Key Themes for 2025 and Beyond

2024 has been another fascinating year for anyone interested in understanding future trends. AI has remained front of mind but the debates around this topic have encouragingly broadened to include cybersecurity implications and whether existing grid infrastructure is robust enough to support growing data demands. Better grids need more inputs, particularly from a diverse range of alternative energy sources. Elsewhere, drugs to treat obesity continue to be a hot topic in the healthcare sector. How the food industry responds and evolves is also highly relevant.

As in our previous compendiums, 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

• Season Six Blog Posts:
  Red-hot innovation (13 February); The transition mission (21 February); California Dreaming: Notes from the Valley (19 March); AI and Its Discontents* (9 April); “There’s never been a better time to be a builder” (25 April); Coffee with the CTO, 2024 edition (29 May); Live from London Tech Week (11 June); 27: “Humanlike decisions, but better” (17 July); Postcard from America (21 August); “Embrace change” (October 2024)
• Relevant Theme Pieces:
  Artificial Intelligence: The rise of the smart machine (April 2016); Your questions on AI (answered by a human) (March 2023)
• Key Statistic:
  The four largest tech companies in the US will spend ~$200bn on AI infrastructure in 2024 (Bloomberg)

““The next industrial revolution has begun… AI will bring significant productivity gains to nearly every industry and help companies be more cost- and energy-efficient, while expanding revenue opportunities”

Jensen Huang, Chief Executive of NVIDIA

Investment in AI has become an arms race, with a potential risk of serious over-spend in the near-term. The longer-term dynamics have already been discussed in our introductory essay. Consider for now that the four largest hyperscalers will spend around $200bn in 2024 on data centres, chips and other related equipment for training and developing AI models, a 45% increase versus 2023 levels, based on consensus Bloomberg data. Much of this spend may go on NVIDIA’s latest Blackwell GPU architecture, which promises AI training at 4 times the speeds and inferencing 30 times more rapid than the prior generation.

Given all the excitement around the technology’s potential, it should also come as no surprise to readers that AI is driving the venture capital ecosystem. The average AI funding round anecdotally exceeds a non-AI one by a factor of six, with an implied valuation up to five times higher. AI has accounted for over 40% of all new unicorns (unlisted businesses worth at least $1bn) formed in the first half of 2024, per Coatue and Pitchbook respectively. The US now possesses over 5,500 AI start-ups and China more than 1,400. Even the UK has some 700, according to Visual Capitalist.

The biggest challenge for the industry remains how to translate infrastructure and ideas into practical products. Consultants, at least, seem to be benefiting, with Accenture detailing a sixfold increase in generative AI bookings in its last reported quarter, versus a year prior.

The C-suite, however, remains more sceptical. 71% of industry leaders polled by BCG earlier this year said they were “pursuing limited experimentation and small-scale pilots” while the same study showed that a majority of executives felt that it would take “at least two years” to move beyond the hype. A study by the US Census Bureau (admittedly conducted in February) showed that only that only 5% of American firms of all sizes said they had used AI. Just 7% at that time said they planned to adopt over the next six months.

Microsoft, however, paints a somewhat different picture, albeit from the perspective of its vested interest in the success of the technology. At its last earnings report, the company said that over 65% of the Fortune-500 businesses now use the Azure OpenAI service. Its interface allows companies to build their own generative AI applications and personalised Copilots. Some large companies have disclosed positive use cases. L’Oréal, for example, has reported that an AI diagnostics machine uplifts the sales conversion rate of its products at retail counters from 10% to 73%. AI for advertising and promotion at L’Oréal is generating productivity gains of up to 15%. In a different industry, payments company Klarna revealed that its AI assistant is effectively doing the equivalent work of 700 full-time agents and has helped reduce repeat inquiries by 25%.

The hope remains that as AI evolves, scales and matures that it can impact both growth and productivity positively. Optimists argue that AI could drive a 15% rise in GDP over the next 10 years (the view of Goldman Sachs) or that it could add contribute 0.3- 0.7pp to annual productivity growth in the US through to 2040 (McKinsey’s perspective). Given the significant Dollars and intellectual capital invested to-date in the industry, were such gains not to materialise, then investors may need to prepare themselves for some disappointment.

DATA

• Season Six Blog Posts:
  The future starts here (4 January); Red-hot innovation (13 February); California Dreaming: Notes from the Valley (19 March); James Bond, for a morning (22 May); New governments, more data centres (9 July); Two ends of the data spectrum (24 September)

• Relevant Theme Pieces:
  The data deluge (March 2011); Drowning in data( October 2012); Ten Years On: The Data Deluge (October 2020)

• Key Statistic:
  The average North American spends five hours a day on their smartphone (World Bank)

“Digital is the most transformational technology our world has ever experienced”

Chuck Meyers, Chairman of Equinix

We are continuing to create huge – arguably unprecedented – quantities of new data even without the upside that AI is set to generate. Consider that 90% of all the world’s data was created in the last 18 months. The total amount of data created and replicated in 2024 is set to reach 163 zettabytes (or 21 zero’s worth), up 80 times since 2010, according to IDC.

If AI’s revolution is to further facilitate human connections with computers, then think of the Internet’s revolution as communication with any person. We do still ‘talk’ in person, but to do so uses a lot of technology. Every minute of the day, humans send 97m WhatsApp messages, 12m iMessages, 2m Snapchats, and make 575,000 posts to X (Twitter). In this same minute, some 5.7m Google searches will occur, over $283,000 will be spent on Amazon and more than 167m TikTok videos may be watched (all data per Visual Capitalist).

A figure of 1.8bn visits to Chat-GPT in March may sound enormous, but it pales in comparison with the 85.4bn for Google, 32.4bn for YouTube and 16.0bn for Facebook that occurred in the same month (based on a report shared by Digital Native on Substack). The answer is compelling: when asked which technology had been “the most revolutionary”, 57% of respondents to a recent YouGov survey cited the internet, versus 27% for smartphones and only 9% for AI. That said, were every iPhone ever sold stacked, the metaphorical tower would extend 250,000 miles, the equivalent distance to the Moon and a third of the way back, based on calculations at NYU.

The reality is that mobile devices account for two-thirds of the world’s internet traffic. Around the globe, some 3.5bn people use them each month. However, while internet penetration stands at 92% in North America – with the average person spending five hours a day staring at their devices – penetration in sub-Saharan Africa stands at just 34% (per the World Bank). In other words, there is a long way to go.

Two subsequent related questions naturally occur to us: how do we secure all this data – for data have zero value without being secured – and how do we build the appropriate infrastructure to support the data deluge? These topics inform our next two sections.

CYBERSECURITY

• Season Six Blog Posts:
  Coffee with the CTO, 2024 edition (29 May); “Human-like decisions, but better” (17 July); Post 39: Hanging out with the cyber experts (8 October 2024)

• Relevant Theme Pieces:
  Watch out! The growing privacy invasion and cybercrime threat (April 2014); Cybersecurity: the next generation (September 2017); Cybersecurity: The more things change…(September 2024)

• Key Statistic:
  Cyberattacks in 2024 will cost the world over $9tr, or more than 20 times Apple’s last reported revenues (Statista)

““Cyber is the most dangerous weapon in the world – politically, economically and militarily”

Bob Gates, former US Defence Secretary

We need to secure the cloud. Remote infrastructure will store half the world’s data by 2025 while more than 70% of workloads will run off cloud infrastructure by the end of the decade (per Gartner). Within the last three years, there has been a fourfold increase in cyberthreats. This year cyberattacks may cost the world $9.2 trillion (according to Statista). For perspective, that’s about onethird of the United States’ GDP or 24 times Apple’s annual revenue in 2023.

Other statistics make for sobering reading. The US lost a record $12.5bn to cybercrime last year, a 22% annual increase. Included within this figure is a loss of $1.3bn on the part of individuals from scammers pretending to be either technicians or from government agencies. More than 70% of firms with annual revenues exceeding $5bn have been attacked in the last year. Ransomware cases rose 300% in the second half of last year relative to the equivalent period the year prior, with cybercriminals pocketing more than $1bn in the process (data from the FBI, KKR and Visa respectively).

Against this background, cyber-security is consistently cited as the top priority for enterprise tech spending. 83% of executives polled by Lightspeed gave this answer, and 66% to Gartner. Further, more than 70% of business leaders believe that generative AI may increase the risk of cyber breaches (this figure courtesy of PWC).

Even if human error remains the number one reason why cyberattacks occur, spend on cybersecurity is set to rise almost 15% in 2024 relative to 2023 levels, per Gartner. At the same time, this may be just a fraction of the spend truly required. McKinsey estimates that the potential addressable market for cyber spend could be up to nine times the level of current investments.

GRID INFRASTRUCTURE

• Season Six Blog Posts:
  The transition mission (21 February); Hot times (7 August)

• Relevant Theme Pieces:
  What happens when the lights go out? (May 2011); What if the sun always shone? The coming energy storage revolution (September 2015); Keep the lights on. Make the grid smarter. (September (2022)

• Key Statistic:
  Global data centre electricity demand in 2026 will be equivalent to Germany’s current total power needs (International Energy Agency)

“The macro demand for electricity is picking up [with AI]… we’re seeing more demand than we’ve ever seen”

Earl Austin, Chief Executive of Quanta Services

Grid modernisation should be front and centre for policymakers globally since it is critical to support economic growth and digitalisation. Grids and their potential compromise are also a matter of national security. Remember, most electricity transmission systems in the developed world were built in the 1950s and 1960s. Given that they were typically designed with a 50-year life expectancy, it is not surprising that they have now “reached or surpassed their intended lifespan”, per a report issued by the American Society of Civil Engineers in September 2020. The problem will only have deteriorated since then. Building new grid capacity can take between 5 and 15 years, depending on the location and complexity, according to the International Energy Agency (IEA).

The growth in all things AI will only exacerbate the problem. A typical ChatGPT query requires ten times the electricity of a Google query. Or, put another way, AI requires four times the power and infrastructure of traditional server farms, increasing the strain on electricity generation (per Berenberg). As a result, power consumption in data centres could more than double between 2022 – before ChatGPT was launched – and 2030. For further context, data centres consume between 10 and 50 times more energy per square metre than does typical commercial office space, based on work conducted by McKinsey.

Data centre energy consumption is driven by IT equipment, which accounts for 40-50% of energy demand and encompasses servers, storage systems and network infrastructure. Cooling systems account for another 30-40% of energy consumption and auxiliary components (uninterruptable power, security etc.) for the remainder. With rising global temperatures, the need for cooling solutions is set to soar. The IEA estimates that the global demand for space cooling will more than treble over the next 30 years. Data centre needs could account for about over 30% of global electricity demand by 2050.

Meeting this demand will not be easy. The IEA estimates that 80m kilometres of power grids worldwide must be added or upgraded by 2040, nearly double the entire existing global power infrastructure. This will take time and cost money. At present, the total capacity of power projects waiting to be connected to the US grid is 30% higher than a year prior, with wait times for connections currently ranging 40-70 months, according to Goldman Sachs. The American Society of Civil Engineers believes that the cost of upgrading the US grid (generation, transmission and distribution) could reach $200bn by decade-end, while BloombergNEF estimates that the global cost could reach $2tr over the same period.

The next question which logically follows is where is all the electricity going to come from? In our view, there can be no transition without transition. Read on.

ALTERNATIVE ENERGY

• Season Six Blog Posts:
  The transition mission (21 February); Winds of change in Spain (26 March); Sunny times (sort of) (14 May); Make Euros while the sun shines (26 June); Hot times (7 August); Watt about the hot rocks? ( 23 Oct0ber 2024)

• Relevant Theme Pieces:
  Winds of change: the growing case for renewables (March 2018); Solar Energy: Everybody loves sunshine (January 2020); Winds of change: offshore edition (September 2023)

• Key Statistic:
  Solar could become the world’s largest primary energy source by 2050 (International Energy Agency)

“The key point is diversity in your [energy] mix… [and] the world cannot decarbonise without access to wind”

Henrik Andersen, Chief Executive of Vestas

Only one country had a net zero goal in 2015. Over 100 did at the end of 2023. If these targets are to be met, then not only will countries need to upgrade their power transmission and distribution lines, but also continue to invest in energy transition technologies. The good news is that global spending in this area rose 17% year-on-year in 2023 to hit a record $1.8tr. This is double 2010 levels, according to BloombergNEF (BNEF).

Even more encouragingly, global investment and spending on energy transition technologies surpassed investment in fossil fuel supply by $671bn in 2023, an increase from the $508bn gap in 2022. Further, for the first time ever, zero-carbon sources made up over 40% of the electricity the world generated in 2023 (again, per BNEF). This figure includes energy sources such as nuclear and hydro, but the share accounted for by solar is growing at a faster rate than any other energy input.

The IEA describes solar as “the fastest growing energy technology in history.” On its analysis, 413GW of solar was installed globally in 2023, up 58% versus 2022. Worldwide solar capacity has doubled in the last 18 months. Whereas it took it took the world a year to install a gigawatt (GW) of solar capacity in 2004, this had fallen to a month in 2010 and a week by 2016. In some days in 2023, a full gigawatt was installed in one day. BNEF forecasts 520- 655GW of installations in 2024. The technology is improving while costs are falling. Lazard estimates that since 2009, the cost of solar power has fallen by 83%, while the cost of producing wind power has fallen by more than half.

Against this background, the wind sector has also seen impressive growth. In 2004, 8 gigawatts of wind power were installed; Last year, the figure was around 110GW, including 12GW of offshore wind (data per BNEF). Just as the average solar project has seen a sixfold increase in size over the last decade, so wind projects are growing in scale too. The latest series of turbine blades being manufactured by Vestas will be taller than the Statue of Liberty.

Project forward and the IEA expects renewables to account for 47% of the global power mix by 2030. Crucially, renewables will account for 80% of new power generating capacity additions through to 2030, led by solar. The growing importance of solar can be seen through assumptions that energy from this source will overtake nuclear by 2026, wind by 2027, dams in 2028, gas-fired power plants at the decade-end and coal-fired ones by 2032. By the middle of this century, solar could become humankind’s largest source of primary energy.

This is exciting, but solar (and wind) should be considered as necessary but not sufficient for solving the world’s energy needs sustainably. Battery storage technologies – since both the sun and the wind are intermittent sources – need to improve markedly. BNEF estimates that if countries globally are to hit achieve their net zero targets by decade end, then investment levels will need to triple over this period, relative to current spend. Countries may also need to think about broadening their alternative energy portfolios and including other potential sources such as geothermal energy.

WATER

• Season Six Blog Posts:
  “Water, water everywhere…” (10 September)

• Relevant Theme Pieces:
  The curse of Coleridge: global water shortages (June 2011); Liquid data: digitalising the water sector (February 2022)

• Key Statistic:
  It takes 1,500 gallons of water to make one semiconductor chip (Xylem)

“Water is increasingly complex – [there is] either too little or too much. Solving water has never been more urgent”

Matthew Pine, Chief Executive of Xylem

When you consider that it takes 1,500 gallons of water to make just one semiconductor chip, then it should be evident that water is a critical component of the digital ecosystem. Global semiconductor sales are forecast to double over the next decade, fuelled by the data deluge we have previously discussed (statistics from Xylem and IDC). More crucially, without water, humanity cannot survive. And there simply isn’t enough.

The World Economic Forum estimates that there will be a 40% gap between water supply and demand by 2030. Other data from the United Nations suggest that over 2bn people lack access to safe drinking water. Meanwhile, more than half the global population experiences severe water scarcity for at least part of the year, according to the World Health Organisation. Populations affected by water scarcity could double between now and 2050.

It is easier to improve water supply than curb demand, in our view. Outdated water-management infrastructure is already causing global economic losses of approximately $470bn annually. By mid-century, water risk could wipe out $5.6tr from global GDP. Against this background, over $6.7tr of water infrastructure investment is required by 2030. This figure rises to $22.6tr by 2050, on World Bank projections. It seems clear that governments need to take action now.

FOOD INNOVATION

• Season Six Blog Posts:
  What we’re going to be eating in 2024 (10 January); My type of fish (4 April); The rise of the vertical farm (2 May); Restaurants of the future (20 June); The lunchtime robot revolution (27 August); On coughs, chocolate and compost (5 September); Sightseeing with salmon (3 October 2024).

• Relevant Theme Pieces:
  You are what you eat: health, wellness and food innovation (October 2014); Alternative meat: beyond impossible (August 2019); Food’s future: how to feed the world more intelligently (October 2021).

• Key Statistic:
  Food tech venture capital investments have grown 40-fold in the last decade (Nesta).

“We are at the tipping point of how food systems work. Sustainable food is not just about protecting the planet, but about creating healthier, more inclusive food systems for future generations”

Emmanuel Faber, former Chief Executive of Danone

Agriculture might be a logical place to begin if you wanted to consider curbing water consumption. The industry accounts for ~70% of global water demand. It needs cheap water in order to produce affordable crops. The World Water Council, an international non-governmental organisation, estimates that it takes around 1,000 litres of water to produce just 1kg of wheat. Almost 13,000 litres are used on 1kg of beef, and more than 20,000 litres are required to generate a kilogram of coffee. Furthermore, nearly a quarter of freshwater resources are lost due to food waste

1bn more people will need to be fed by 2050, resulting in 50-70% higher food demand, according to the United Nations Food and Agriculture Office. At the same time, some 2bn people today lack access to safe nutrition. How to match demand with supply constitutes one of the major challenges facing the world, particularly when environmental considerations are taken into account. Animal farming alone uses over 25% of all land, equivalent to the size of the Americas. The food industry is also the single largest driver of biodiversity loss, says The Circularity Gap Report.

It is clear that we need more sustainable solutions. 84% of consumers list sustainability as “very important” when making a purchase, per a recent McKinsey study. More fish might be one potential solution. Its carbon footprint and water usage is markedly lower than for other comparable sources of animal protein. Although 70% of the Earth’s surface is covered by the oceans, fish currently accounts for just 7% of all protein sources produced for human consumption. Fish farming is, however, the fastest growing form of food production in the world. Global fish consumption is therefore set to expand by 15% over the ten-year period to 2030 (data from the United Nations and the OECD respectively).

Go a few steps further and consider the potential of plant-based solutions, alternative meats and even lab-grown products (your author has tasted and enjoyed all the above, including salmon from a laboratory earlier this year). Plant-based beverages already account for 15% of all milk sales by value in the US and 11% in the EU. For context, the global dairy industry is worth $900bn. Expect further share gains. The alternative meat market is currently growing at three times the pace of the conventional meat market. There is scope for expansion here too. Meanwhile, the cell-cultured meat market could reach $6.9bn in value by 2030 (statistics from the Good Food Institute, Kerry Group and Grand View Research respectively). Venture capital investment into food tech has increased 40-fold over the last decade, per Nesta. More innovation should mean more novel products.

HEALTHCARE

• Season Six Blog Posts:
  Hot topics in healthcare (18 January); “The job is far from done” (12 March); On coughs, chocolate and compost (5 September); Wonder drug, or, just too good to be true? (19 September)

• Relevant Theme Pieces:
  Fat profit potential: bulging bellies and growing obesity (April 2012); Reinventing healthcare and the coming age of personalised medicine (November 2012); Longevity science: Who wants to live forever? (October 2019); Alzheimer’s: The final healthcare frontier (April 2024)

• Key Statistic:
  To treat all obese Americans with GLP-1 drugs such as Wegovy could cost $1tr annually, or 4% of US GDP (JP Morgan)

“The need to treat diabetes and obesity is massive and unmet”

Lars Freurgaard Jorgensen, Chief Executive of Novo Nordisk

While it is shocking that 690m people suffer daily from acute hunger, it is also highly concerning that every tenth adult in the world suffers from obesity. More than 1bn people, including 7% of girls and 9% of boys are now classified as obese. In 2019, it led to around 5m deaths, 20 times as many as malnutrition did. The number of people with obesity has doubled since 1990. By the end of this decade, more than 3bn people globally could be classified as either overweight or obese (all data per the World Health Organisation).

Obesity is a major problem for healthcare systems. 80% of their cost burden goes on managing chronic diseases. And if patients experience either diabetes or obesity (the two are highly correlated), then their odds of suffering from a co-morbidity are high. In the US, obesity is associated with ~40% of cancer cases, according to Cornell University. Those with type-2 diabetes (a function of lifestyle choices as opposed to heredity) have a 30% likelihood of suffering from cardiovascular disease and a 40% chance of experiencing chronic kidney disease, based on data shared by Novo Nordisk.

An even bigger challenge for healthcare systems is that fewer than 40% of people with diabetes are currently being treated, while the figure drops to just 2% for obesity. With ~3% of global GDP and ~8% of global healthcare budgets are spent on treating obesity today (according again to Novo Nordisk), the need for solutions should be evident. In the UK, the NHS reports that obese patients cost the system roughly twice as much as their non-obese counterparts. The Milken Institute in the US calculates that the overall cost of obesity to the US economy (if lost productivity is included too) could amount to $1.7tr.

The growing availability of GLP-1 (glucagon-like peptide) drugs such as Wegovy and Zepbound for helping people manage obesity has dominated headlines over the past year. While by no means a panacea, they can play a clear role in helping solve the growing obesity burden. More than 70% of people polled by McKinsey say they want to be healthier, and 50% rank it their top priority. It should come as no surprise then, that some 45% of US adults say that they would be interested in taking a safe and effective prescription weight-loss drug. Prescriptions for GLP-1 drugs in the US grew 300% between 2020 and 2023, albeit from a low base. Project forward and they are expect to see compound annual growth of over 50% through to the end of this decade. Correspondingly, medication sales in this area could reach $80bn by 2030, according to estimates from JP Morgan.

Benefit consultants estimate that 30-40% of employers already have plans in place to cover drugs such as Wegovy for obesity treatment. This figure could rise to 60-80% in the next few years, using the same research analysis from JP Morgan. While the cost of these drugs will inevitably fall as production scales and more generics become available, the current price point (of between $300 and $1300 a month) means that GLP-1s alone cannot be the main solution for the world’s obese people, for they cost too much. To treat every obese American might amount to $1tr annually, or 4% of US GDP.

Technology might provide the answer. Healthcare spend on AIrelated hardware is set to triple over the next five years, while over $30bn of new equity funding has gone into start-ups seeking to leverage AI to help solve healthcare challenges over the last three years (per Research & Markets and CB Insights, respectively). In case you missed it, in 2023, ChatGPT passed all three parts of the US Medical Licensing Exam. This accomplishment was a defining moment for AI’s role in healthcare as it proved that a general-purpose AI model could demonstrate human-level complex medical reasoning. Look forward and AI could become increasingly helpful in respect of not only drug development but also medical diagnostics, predictive analytics and robotic surgery.

HONOURABLE MENTIONS

There is, sadly, never enough space in this publication to cover all the topics that deservedly merit attention. Below follows a brief comment on three other themes that we continue to follow with interest.

• Humanoid robotics:
  Walk into the most modern BMW car factory in California and you may see a two-legged, fivefingered robot working alongside human counterparts. Robotic performance has improved 30- fold in the last decade and 2025 might prove to be the year when humanoid robots will become increasingly viable (and visible) within factory settings. Such robots can help not only improve productivity but address shortages in manufacturing labour. An additional use-case might be to help satisfy elderly care demand. Some analysts (such as Goldman Sachs) believe that the humanoid robot market could be worth $6bn by 2035. Watch this space. Our first theme piece next year will explore this topic in more detail.

• Autonomous vehicles:
  If the future seems bright for humanoid robots, then prospects for autonomous vehicles look bleaker, in our view. While we have written extensively about the importance of both electric vehicles and transport as a service, the reality is that the industry has grown far more slowly than experts had initially predicted. The auto industry remains driven (no pun intended) more by cyclical factors than secular ones. We are still in an autonomy winter. Advocates, of course, remain. Waymo (a business backed by Alphabet, Google’s parent company) says that its autonomous vehicles successfully completed 100,000 paid rides a week for the first time this summer. Further, Waymo says that in over 3.8m miles driven without a human behind the steering wheel, there have been zero bodily injury claims, Humans typically incur 1.1 for every million miles driven. The expectation (or hope) of some, such as BCG, is that 25% of all new cars sold by 2035 could be autonomous.

• Plastic waste:
  On a different topic, plastic waste remains a subject close to our heart and a priority for the world to address. The statistics are shocking: 1.2m tons of plastic waste, or over 80% of all that is produced, finds its way into the oceans annually. Every second of the day, 20,000 plastic bottles are sold. Fewer than 15% of these are currently recycled (all data per the OECD). The good news is that sustainability is becoming an increasingly significant priority for both manufacturers and consumers. The latter are pressuring brands to clean up their act. Retailers currently account for ~25% of global carbon emissions. Scientists estimate that the textile industry alone contributes to ~35% of microplastics found in the ocean. Hopefully by the time we produce next year’s ‘how we see the world changing’ piece, there will have been tangible improvements in all the above metrics.