Narrow Intelligence and AGI or Artificial General Intelligence. The latter is more akin to the way we picture robots and computers replacing and doing everything that a human can and the former, the more applicable, refers to areas such as farming, wealth management, self-driving cars and the IoT or the Internet of Things. Think of smart speakers for your home and fridge’s that learn your consumption trends and place grocery orders without needing a directive. Because of the growth of the former, we do see lots of scaremongering around AGI. But the reality is that we are nowhere near being able to master and build machines that can accomplish AGI.
Today I will give a brief outline of what AI is, the general terms such as machine learning, provide some key areas this may be applicable and then focus on one key area, wealth management where we have already seen rapid progress.
Artificial Narrow Intelligence Artificial General Intelligence
(Self-driving vehicles, Home Automation, IoT) (Capable of anything a human can do)
What is AI and General Applications
‘To the moon’ is often a term we often hear in investing. When people ask what the possibilities of AI are, I feel this terminology is fitting. Machine learning is one of the key concepts in AI and one most of us may have heard about and actually already benefitted from. This is where a computer is ‘supervised’ and given simple learning parameters and then allowed to process and adapt its own thinking. There are some great examples out there.
Think of self-driving cars. The computer is given parameters to use via its radar, which it uses to gather data on the position of other cars and the lanes. The machine learning then builds up a data pool of how a driver drives and anticipates how both the driver of the car would like to be driven and adapts to how the surrounding drivers are behaving. We are seeing this use cases in practice today. Just the other week I was in a friends Tesla and he was happily showing this work in practice and I must say, it worked exceptionally well. Will this be the future, will we no longer have drivers or taxis, will we simply have our cars pick us up and drive us home?
Other everyday examples can be seen in areas such as spam filters on our computers, audio translation, speech recognition (which most of us now use daily), as well as in areas like manufacturing where product inspection is now carried out by supervised machines rather than human inspectors.
Possibly the most lucrative and an area that I hear so many people commenting on, is online advertising. Especially where speech recognition is combined with this. As an example, how many of us have had a personal conversation on say the phone and said, ‘I want to go on holiday’ and then all of a sudden, we receive message, emails and target advertising all around holiday destinations? Because it is new, it somehow feels slightly creepy and as if ‘big brother’ is watching. But in fact, the benefits are that we get more targeted advertising around products we actually like and may consume and the companies that employ so many of us, well they get to reduce wastage in areas like advertising. Surely, in time and with our agreement, this could be something we grow to actually value?
So why has the sector picked up so rapidly now. Well this is for a number of reasons but mainly around technology. What we have today is advanced CPU’s and computer capabilities accompanied by the ability of collecting ‘big data’ (the more data points the more effective machine learning and AI). If we add the benefits of DLT/Blockchain and the advancements in areas such as smart contracts, we should see this advancement grow even faster over the next decade.
If the 2010’s was dedicated to the growth of blockchain and distributed ledger technology, the 2020’s will be the decade where AI capabilities expand exponentially. Think of your home being fully automated where your house knows when you are coming home and sets the lights and temperature before you arrive, where the kettle has been set to boil for when you wake up or the slow cooker or oven preheated according to when you prefer dinner daily. When you run out of milk, your fridge has already ordered a new supply this week but knows not to next week as you will be away on holiday.
AI in wealth management
One area that I would like to expand briefly on is AI in Wealth Management. The core concepts for discussion here are Robo-Advice, in the area of defining a customer’s investment strategy, target dates for retirement and required funds and then Asset-Allocation where portfolios are created to hit those defined objectives.
Robo-Advice as it sounds is a concept whereby a customer is asked a series of questions, the same questions they would be asked by a face to face ‘human adviser’. Those data points then feed into a financial plan. Complex areas such as tax can also be accounted for in this plan. The financial plan would incorporate time horizons, target required money for retirement (for example) and also account for risk profiles of the customer.
This all then leads to the next stage, which is portfolio construction. Concepts like target dated funds selection and ‘lifestyling’ where risk is reduced the closer a person gets to retirement can all be included. Portfolio construction concepts such as correlation, standard deviation and the efficient frontier will be incorporated in stock/fund selection with equal equally effectiveness.
It is also possible that some areas can be undertaken with greater efficiency such as Rebalancing. Robots and machines are less influenced by emotions and as we all know there are hyper emotions when it comes to investing. Removing this from the equitation is more likely to achieve optimal results. We have also seen the growth of Algo’s being used in trading floors across the world. In fact, many estimates suggested that around 90% of capital markets trading takes place through the use of ‘bots’.
AI in developing economies
For me, no conversation would be complete without a quick summary and focus on how AI may affect the developing world. As you know, I strongly believe that technology has the ability of leveling the playing field for developing economies. Will AI be able to further enhance this?
We all know the concept of the economic ladder and how developing countries climb this ladder. For example, through manufacturing. AI has the potential to knock out many of the lower rungs of the ladder, which means less jobs for the lower end of the scale and less chance for advancement. So, in that sense it is not a good thing for those economies in a developing state. Or is it?
Well, we have already seen how developing economies can leapfrog technologies in areas such as telecommunications, where there has been less of a need to have landlines as most people go straight to mobile and smart phones. This incumbent pull and requirement on the ladder can then be less restrictive. I believe that education may be the key and technologies ability to move education onto an online platform (assisted no doubt by Covid) will be one of the routes for success here.
Another poignant fact is that AI is in its infancy. There is absolutely NO reason why developing economies can’t be at the forefront of this advancement. If China can compete with the US in this area, and be more advanced than places like the UK, why can’t African countries also follow this example. It just requires governments to actually sit up and say, yes, we understand this and will commit to achieving this.
We at the Emerald Group have been advertising this idea to various governments across Africa. To build centers of excellence and technology hubs within their regions and take advantage of the ability to leapfrog up the ladder. Either they do that, or the ladder will become harder to climb.
The most efficient way of doing this is through a few core focuses.
- Invest in education.
- Focus on new verticals like AI, Blockchain and Tech. where they can compete as the sectors are new.
- Commit to moving the economy away from reliance on the old sectors such as industrial.
- Build strong public-private partnerships
- Ensure that this forms part of the government core mandate.
The simple fact is that not every part of China has developed as fast as others, as we have also seen in countries in Europe and states in America. And this will be true for Africa. It is those countries that grab the bull by the horns and actually and commit to advancement, that will adapt and assure not only their economic survival but affluence.
AI creates the best of both worlds
I think the best of both worlds approach is probably going to deliver optimal results. This is where AI, computers and robots are used to enhance service offering and efficiency but where we keep a human touch. There is a real chance for developing economies to leapfrog various economic rungs and I truly hope that governments see the potential and try and take advantage of this in the coming years.