What will be the Next Big Thing?

 Right now, generative AI is the “Big Thing” in global equity markets. Equity markets also remain excited about GLP-1 weight-loss drugs. What may be next?

 (Image generated using Google Labs)

Over the past 12 months, we have witnessed an incredible increase in demand for the infrastructure needed to enable generative AI, including semiconductors, servers, and data centres built specifically for training generative AI. Over the same period, share prices of companies that supply the semiconductors and other infrastructure enabling generative AI have risen dramatically. The massive increase in Nvidia’s share price has been well reported, but not everyone realises that the share prices of Dell Technologies, Super Micro Computer, and Vertiv have been even stronger over the past year. Vertiv provides power solutions for data centres, while Dell and Super Micro Computer build AI servers.

The market’s excitement about the rapid demand growth for generative AI has plenty of precedents. For example:

• The share prices of Novo Nordisk and Eli Lilly, the two leading producers of GLP-1 based weight loss drugs have both roughly quadrupled over the past 3 years in response to the rapid growth in demand for these drugs.
• The share price of Tesla rose 12-fold over 2020 and 2021, as markets got excited about growth in demand for Electric Vehicles (and Teslas in particular). At about the same time that Tesla’s market cap was hitting US$1 trillion, we saw startup electric vehicle companies Rivian and Lucid getting listed on the share market and immediately trading at eye-watering valuations (at about the same time that they began to sell their first vehicles).
• The prices of many companies enabling investment in solar power (such as Enphase, SolarEdge, and SunRun) rose more than five-fold over 2019 and 2020, as US investment in solar panels accelerated.
• Excitement about rapidly growing demand for services delivered over the public cloud has also contributed to some massive increases in market capitalisation in recent years, although the share prices of the biggest providers of cloud services (Amazon, Microsoft, Alphabet) did not generate such eye-watering returns, as these were already large companies, generating most of their revenues from areas other than the public cloud.

Looking even further back in history, we see many similar episodes of investor excitement about rapid growth in demand, featuring then-newish technologies that were experiencing a rapid increase in demand, such as smart phones, internet infrastructure, internet retailing, cell phones, micro-computers, and so-on. You can find this pattern throughout history, right back to the early days of automobiles and railroads.

In some cases, history shows that investors were right to get excited, but in other cases their optimism was excessive. Even in those cases when demand for the technology continued to grow, competition has sometimes meant that market leadership has changed or that profit margins narrowed, such that buying and holding the leading companies behind a technological revolution has not always worked out well for investors.

So, buying into a “Big Thing” at close to peak market excitement carries a fair degree of risk. However, hindsight shows that identifying these trends early on would have generated fantastic investment returns for the first couple of years of excitement about the new technology.

 What comes next?

In the rest of this article, I will try to identify some emerging or potentially disruptive technologies that could possibly generate similar excitement in the future. To my mind, this is no easy task, and our approach at Te Ahumairangi places more emphasis on identifying more predictable sources of cashflow, rather than trying to predict what will be the next hot thing.

However, given the potential for amazing medium-term returns from catching a wave of investor excitement, I think it is worth at least thinking about some technologies that have the potential to generate similar excitement in the future.

It is worth noting that most of the previous “Next Big Things” did not come completely out of the blue. For example, people have been playing around with AI for several years, without generating much excitement. (For example, in 2015 my brother used a recurrent neural network to win an international competition looking for software to identify authors of anonymous text. His neural network used some of the same techniques now used in large language models like Chat-GPT, and beat expert systems built by large multi-disciplinary teams of academics, but AI researchers continued to focus on building expert systems for a few years after that). Similarly, the first indications that GLP-1 drugs might help with weight loss were reported on over a decade ago, but investors did not get excited, as many other drugs have also shown the potential to help people lose weight.

These precedents show that picking the next Next Big Thing does not mean that we need a crystal ball to tell us what completely new technologies may be developed in the future. Many future investment fads will likely come from technologies that are being developed now, but may still be a few years away from being good enough to stimulate mass up-take. In many cases, we may be able to observe a trend of technology improving, such as Moore’s law of semiconductor processing power doubling every 18 months, and extrapolate this trend to work out when a technology may be ready for mass up-take.

Here are a few ideas as to what disruptive technologies may generate excitement in the future:

 Quantum Computing?

Classical computing essentially works by calculating one answer at a time. Given how fast computers can compute, and the option of parallel processing, this works fine for most problems. But it can fail if there is a sufficiently large number of possible solutions to explore. For example, if a courier has several parcels to pick up and deliver through town and wants a conventional computer to tell them the shortest route, conventional computing will likely strain to solve this problem.

By utilising the wave/particle duality of electrons, Quantum computing can in theory solve these sorts of “optimisation” problems by considering all the possible solutions simultaneously.  Researchers have already built experimental quantum computers, but at this stage they have tiny computing power and are error prone. However, the technology is improving, and Google claims to have achieved “quantum supremacy” by using a quantum computer to solve problems that cannot feasibly be solved with conventional computers.

Quantum computing should also be good in theory at factorising large numbers, and therefore has the potential to be used to break the public key encryption which is used protect passwords being sent over the internet. There are fears that the Chinese hack of the emails of several US public officials last year may indicate that the Chinese may already have a quantum super-computer that is capable of decrypting public key encryption (as no one could work out any other way that the hackers could have accessed so many passwords protected by public key encryption).

It seems possible that within the next few years, quantum computing could develop to the point that it has significant real world applications. If corporate customers see potential in quantum computing similar to the potential that they’re currently expecting from generative AI, this could lead to another spending boom.

 Improved forms of energy storage?

The world is trying to wean itself off fossil fuels by using renewable energy. However, the big problem with this is that renewable energy comes in the form of electricity, and electricity is difficult to store, as batteries are expensive, heavy, and tend to have finite lives.

Many alternatives to batteries, such as using electricity to produce hydrogen, are very inefficient. Using current technology, there is typically an energy loss of more than 50% when electricity is used to produce hydrogen, and a further energy loss of almost 50% when the hydrogen is converted back into energy using a fuel cell.

The problem of storing electricity will increasingly be a disincentive for large scale investment in solar and wind. This is because the output of solar and wind generation fluctuates over time (due to changes in wind and sunshine), such that in an electricity market dominated by either wind or solar (but with limited storage capacity), the electricity price would frequently fall to zero when the wind is blowing or the sun is shining. (The electricity price may also spike up towards infinity when the wind and solar output is low, but this would be little solace for the owners of the wind and solar generation, as they would have no electricity to sell while the price is high). For solar or wind to become a large proportion of a country’s electricity supply (without government subsidies) there would need to be a lot of available storage capacity that would suck up surplus electricity whenever electricity prices get unusually low.

The problem of storing electricity is also a big problem for long-haul air travel, where weight really matters. Using current battery technology, it would be pretty much impossible to build a commercial plane that could travel 10,000 km using electric power, as the weight of the batteries required to make it fly for that distance would be too heavy for it to ever get off the ground. Battery weight also contributes significantly to the energy usage of terrestrial electric vehicles, such that a reduction in battery weight could lead to EVs having a clearer economic advantage over internal combustion engine vehicles.

One technology that offers the promise of a partial solution to this problem is solid state batteries (i.e. batteries that use a solid rather than a liquid electrolyte), which will likely weigh a lot less than conventional batteries with similar capacity. Last year, Toyota announced that they have made some technological advances with solid state batteries and plan to begin producing them for automobiles by 2027. Due to their lighter weight (and other advantages like better safety), solid state batteries have the potential to significantly improve the economics of electric-powered transport. Depending on how this technology performs and how much it costs, solid state batteries could lead to an acceleration in the shift to electric vehicles, and feasible options for electric air travel.

Another possible technological development that could improve options for energy storage and assist the use of green energy for transport is that the efficiency of producing hydrogen (or other forms of chemically stored energy) from electricity will continue to improve. Hydrogen probably only has niche potential while the round trip of turning electricity into hydrogen then turning hydrogen back into electricity results in a 75% energy loss (as is the case with current technology). However, it seems possible that the energy cost of this round trip might ultimately fall to something like 50%, which would probably be good enough to make it an economic solution for a lot of transport. In this scenario, hydrogen production facilities could be built that only use electricity to produce hydrogen when electricity prices are cheap, or they could be built next to solar power plants in sunny deserts where there is a lot of sun but not much electricity demand.  

 Low Earth Orbit Satellites?

Over the last few years, Elon Musk’s Starlink has put over 6,000 low Earth orbit satellites into space. It was recently reported to have 3 million broadband subscribers, a 3-fold increase over the last 18 months. Other companies have also put numerous low Earth orbit satellites into space, and are using them for numerous applications, including gathering weather data, monitoring crop growth, and various forms of communications.

Clearly, using low Earth orbit satellites for broadband and communications is technically feasible, but the economics are hard to judge, as Starlink does not publish its financial statements. Starlink reportedly broke even on US$1.4 billion in revenues in 2023, which would work out at annualised revenues of about US$1,000 per customer. From these numbers, it is easy to imagine that with increased scale, Starlink might be able to make a profit on average revenues of about US$500 per customer, which would put it in a similar price range to broadband and mobile offerings of many telcos around the world.

Starlink have previously talked of spending US$10 billion on putting their satellites into space, but it is unclear how much capacity this buys them. Will they need to keep launching more and more satellites as their subscriber numbers grow? Or does $10 billion buy them a satellite network that can support several million subscribers?

If this model works for Starlink, it should presumably work for other companies as well. This could potentially include both startups and existing telecommunications providers looking to expand their offering. It is easy to envisage a big investment boom in launching low earth satellites if the economics of Starlink’s business model seem to stack up (these economics will likely become clearer if Starlink eventually lists on the share market). This could be good news for New Zealand’s Rocket Lab, although it also highlights a competitive issue that we are watching carefully at Te Ahumairangi, as we have significant investments in Verizon and other ground-based telecommunications companies.

It is likely that there will be some niches that are better suited to satellite communications and other niches that are better suited to conventional telecommunications. Conventional cellular technology divides the land up into “cells” that can each be serviced by different mobile phone towers, such that conventional telecommunications can pack a lot of capacity into a small densely populated area, such as New York City. By contrast, when you launch a low Earth Orbit satellite, you unfortunately know it is going to spend as much time flying over the Chatham Islands as it will spend flying over New York City. This indicates that if companies like Starlink do succeed, they will likely be ultra-competitive and have plenty of capacity in sparsely-populated areas like the Chatham Islands, but are less likely to make a big competitive dent in densely populated areas like Western Europe, Japan, or the more densely populated regions of the United States.

 Autonomous Vehicles?

Autonomous vehicles have already been developed and are already being tested in many places around the world, and although the technology is still being ironed out, the results to date are encouraging for the technology. For example, Alphabet’s Waymo has been offering a driverless Uber-like service in Phoenix, Arizona, and claims that after Waymo driving over 10 million kilometres for this service, they have achieved an 85% reduction in injury-causing crashes compared to what would be expected from human drivers.  

It seems likely that autonomous driving technology will be rolled out around the world over the next few years, subject to regulatory approvals. The technology is likely to be available both in cars sold to the public, and in driverless taxi services such as what Waymo are offering in Phoenix.

Driverless taxi services could be a huge risk to automobile manufacturers, as it seems reasonable to expect that driverless taxi services may ultimately be priced at less than half the price that you would currently expect to pay for a taxi or an Uber. For people (like myself) who never drive more than a few thousand kilometres in a year, owning and driving a car will therefore become a more expensive option compared to the use of driverless taxi services. And if this stimulates a profusion in the number of these driverless taxis, it could also make them the more convenient option in most cases, if the autonomous vehicle can be relied on showing up only a couple of minutes after you order it, and without the hassle of finding a park.

In a future world where driverless vehicles dominate, some car manufacturers may struggle due to less motor vehicles being sold each year, and because the main purchasers may be companies operating driverless taxis (that may be ruthless in seeking discounts for volume) rather than brand-conscious consumers. It is also conceivable that the biggest profit share of each vehicle sold might go to the software company (e.g. Alphabet’s Waymo) rather than the company that builds the physical vehicle.

 Anti-ageing drugs?

At the fringes of medical science, some writers and academics are suggesting that the process of ageing can be understood in terms of the progression of a small number of biological pathways that are becoming better understood. While there is evidence that lifestyle choices (like exercise, eating whole foods, and avoiding over-eating) can slow down these ageing processes, these writers suggest that drugs, supplements, and other medical therapies might directly “fix” the biological pathways that cause ageing.

I would be sceptical about the likelihood of anti-ageing therapies leading to any significant slowdown in how most people age, as I note that despite our increased understanding of how our body uses vitamins and other micro-nutrients, there is no evidence that nutritional supplements do any good for anyone who doesn’t have a known deficiency. However, I don’t think that we can entirely rule out the possibility that drug companies may discover therapies that are proven in clinical trials to slow down indicators of ageing. If anti-ageing therapies are proven to work, I find it easy to imagine a huge potential market for these therapies from people who are keen to live healthier for longer (particularly if they can do that without hard exercise or eating their greens!). Such therapies could potentially represent an even bigger market than GLP-1 based weight loss drugs.

 Many more!

A shortage of time and space means that I’m not going to try to discuss every emergent technology that may create cause for future investor excitement, but I would be interested in any technologies that readers of this article believe could be the “Next Big Thing”.      

 Nicholas Bagnall is chief investment officer of Te Ahumairangi Investment Management

Disclaimer: This article is for informational purposes only and is not, nor should be construed as, investment advice for any person. The writer is a director and shareholder of Te Ahumairangi Investment Management Limited, and an investor in the Te Ahumairangi Global Equity Fund. Te Ahumairangi manages client portfolios (including the Te Ahumairangi Global Equity Fund) that invest in global equity markets. These portfolios hold shares in Alphabet, Toyota Industries, and Verizon, which were mentioned in this column.