Richard Warne, Senior Portfolio Manager, Copia Capital
It is hard to get through a week these days without hearing something new regarding Artificial Intelligence (AI). Whether it be regulation or innovation, praise or criticism, most people have had their say by now.
As an investor, it may have initially been hard to look beyond the providers at the centre of AI development. For a while, the group dubbed the ‘Magnificent Seven’, consisting of Apple, Microsoft, Amazon, Alphabet, Meta, Nvidia and Tesla appeared the place for unmitigated growth. These are all still well-performing stocks, but geopolitical events and subtle shifts in the business environment have left people questioning if they are indeed in the best place to see strong returns going forwards.
Who are the real beneficiaries of AI development
It has been said that if you look back over time, the Magnificent Seven all traded on their own distinct narrative. However, over the last two years, they have become increasingly homogenous, particularly when it comes to AI development. This makes it tricky for an investor to back the winning horse, assuming there is one.
Despite significant capital expenditure on AI innovation by these tech behemoths, it is not necessarily translating into direct additional revenue. If you consider any new smartphone for instance, it has significantly more AI technology built into it than previous models, but it’s not significantly more expensive; the provider is not seeing elevated revenue in return for its investment in that development.
Fears were also stoked earlier this year, when China’s DeepSeek revealed its capabilities, allegedly at a fraction of the cost going into Stateside investment. Though their spending figures now look understated, and any initial stress share prices suffered has been recuperated, it raises questions over the wisdom of relying too heavily on the Magnificent Seven.
As AI still has a lot of development to go through, we think there are other areas where growth opportunities could present themselves.
Powering AI
If there is any certainty around the future evolution of AI, it’s that further development will require power, a lot of it. When looking for potential winners in the AI revolution, the energy and utilities sector appears well-positioned to benefit.
According to Polar Capital, US power demand is forecast to increase six-fold between 2020 and 2040, and data centre electricity consumption could account for between 7% and 12% of the US total by 2028. Understandably, there are reservations about the sustainability – materially, financially and environmentally – of relying on fossil fuels as the power of choice to fuel this rise in demand.
Renewable energy is an option. If we consider the cheapest energy per wattage, solar would be the obvious choice. However, solar energy suffers from the limitations we see in renewables – primarily their dependence on seasonal and geographical factors. As a result, solar alone would be insufficient to provide the high, consistent baseload that AI data centres require.
Nuclear energy and AI
Nuclear energy is certainly gaining a lot of traction around the world as a strategic clean energy source. There are varying motivations for this, including the desire to transition away from fossil fuels, the US wanting to compete with China, and Europe’s need to become more self-sufficient in energy following Russia’s invasion of Ukraine. Only recently, the UK government indicated its commitment to the energy source, funding the Sizewell C nuclear plant to the tune of £14.2bn.
Meanwhile, the Trump administration has conceded that they are behind China in terms of energy production and has signed four executive orders aimed at quadrupling nuclear energy capacity over the next 25 years. Despite the rhetoric surrounding “drill, baby, drill”, nuclear is more likely to be seen as having the growth potential to bridge the gap with China.
There are, of course, security and safety concerns. Nuclear infrastructure is not something that can be rushed out without adequate safeguards implemented to ensure there are no incidents like Chornobyl or Fukushima. However, it is important to note that the technology has advanced significantly in the past 20 years.
In terms of its implications for AI, there are already indications that nuclear will be significant in powering any further AI innovation. Labelled a ‘world first’, last October Google ordered up to seven small nuclear reactors from California’s Kairos Power to power their data centres. Furthermore, this month, Meta signed a deal with an Illinois nuclear plant Constellation Energy to keep one nuclear reactor for twenty years to power AI and datacentres.
Nuclear energy is seen as a good provider of the clean baseload fuel required for AI development’s intense and constant energy consumption.
Looking ahead
The primary drivers in AI development and nuclear expansion will arguably be tethered to the Sino-US technology war, with both committed to significant spending on AI innovation. Early into his presidency, Trump pledged $500bn investment into expanding AI data centres. Even if the accuracy of the DeepSeek numbers is questionable, capital expenditure is likely to be substantial.
However, the situation develops, further energy capacity is going to be necessary in facilitating any significant AI innovation in the coming years. As the appetite for fossil fuels decreases and renewables remain limited, it is likely nuclear will play a significant role.
This article is intended for regulated financial advisers and investment professionals only. Copia does not provide financial advice. This information is not intended as financial advice and should not be interpreted as such.
Free at 10am on Tuesday 15 July? Join Tony Hicks and Pete Wasko for our second Quarterly Market Review of 2025! Register now >
20th June 2025
Who will be the real winners in the AI revolution?
Richard Warne, Senior Portfolio Manager, Copia Capital
It is hard to get through a week these days without hearing something new regarding Artificial Intelligence (AI). Whether it be regulation or innovation, praise or criticism, most people have had their say by now.
As an investor, it may have initially been hard to look beyond the providers at the centre of AI development. For a while, the group dubbed the ‘Magnificent Seven’, consisting of Apple, Microsoft, Amazon, Alphabet, Meta, Nvidia and Tesla appeared the place for unmitigated growth. These are all still well-performing stocks, but geopolitical events and subtle shifts in the business environment have left people questioning if they are indeed in the best place to see strong returns going forwards.
Who are the real beneficiaries of AI development
It has been said that if you look back over time, the Magnificent Seven all traded on their own distinct narrative. However, over the last two years, they have become increasingly homogenous, particularly when it comes to AI development. This makes it tricky for an investor to back the winning horse, assuming there is one.
Despite significant capital expenditure on AI innovation by these tech behemoths, it is not necessarily translating into direct additional revenue. If you consider any new smartphone for instance, it has significantly more AI technology built into it than previous models, but it’s not significantly more expensive; the provider is not seeing elevated revenue in return for its investment in that development.
Fears were also stoked earlier this year, when China’s DeepSeek revealed its capabilities, allegedly at a fraction of the cost going into Stateside investment. Though their spending figures now look understated, and any initial stress share prices suffered has been recuperated, it raises questions over the wisdom of relying too heavily on the Magnificent Seven.
As AI still has a lot of development to go through, we think there are other areas where growth opportunities could present themselves.
Powering AI
If there is any certainty around the future evolution of AI, it’s that further development will require power, a lot of it. When looking for potential winners in the AI revolution, the energy and utilities sector appears well-positioned to benefit.
According to Polar Capital, US power demand is forecast to increase six-fold between 2020 and 2040, and data centre electricity consumption could account for between 7% and 12% of the US total by 2028. Understandably, there are reservations about the sustainability – materially, financially and environmentally – of relying on fossil fuels as the power of choice to fuel this rise in demand.
Renewable energy is an option. If we consider the cheapest energy per wattage, solar would be the obvious choice. However, solar energy suffers from the limitations we see in renewables – primarily their dependence on seasonal and geographical factors. As a result, solar alone would be insufficient to provide the high, consistent baseload that AI data centres require.
Nuclear energy and AI
Nuclear energy is certainly gaining a lot of traction around the world as a strategic clean energy source. There are varying motivations for this, including the desire to transition away from fossil fuels, the US wanting to compete with China, and Europe’s need to become more self-sufficient in energy following Russia’s invasion of Ukraine. Only recently, the UK government indicated its commitment to the energy source, funding the Sizewell C nuclear plant to the tune of £14.2bn.
Meanwhile, the Trump administration has conceded that they are behind China in terms of energy production and has signed four executive orders aimed at quadrupling nuclear energy capacity over the next 25 years. Despite the rhetoric surrounding “drill, baby, drill”, nuclear is more likely to be seen as having the growth potential to bridge the gap with China.
There are, of course, security and safety concerns. Nuclear infrastructure is not something that can be rushed out without adequate safeguards implemented to ensure there are no incidents like Chornobyl or Fukushima. However, it is important to note that the technology has advanced significantly in the past 20 years.
In terms of its implications for AI, there are already indications that nuclear will be significant in powering any further AI innovation. Labelled a ‘world first’, last October Google ordered up to seven small nuclear reactors from California’s Kairos Power to power their data centres. Furthermore, this month, Meta signed a deal with an Illinois nuclear plant Constellation Energy to keep one nuclear reactor for twenty years to power AI and datacentres.
Nuclear energy is seen as a good provider of the clean baseload fuel required for AI development’s intense and constant energy consumption.
Looking ahead
The primary drivers in AI development and nuclear expansion will arguably be tethered to the Sino-US technology war, with both committed to significant spending on AI innovation. Early into his presidency, Trump pledged $500bn investment into expanding AI data centres. Even if the accuracy of the DeepSeek numbers is questionable, capital expenditure is likely to be substantial.
However, the situation develops, further energy capacity is going to be necessary in facilitating any significant AI innovation in the coming years. As the appetite for fossil fuels decreases and renewables remain limited, it is likely nuclear will play a significant role.
This article is intended for regulated financial advisers and investment professionals only. Copia does not provide financial advice. This information is not intended as financial advice and should not be interpreted as such.
Free at 10am on Tuesday 15 July? Join Tony Hicks and Pete Wasko for our second Quarterly Market Review of 2025! Register now >