The development of artificial intelligence (AI) is one of the most talked about issues of our time. Usually anxieties about its development dominate many discussions, and regularly form the basis of some dystopian future when it comes to adopting it within a narrative.
From an investment perspective, however, the future is far more upbeat when it comes to AI and robotics.
In fact, the rise of such an innovation has become the investment theme most cited by institutional investors. And this usually involves the N word – Nvidia – which is an ever-present reference within the AI debate for obvious reasons.
Indeed, anyone who doubts that AI will change our lives should take a look at Nvidia. It makes the chips that AI applications run on. In its latest quarterly earnings, the company reported revenues of $18bn (£13.7bn) – an increase of $12bn (£9bn) against the same three months of the previous year. Analysts are forecasting that in just two years, Nvidia’s annual revenues will have grown by more than $60bn (£45.7bn).
If that much is being spent on the components that enable the technology to be deployed into real world uses, it seems fair to assume that quite a lot of change is going to happen. There will be a few big winners within this: processor designers, hard- ware producers, datacentre operators and all that.
“For me, it is the potential for companies to up their returns by deploying the technology to cut through bottlenecks and lift service standards that is most exciting,” says Steve Clayton, head of equity funds at Hargreaves Lansdown.
Whole new product categories are emerging. Microsoft’s Copilot costs $30 (£22) per month but reportedly can boost productivity by far, far more. So one day, we could potentially see hundreds of millions of workers using Copilot. That could add up to a lot of GDP growth and a lot more revenue for Microsoft.
Hargreaves Lansdown holds positions in Microsoft and Nvidia. “These businesses are some of the core beneficiaries of the early stages of the surge in AI development,” Clayton says.
An AI promise
AI has been stealing the limelight for good reason as it promises to turbocharge the power of the digital economy by making systems smarter.
The range of sectors AI will influence is also vast. “In healthcare, AI can help to boost accuracy and output in areas like radiology by highlighting areas that need particular attention by the radiologist and speeding up the interpretation of scans,” Clayton adds.
Connecting sensor data gathered in hospital wards to AI engines will enable earlier interventions by medical teams.
The possibilities are huge.
“Deploying AI into real-world situations will lead to more digital equipment, from sensors to servers being required. Digital businesses will be more capable and require more capacity as a result,” Clayton says.
And on the road cars will have more tech in them, they will be smarter and possibly safer as a result.
Aptiv is one of the largest manufacturers of core electric vehicle components and intelligent electronics for the new generations of automobiles.
And in a broader sense, technology is ubiquitous across economic activity providing new investment opportunities for investors, the most recent of which is the transition from digitalisation to artificial intelligence in the services sector.
“We have seen the acceleration of autonomation through AI, notably several industrial companies have spent heavily on robotics and automation in order to improve output quality across their production workflows,” says Marco Barresi, equity research analyst at Lombard Odier.
“This trend also includes ‘embodied Al’, which focuses on how technology interacts with the physical world, typically as a robot or ‘cobot’ (collaborative robot), including autonomous vehicles and drones,” Barresi adds.
Robots can use Al to learn from their own interactions with the world, and to show conversational and situational awareness as their abilities to read labels, interpret signs, understand shapes, calculate volumes or package objects improves. This is driving a shift from automating simple tasks to autonomous machines, according to research by Intel.
“While Al has been around for years, we are on the cusp of a major inflection point as the various prerequisites are rein- forced by the growing infrastructure behind data centres,” Barresi says. This is illustrated, he says by the “tremendous increase” in semi-conductors designed to support Al’s architecture.
“We expect this infrastructure-related spending to filter down into devices – smartphones and desktop computers – that access networks, and where most data is generated,” he adds.
High-paced innovation cycles will spur new use-cases across many sectors, thanks to customised solutions based on differentiated industry know-how and proprietary data, Barresi says. “This offers investors plentiful opportunities to benefit from exposure to well-positioned companies in the Al value chain.”
Moreover, business intelligence and analytics, increasingly powered by Al will now remain, or become, a top priority for investment as enterprises continue to digitise and consumer behaviour shifts towards digital channels, he adds.
Big change
Yet many institutional investors do not fully grasp the changes, according to Jonathan Curtis, chief investment officer and portfolio manager at Franklin Equity. “Despite witnessing the phenomenal growth of AI infrastructure businesses, we are convinced that many investors do not fully comprehend the improvements in AI models that are yet to come,” he says. “They also do not foresee how these improvements, once employed, will enable innovation, accelerate growth and boost the efficiency of the global economy.”
Curtis adds that in his view, we are on the cusp of radical change as models evolve from text-centric chatbots and media creation engines to active agents.
“These agents will be able to reason and use computers, behaving more like the world’s most advanced human digital workers, such as scientists, mathematicians, programmers and AI researchers. We believe this is when innovation, growth and efficiency will begin to accelerate in a meaningful way,” Curtis says.
It will all contribute to change in some shape or form. Zehrid Osmani, head of the global long-term unconstrained team at Martin Currie, highlights three “seismic thematic shifts” that are likely to shape investment opportunities and will be structural growth drivers over the next decade and beyond.
These are the energy transition, the ageing population and artificial intelligence. “Artificial intelligence will influence all three of those, in a broad sense,” Osmani says.
New themes
The nature of artificial intelligence is changing things so dramatically that different theories emanate from it. Osmani foresees four themes as being important for investors in this regard: one is technological and geopolitical fragmentation; two is cloud infrastructure and cyber security; three, robotics and automation; and four, metaverse and quantum computing.
“Robotics and automation is a rapidly growing structural trend, which has been further boosted by the advent of AI. It will lead to a rapid take-up of smarter robotics, more autonomous automation, and various enhancements in industrial and services sectors,” Osmani says.
As part of this, Martin Currie forecasts robotics and automation to grow at a compound annualised growth rate of +25% over 10 years, which highlights the immense magnitude of the growth opportunity for investors.
“As we get into a future of robotics and automation interacting in a more ubiquitous manner with humans – think autonomous driving, autonomous robots, etc – there will be a need for faster computing power, which means that companies such as Nvidia are well placed to capture that structural growth segment,” Osmani says.
AI pyramid
Tom Riley, lead portfolio manager of the AXA Robotech strategy, believes the AI Investment universe looks like a pyra- mid, split in to three main areas of infrastructure, technology and applications.
The investor focus is at the bottom of the pyramid – the infrastructure and technology segments. These are the semi-conductors and the hardware used to build AI ecosystems as well as the cloud computing platforms on which they run.
“At present, it is particularly the infrastructure areas where companies have seen the most immediate impact and share prices have risen the most over the last 18 months,” Riley says.
Currently, the applications are more limited, the development of Generative Artificial Intelligence and Large Language Models (LLMs) – such as ChatGPT – are at the early stages of commercialisation, Riley says. Indeed, the true range of applications of AI are still to be determined.
“In the coming years, we believe the pyramid should invert as the opportunity for applications of AI become more significant,” Riley says.
There are parallels here with other technology changes – mobile internet, for example – the hardware companies derived some value from these technology shifts, but the real economic value was created in the new businesses which leveraged 3G, 4G, 5G – social media, ecommerce, streaming and the much used Ride Hailing app to get your Uber.
“Many of the beneficiaries will be outside of the technology sector, so traditional technology investment funds may not be positioned to capture all of these opportunities,” Riley says. “‘Robotech’ should be sufficiently flexible to capture many of the most exciting AI-enabled opportunities across the economy.”
There are therefore other extension areas for investors to consider within this, particularly in the rapidly growing area of robotics.
For example, factories worldwide introduced a record number of industrial robots in 2022 – a total of 553,052, 5% more than in 2021, according to the International Federation of Robotics. Between 2023 and 2026, the market is expected to grow by 7% per year.
The investment universe for robotics and AI is much broader than most people anticipate, with opportunities in industrial applications, transportation, healthcare and the technology enablers. “We see AI moving into the warehouse automation space and have investments in companies that have autonomous bots that move around warehouses collecting packages,” Riley says.
“In the transportation space we are getting increasingly closer to breakthroughs and adoption of autonomous driving applications,” he adds. In the healthcare space, Riley says there is huge amount of development going on in the robotic-surgery space.
The leading provider in the healthcare space, Intuitive Surgical, which carries out around 2.5 million surgical procedures per year, has just announced a new robot for the first time in many years. This Robot has 10,000 times the compute speed of the prior generation.
And perhaps the most important are the technology enablers. “These are the semi-conductors, the sensors, the software, the AI that drives the growth of the industry,” Riley says.
The growth of the robotics industry has accelerated over the past decade due to new technologies allowing machines to serve new markets. “We believe that AI should be a further catalyst to drive this going forward,” Riley adds.
But this is a particularly interesting time to be looking at the area as three things are happening at once: one, the structural growth of the industry continues. Two, the technology change from AI supplements this growth opens new markets. And three, a cyclical recovery emerges for the first time in many years. Alongside more traditional robotics, there are headlines at the moment on the concepts on physical AI and humanoid robotics.
“These concepts are in their infancy, so we would be cautious in making too many predictions here, but they are interesting developments to monitor,” Riley says.
“We believe physical AI could become the next major wave in AI and enable the digitalisation of heavy industries,” Riley adds. Physical AI – or AI that understands the laws of physics – is crucial for robotics and industrial digitalisation.
And guess who is a pioneer in this area? Yes, it is Nvidia. Their Omniverse simulation platform and its Isaac robotics platform are being adopted by such key industry participants as Delta Electronics, Foxconn, Pegatron, Wistron, BYD Electronics, Siemens, Teradyne and Intrinsic. “This is essentially the AI version of the next industrial revolution that remains ahead of us,” Riley says.
“We view the focus on physical AI at its flagship conferences by Nvidia as encouraging,” he adds. The growing number of partnerships Nvidia has within the industrial sector and the breath of markets covered help raise the opportunity for potential breakthroughs.
Harnessing robotics
Significant advances in AI coupled with labour markets that continue to be tight are ramping up the use of automation, as an ever-rising number of companies increasingly harness robotics to bolster efficiency, safety and precision, Riley says.
The car industry accounts for around a third of all robots in factories worldwide and most car manufacturers use automation at nearly every stage of vehicle production, with battery manufacturing for electric vehicles proving to be a large new market for automation technologies.
“Exciting developments in the use of artificial intelligence to develop electric vehicles, as well as cars that can optimise fuel usage, navigate in real time, alert drivers to maintenance issues and even drive autonomously,” Riley says.
Driving all this are software, semi-conductor and other key component providers. “The use of robotics and automation can reduce costs and increase efficiency, quality and safety,” Riley says. “Connected or smart factories can gather and assess data which can optimise processes, helping manufacturers respond swiftly to changing demand, as well as recognising problems before they cause wider disruption.”
Some of this has happened faster due to Covid. The pandemic boosted demand for online shopping and meant many retailers and distributors needed to scale up their fulfilment of orders, increasingly turning to automation.
“From Amazon to Ocado, companies are using warehouse robotics for managing orders, picking, sorting, packing and more – and this market is forecast to almost double from $12.9bn (£9.8bn) in 2023 to near $25bn (£19bn) in 2028,” Riley says.
New and old
But it’s not just a case of new innovations. Investment in new technologies is coming as manufacturing systems age, in the US in particular.
“Across manufacturing, US machinery is the amongst the oldest it has ever been,” Riley points out. “The average age of manufacturing capital stock is 11 years, according to 2022 data, compared to an average of eight years over the past several decades, indicating that upgrades are imminent.
“Industrial companies worldwide expect to spend heavily on robotics and automation, led by logistics and fulfilment firms,” he adds. “But successful implementation is a concern among businesses, suggesting the most successful robotics and automation providers will be those that can help their industrial clients overcome the challenges.”
The combination of changing consumer and corporate demand, and the backing of new government policies like the Inflation Reduction Act (IRA), means a considerable potential boost for the robotics and automation sector.
“As demands for technology become more complex, we see potential investment opportunities in semi-conductor firms, such as Nvidia and Taiwan Semiconductor Manufacturing Company, that innovate to keep pace with these changing needs,” Riley says.
Meanwhile, sectors that are at the forefront of harnessing the benefits of AI, including much cited transportation and healthcare, look set to continue to benefit as the use of robotics and automation becomes further embedded.
“We believe the robotics and automation sector is a significant growth market, and one where we are still in the fledgling stages of its development; not only in terms of its expansion potential but also in terms of the long-term investment opportunities it presents,” Riley says.
And when it comes to humanoid robots, after unveiling its first humanoid robot, Optimus, in 2022, Tesla now forecasts that several thousand Optimus robots will be working in its factories by 2025 with sales of Optimus Version 2 scheduled for external customers a year later.
Elon Musk forecasts more than 1 billion humanoid robots will be in operation by 2040.
Taking a step back, for all this elaborate and enthusiastic talk surrounding AI, is it possible that could AI be hot air and mere hype? It wouldn’t be the first time that the investment world has fallen for such hysteria.
Addressing this, Osmani gives a different slant to such an idea.
“There is a debate on whether AI as a theme is overhyped, but we continue to see the implications of the advent of AI as being underestimated by the market in terms of potential market size, and in terms of speed of take-up,” he says.
In short, the opportunities for investors are like AI itself, somewhat mind-boggling. And institutional investors have the intelligence to fully exploit those opportunities.
It is therefore true to say there is nothing artificial about AI and robotic investment.
Comments