Artificial intelligence is going to become an omnipotent part of our world and everyday life. Most experts believe that the global market for A.I. will be worth $3.9 trillion by 2022 and as much as $15.7 trillion by 2030. But I sincerely believe that these predictions wholly underestimate the true economic potential of A.I. by a factor of 10. The potential applications of the global A.I. economy are at least three times greater than those of the internet economy, and today I would estimate that internet economy to be worth $50 trillion.
To gain long-term exposure to the entirety of this growth requires some lateral thinking. A.I. is still some years away from being the ‘platform’ that the internet represents. The true A.I. revolution will come once that companies and entrepreneurs can ‘plug in’ to a user-friendly interface that enables them to easily derive the full benefits of A.I.. In early 2019 IBM made Watson open source in the cloud, which history will likely view as a significant milestone, but there are still many components of A.I. that are yet to be mastered.
For artificial intelligence to work, it must first be able to capture huge amounts of data from the real world, often in real time. According to the National Science Foundation, the Internet of Things (IoT) is on track to connect 50 billion “smart” things in 2020 and 1 trillion sensors soon after. 5G technologies mean that captured information can now be processed in real time for A.I. to make a decision before taking an action (often a physical action). For this, the world is also going to need a huge increase in the abundance of smart robotics.
Focusing on these ‘component parts’ of A.I. is how investors can find good long-term exposure. Blue Prism Group provides exposure to the growth in robotic process automation software for enterprises. And there are ETFs such as the WisdomTree Artificial Intelligence UCITS ETF that promise direct exposure to A.I. enablers. But to create long-term exposure to the entirety of globally omnipotent A.I., one must look to invest not in just in the component parts of A.I., but in the electronic components that these parts will rely on.
Unfortunately for UK-focused investors, opportunities are very limited. According to the McKinsey Global Institute, A.I. is expected to provide a 22% boost to the British economy by 2029. But one would not know that if you looked at the UK stock markets… To find stocks whose prospects will correlate with the entirety of A.I., one has to turn to the US markets.
Micron Technologies (NASDAQ: MU) is a producer of computer memory and computer data storage. I have held a long position in this stock for quite some time, and I expect to benefit from long-term equity growth. Equity growth is the only play available because, like many in the semiconductor industry, Micron has never paid a dividend.
Over 12 months Micron has had a wide 52-week range for a company with a market cap of $57 billion, but this ‘volatility’ is easily attributed to wider supply and demand factors. Revenue grew 50% in 2017 to $30.39 billion but then dropped to $23.41 billion in 2019 on the back of slowing demand caused by oversupply in the whole semiconductor industry. Future trade wars between the US and China also risk exacerbating this trend.
However, I take a very long-term view on Micron Technologies. It has the scale, capacity and technical expertise to capitalise on huge demand from A.I.. The recent glut in supply should also have cleared out the worse inefficiencies from the sector. And I predict that as the global A.I. economy grows exponentially, memory chips are going to be in huge demand in perpetuity. Those one trillion sensors are going to produce an unprecedented amount of data, and if A.I. is going to use that data, it will need to be stored. Equally a 5G-enabled IoT will need to process a lot of data, very quickly and in real time, and in terms of technical sophistication and capacity, only NVIDIA chips compete with Micron.
My feeling is that analysts are greatly underestimating A.I. growth, and that a huge future A.I. demand for highly sophisticated memory chips is not yet priced into Micron Technologies.
