Key Themes for 2024 and Beyond

2023 has been another fascinating year for anyone interested in understanding future trends. The rise of AI in general and Chat-GPT specifically has dominated more headlines than almost any other topic, but record temperatures and extreme weather conditions have also been front of mind for many. The launch of a potential ‘wonder drug’ for treating obesity has been another hot issue.

As in our previous editions, we present below a summary of how and where we see the world changing most rapidly. Please note, this review is non-exhaustive; think of it rather as a series of interlinked high-level perspectives. Our stance remains that as diverse future trends overlap and intersect, they become mutually reinforcing.

AI, Data And Digital Transformation

• Season Five Blog Posts: Eyes on the future (5 January); Explain AI to my mum (11 January); Data dreams and AI aspirations (9 March); Did Chat-GPT steal my homework? (3 May); The “most transformational” technology (27 June); Coffee with the Head of IT (6 July); Get on the train! (5 October)
• Relevant Theme Pieces: The rise of the smart machine (April 2016); Your questions on AI (answered by a human) (March 2023). See also The data deluge (March 2011) and Ten years on; the data deluge 2.0 (October 2020)
• Key Statistic: Generative AI could add $7tr to global GDP over the next decade (Goldman Sachs)

“A race [is now on] to apply generative AI into every product, service and business process”

Jensen Huang, Chief Executive of NVIDIA … but

“The responsible application of AI should enhance human intelligence, not replace it or repeat its mistakes”

Eric Schmidt, former CEO of Google

In 2023, it was hard not to have escaped the whole AI bubble. Think of the comment from Bill Gates that AI is “as important as the PC, as important as the Internet” or from Jensen Huang, CEO of NVIDIA, who believes that a “new computing era has begun.” With 100m users in its first two months since launching (faster than the adoption of TikTok by a factor of almost five – per UBS), think of Chat-GPT as the first killer app of the AI era, useful for consumers and enterprises. For those who have yet to use it, the service allows for out-of-the-box experimentation, prompt engineering and custom modelling.

Using Chat-GPT today, however, may not be dissimilar to using a Nokia mobile handset in the 1990s. This was the first mobile that the average person could appreciate, but looking at it then, there was no way to predict how smartphones would end up a generation later. More broadly, at this stage, it is unclear whether adding AI to the mix will make us better or worse off, since every technological change brings unintended consequences. While there is no doubt that AI machines can perform reflexive statistical analysis, they have little or no capacity for deliberate reasoning at this stage. Sure, these machines can ‘learn’, but this largely revolves around the statistics of words and appropriate responses to prompts. What they are not doing is learning abstract concepts. Unlike humans they are not seeking to understand the world.

From an investment perspective, it is also worth considering what if generative AI becomes so ubiquitous that it becomes commoditised? In this world view, how might companies charge a premium or seek to monetise it? Everyone these days wants to describe their company an AI business. But it’s not possible to have it both ways: insisting that AI is crossing a boundary into the unknown, and at the same time, defining it so broadly that it includes AI, Data And Digital Transformation Key Statistic: Generative AI could add $7tr to global GDP over the next decade (Goldman Sachs) “A race [is now on] to apply generative AI into every product, service and business process” Jensen Huang, Chief Executive of NVIDIA … but “The responsible application of AI should enhance human intelligence, not replace it or repeat its mistakes” Eric Schmidt, former CEO of Google 18 19 Future Trends Future Trends almost any large-scale, statistically driven computer programme.

At a practical level and irrespective of how AI evolves over the longer-term, the technology is being practically adopted today. The reason is simple: over 50% of US firms, more than 75% of Eurozone businesses and greater than 90% of Japanese firms currently report a shortage of skilled workers (per KKR) AI can mitigate if not resolve the worker shortage dynamic. A recent study from McKinsey suggested that a third of the businesses it surveyed globally say their organisations are using generative AI in at least one business function currently. Another study (from the Boston Consulting Group) suggests that using AI can make people complete tasks 25% faster and increase the quality of their work by 40%. Correspondingly, more than 40% say they expect to increase their investments in AI owing to recent improvement in its capabilities. A KPMG report highlights that 85% of businesses expect to be using more AI in the coming years.

It is perhaps not surprising then that spending on generative AI software is expected to reach $18bn in 2027. With less than $1bn spent in 2022, this is equivalent to a compound annual growth rate of more than 190% (per TD Cowen). A recent study from Stanford University shows that workers can become at least 15% more productive when given generative AI tools vs those who did not, with least-skilled workers benefiting the most. Least-skilled workers can apparently get their work done 35% faster, per the study. Against this background, Goldman Sachs is forecasting a ~$7tr increase to global GDP from generative AI over the next decade, primarily from the automation of tasks associated with numerous jobs.

A corresponding ecosystem has already begun to develop. Pitchbook believes more than $676bn of venture capital money was allocated to AI in 2022. The figure will undoubtedly be much higher for 2023. A recent analysis by New York University estimates that there are over 100 AI unicorns (private businesses with a valuation of more than $1bn) in the US and almost 40 in Asia. Nonetheless, progress won’t be linear. For context, although the internet began to be used by some companies in the early 1990s, it was not until the late 2000s that two-thirds of American businesses had a website (per a study conducted by The Economist).

Nonetheless, progress won’t be linear. For context, although the internet began to be used by some companies in the early 1990s, it was not until the late 2000s that two-thirds of American businesses had a website (per a study conducted by The Economist). The biggest reason cited for not adopting AI currently is cost. More than half of all AI decisionmakers in top companies say that they are facing cost barriers to deploying the latest AI tools (according to S&P Global). Correspondingly, around 70 firms within the S&P 500 Index still show no interest in AI. Likewise, one-third of small businesses in the US and Canada say they have no plans to try generative AI tools in the next year. For early movers, this could prove to be advantageous. Scale matters. Consider that JP Morgan employs some 600 dedicated machine learning engineers while Eli Lilly has over 100 projects on the go involving AI.

At the heart of any AI development lies data. This was the point that we made in our very first theme piece (reproduced in this volume) over a decade ago. Data have no value unless stored, secured and analysed. In the world of AI, perhaps the best analogy for data is less as oil – a valuable commodity over which countries may go to war – and more as sand; it is only valuable when aggregated into the billions. Digital transformation continues to change the world. It is growing at eight times the pace of the broader economy and will create a $100tr economic opportunity over the next decade (per IDC and the World Economic Forum respectively). No surprise then that the semiconductor market is expected to double over this period. Admittedly this forecast is courtesy of ASML, the business that is market leader in the manufacturing of the machines that make semiconductors, but as Peter Wennink, its CEO, notes, “we have consistently continued to underestimate demand.”

To believe in the beneficial nature of digital transformation, of course, requires an optimistic world view. A recent survey from the Stevens Institute of Technology found that 74% of respondents believed that AI could lead to a loss of personal privacy, while 71% were of the view that it would reduce employment opportunities. 60% felt that it would increase political polarisation. The job loss concern is perhaps the easiest to rebut. Consider that 150 years ago, 60% of the world’s population worked in agriculture. Now, fewer than 3% do. There will always be opportunities. At the same time, it behoves both politicians and business leaders to consider how to manage the development of AI responsibly.

Healthcare (And Food Innovation)

• Season Five Blog Posts: Happy Birthday HGP (26 April); In conversation with Dr Gordon Sanghera (23 May); Diagnostics for all (3 August); Our Future Health (17 August)

• Relevant Theme Pieces: Fat profit potential: bulging bellies and growing obesity (April 2012), Reinventing healthcare & the coming age of personalised medicine (November 2012), The man-machine merger (September 2016), Time for a DNA upgrade? (June 2017)

• Key Statistic: 4bn overweight and obese people could cost the world economy $41tr by 2035 (World Health Organisation)

“AI could usher in a new renaissance of drug discovery”

Demis Hassabis, co-founder of Google DeepMind

If AI can be thought of as the age of wonder, then consider its potential for accelerated scientific discovery. Drugs can take a decade to emerge, cost billions of Dollars and succeed only 10% of the time. Even a small improvement in speed and efficiency would be hugely valuable. Machine learning makes it possible to sift through piles of information, from clinical patient data and genome sequences to images of body scans. Unlike a human researcher, it can also do this 24 hours a day, every day of the year. Much of the current buzz in healthcare therefore revolves around AI trained on biological data that could improve the hit-and-miss process of drug discovery. Forecasts from Gartner suggest that over 30% of innovative pharmaceuticals, materials and synthetic marketing content will be created by generative AI by 2030.

Such a development would be highly welcome given the huge cost burden to all developed economies of healthcare, a trend that will only be exacerbated as populations age. Consider that 90% of all US healthcare costs are accounted for by treating preventable, chronic conditions. The biggest of these is obesity. Nearly 70% of American adults are overweight and over a third are obese. Almost half of Americans will be obese by 2030, research by Harvard University has found. About 18% of healthcare spending by this data would be required to treat it and other related conditions.

Obesity may be profound in the US, but it is also a global problem. It currently affects about 650m people worldwide. By 2035, this figure could swell to more than half, with a staggering 4bn people overweight or obese (based on data from the World Healthcare (And Food Innovation) Season Five Blog Posts: Happy Birthday HGP (26 April); In conversation with Dr Gordon Sanghera (23 May); Diagnostics for all (3 August); Our Future Health (17 August) Relevant Theme Pieces: Fat profit potential: bulging bellies and growing obesity (April 2012), Reinventing healthcare & the coming age of personalised medicine (November 2012), The man-machine merger (September 2016), Time for a DNA upgrade? (June 2017) “AI could usher in a new renaissance of drug discovery” Demis Hassabis, co-founder of Google DeepMind Key Statistic: 4bn overweight and obese people could cost the world economy $41tr by 2035 (World Health Organisation) 20 21 Future Trends Future Trends Health Organisation). The annual cost to the world economy of excess weight could reach $41tr by 2035 (or 2.9% of global GDP, up from 2.2% in 2019). This includes spending on healthcare and working time lost to illness and premature deaths tied to obesity.

The relative good news is that there are treatments on hand already. Wegovy, an anti-obesity drug developed by Novo Nordisk, has not garnered quite as many headlines as AI year-to-date, but its success has been almost as revolutionary. The drug (and those being developed by competitors) have been de facto clinically proven to reduce body weight by 15% or more in some people. They work by decreasing appetite and in turn can decrease calorie intake by 20-30% daily. Some estimates (for example, from Jefferies, an investment bank) suggest that the market for such drugs could be worth over $150bn by 2031. For context, the market size of all drugs to treat cancer was worth $185bn in 2021.

Treating obesity matters, but is however, just the tip of the iceberg. Alzheimer’s is currently the most expensive disease in the US (per a study from Harvard University), given related palliative care costs. More than 55m people live with a dementia disorder worldwide, and by 2050 that number is expected almost to triple. The global cost of treating dementia disorders is estimated to be $1.3tr and will more than double over the next decade. As important as drug development is, lifestyles decisions do also matter, whether it be increasing exercise or making better dietary choices. Food innovation remains an important future trend. Were space to permit, we would discuss it more, although it has featured regularly as a topic in our Blog posts over 2023 (see posts #18, #22, #24, #33).

Nonetheless, whether it be obesity, cancer, Alzheimer’s or any other disease, the logic for datadriven drug discovery is high. A recent KPMG study highlighted that 82% of healthcare and life sciences executives are currently seeking a more aggressive adoption of these technologies. Venture capital funding for healthcare AI solutions is set to reach over $10bn this year (per CB Insights).

Alternative Energy

• Season Five Blog Posts: “Nothing has made a difference” (25 January); “There’s a piece for everyone” (3 March); Not quite full speed (20 April); Turbine-tastic (27 July); Staying cool and getting greener (14 September); The sun always shines in Ohio (21 September); Get on the train! (5 October)

• Relevant Theme Pieces: What if the sun always shone? The coming energy storage revolution (September 2015), Winds of change (March 2018), Everybody loves the sunshine (January 2020), Winds of change: offshore edition (September 2023)

• Key Statistic: Despite more than $6tr spent on renewable energy since 2005, the world is still ~80% reliant on fossil fuels (JP Morgan)

“Our world needs climate actionon all fronts”

UN Secretary General, António Guterres

“We’re on the edge of a new age of electrification; everything that can be, will be”

Mark Widmar, Chief Executive of First Solar

Another enduring memory for all of 2023 will be just how hot the year was. Temperature records were broken globally. Climate change may be the culprit but consider too the impact of technology. Data centres consume vast amounts of energy, equivalent to a range of 1-7% of total electricity usage, depending on country. Cooling can represent up to 50% of a data centre’s power consumption (per KKR). With the amount of computing power required to train a cutting-edge AI model currently doubling every five months, there is only upside (or downside if you’re concerned about climate change) to this estimate.

From a bigger picture perspective, although some $6.3tr has been spent on renewable energy and another $3.3tr spent on electricity networks since 2005, global energy use is still ~80% reliant on fossil fuels (per JP Morgan). This is because – and forget data centres for a moment; this is a newer problem – the main pillars of modern society are still made primarily using fossil fuels. Think of the role played by cement, steel, ammonia and plastics. Take the former. About three tonnes of concrete are poured annually for every person in the world. It is (according to work published in The Economist) responsible for ~8% of manmade CO2 emissions. Were concrete a country, it would be the third-largest global emitter of CO2, after China and the US. There is a lot, therefore, that needs to be done.

The good news is that over 70 countries, 1200 companies and 1000 cities have made net-zero pledges (per the United Nations). Further, thanks to cost efficiencies, 90% of new global power capacity is forecasted to come from renewables by 2027, with solar making up the major share. The price of solar modules has declined by 99.6% since 1976 – the year your author was born – with the price fall in the past year being equivalent to more than 50% (data from IEA and Bloomberg respectively).

As a result, global capex on wind and solar assets was greater than investment in new oil and gas wells for the first time in 2022 (per Rystad Energy). Combined power from wind and solar overtook natural gas in the EU for the first time last year. In the US, wind and solar generated more power than coal during the first half of 2023 (data from Ember Analytics and Future Crunch respectively).

Looking forward, the EU and US are now expected to install over twice the amount of wind capacity in the next five years compared to the prior fiveyear period (per Jefferies). While Europe has had a long history of seeking to develop alternative energy projects, the biggest global game-changer has been in the US. Admittedly, geopolitics has played a role in increased political fervour for alternative energy in much of the west, but since the Inflation Reduction Act (IRA) was passed, over 50 new or expanded solar projects have been announced. By 2026, the US solar ecosystem will be over 15 times its pre-IRA size (per the Solar Energy Industry Association).

Against this background, renewables could become the largest source of global electricity generation by early 2025, surpassing coal. Their share of the power mix is forecast to increase by 10 percentage points over the forecast period, reaching 38% in 2027. In the International Energy Authority’s (IEA) study on this topic renewables are the only electricity generation source whose share is expected to grow, with declining shares for coal, natural gas, nuclear and oil generation.

As encouraging as such developments sound, more needs to be done. In order to reach the United Nations’ (perhaps optimistic) target of net zero by 2050, annual investments in renewable energy solutions would need to be at least $1tr, a consensus shared by most forecasting bodies. Today’s sums are markedly lower. How to move forward? Improved support from governments and regulators matters. The IEA estimates that renewables generation could rise by an extra 25% by 2027 were bureaucratic and financing barriers were removed.

Further, in any analysis, security of supply also matters. Energy transition requires critical minerals such as lithium, nickel, and cobalt. Demand for these could be 20 times higher in aggregate in 2035 than it was in 2021 (per S&P Global). Notably, China refines 60% of the world’s lithium, 68% of its nickel, and 80% of its cobalt. These dynamics also matter when considering the car of the future, our next topic.

Car Of The Future

• Season Five Blog Posts: The autonomous bus stops here (17 May); Get on the train! (5 October)

• Relevant Theme Pieces: The long road to autopia (April 2015), The new transport revolution (Feb 2017), Disrupt the car, and rethink the city (April 2021)

• Key Statistic: Almost one-in-five cars sold this year will be electric vehicles (IEA)

“Our industry is going to be shaped by safe, green and connected megatrends”

Kevin Clark, Chief Executive of Aptiv

There is perhaps no stronger confluence of the two themes of data and alternative energy than when it comes to thinking about the car of the future. The success (or failure) if electrifying and automating the mobility industry will depend on its ability to efficiently and economically produce vehicles that can utilise two critical resources: electricity and data. This will be addressed with both hardware and semiconductor and software technology innovation. For context, electric vehicles (EVs) need around twice as many semiconductor chips as a conventional car.

That EVs are here to stay is not open to debate. Rather, the consideration is how quickly can they scale. Global EV sales surpassed 10m units in 2022 and this year are forecast to reach 14m, equivalent to 18% of all new cars sales (up from 14% the year prior). Nowadays, motorists can choose between 500 or so EV models. Look further ahead and 100m passenger EVs are expected on the roads by 2026 and over 700m by 2040 (per the IEA).

The challenge is that the supply of the lithium required to make EV batteries must grow by a third every year this decade to meet estimated demand (per Bloomberg New Energy Foundation). This is exacerbated by the fact that c90% of the EV battery supply chain relies on China, with the largest Chinese battery companies controlling more than half the global market. To change this paradigm will take time. Work conducted by Cowen, a US broker, suggests that it takes roughly two years to construct a battery plant, while the upstream mining buildout required to secure the raw component materials for battery end products requires a much longer lead time of between five and ten years. Irrespective of data, geopolitics will almost certainly play a role in determining how quickly the EV industry develops.

Cybersecurity

• Season Five Blog Posts: Coffee with the Head of IT (6 July); Experience Centres experienced (28 September)

• Relevant Theme Pieces: Watch out! The growing privacy invasion and cybercrime threat (April 2014), The next generation (September 2017), Ten years on: the data deluge 2.0 (October 2020)

• Key Statistic: Cybercrime cost the world $8tr in 2022 (Palo Alto Networks)

“Everyone thinks they have a plan until they get punched in the face”

Mike Tyson, former boxing champion

A remarkable $100bn has been spent on self-driving cars cumulatively (according to McKinsey), but little progress has been made. Some automakers have scaled back ambitions, while Ford and Volkswagen have pulled the plug, for now, on their self-driving car efforts completely. This has primarily been as a function of too many unpredictable ‘edge cases’ for autonomous cars to figure out. Nearly 70% of drivers say they’re afraid of fully self-driving cars, per a survey from the American Automobile Association. About a quarter say they’re unsure about them, while just 9% say they trust them.

The prior paragraph links our earlier discussion of the future car with our final major topic in this report: cybersecurity. The key data point relates to trust. Data, as we have said many times, have no value unless secured, stored and analysed. This is as relevant in the world of cars as it is in drug development or any other industry. Consider that the annual global cost of cybercrime reached $8tr last year. More than 10m people and over 1700 organisations were affected by cyberattacks in 2022. There has been a 20-fold increase in malware programmes over the last decade (per Palo Alto Networks).

There’s more. The average data breach cost rose to $4.5m per incident in 2023, a 15.3% increase relative to the start of the decade. Ransomware revenue (payments) in the first of half of 2023 was already at 90% of 2022 levels with the number of ransomware attacks globally having more than tripled since 2019 (data from IBM and Chainalysis respectively). No surprise then that almost two-thirds of executives are concerned about cybersecurity, up from 58% the prior year. Investment in IT budgets is on the rise too, with one-third of businesses increasing dedicated cybersecurity budgets by 14% (per Microsoft)

As a result, the cyber industry is worth over $175bn and still growing at a double-digit rate (according to Gartner). AI – to return to our first discussion topic in this report – could be a game-changer. The threat lies in the fact that there is the potential for more sophisticated attacks now to be perpetrated, but one of the largest opportunities, or use-cases for this new technology could come from the effective deployment of preventative algorithms.

What we didn’t have time to cover in more detail…

Space in these reports is sadly finite and so there was no opportunity to cover on this occasion other important future trends such as automation, the circular economy, drones, hydrogen, logistics, payments, robotics and more. Many of the above have featured in our 2023 Blog posts and all have been addressed in thematic white papers over the years.

If we were to select just one additional topic to discuss briefly, then it would be water. Although the sector accounts for just 1% of global GDP, the world economy depends on water to exist. Ensuring access to water and making sure that it flows efficiently therefore matters, particularly in the context of a forecasted 40% gap between global water supply and demand by 2030 (per the World Economic Forum). Outdated water-management infrastructure is already causing global economic losses of approximately $470bn annually. If not addressed, by mid-century, water risks could wipe out $5.6tr from global GDP. Required investments in global water infrastructure therefore range from $6.7tr by 2030 to over $22tr by 2050 (per 13D Research). Beyond this, consider that the United Nations estimates that 640m people globally have critical levels of water insecurity. Water matters.