Cryptocurrencies and personal AI: Tech to watch out for in 2025

At the end of 2022, the outlook for the cryptocurrency industry was bleak. One prominent company, FTX, collapsed, leaving as much as $8 billion (£6.3 billion) of customer funds unaccounted for. In March 2024, the company's co-founder, Sam Bankman-Fried, was sentenced to 25 years in prison for defrauding customers and investors. This scandal shook confidence in the entire industry.

Cryptocurrency seemed destined to remain a niche product with a passionate but relatively limited following. However, just months later, the industry is once again brimming with optimism. This optimism is fueled by Donald Trump's victory in the November 5th presidential election. It is widely believed that he will be more favorable to the cryptocurrency industry, and so far, this appears to be the case.

In early December, Mr. Trump indicated that he would nominate former US Securities and Exchange Commission (SEC) commissioner Paul Atkins to the top chair position at the Wall Street regulator. Mr. Atkins is considered to be more supportive of cryptocurrencies than the outgoing head, Gary Gensler. This news helped push the price of Bitcoin, the largest cryptocurrency, past the $100,000 mark.

Standard Chartered's global head of digital asset research, Geoffrey Kendrick, said: "With Trump winning, you could imagine that in 2025, there will be positive regulation. Some of the negative regulation will be unwound, and that will allow banks and other institutions to get into this space." Mr. Kendrick specifically pointed to a guidance issued by the SEC called SAB 121. In effect since 2022, it has made it difficult for banks and other financial firms to offer cryptocurrency services. This move could help Mr. Trump fulfil his pledge, made in July, to make the US the cryptocurrency capital of the world.

If he delivers on this promise, it will be a stark contrast to 2021 when Mr. Trump described Bitcoin as a "scam". As AI tools enter our phones – with Apple, Google and Samsung all launching services that can edit photos, translate languages and carry out web searches – we are at the beginning of an era where AI becomes an intrinsic part of our digital lives and increasingly helpful on a personal level. That is, of course, if we allow it to, as it does require a degree of trust.

Take diary management, for example. If you allow an AI tool access to your diary, it can effectively manage your schedule. But how far should this go? To be truly useful, does it also need to know who you would rather avoid seeing, or relationships you want to keep secret? Do you want it to provide you with summaries of consultation meetings or medical appointments? This is all very private information that could be very embarrassing and valuable if a glitch led to it being leaked. Do you trust big tech companies with this kind of data?

Microsoft is pushing hard in this area. The company got into hot water in 2024 for demonstrating a tool called Recall, which took a snapshot of a laptop's desktop every few seconds to help users find something they had seen but couldn't remember where. The company has now made multiple changes to the product, which was never released, but is still persevering with it. "I think we are moving into a new era where there are going to be ubiquitous, persistent, very capable co-pilots in our daily lives," Microsoft's AI chief, Mustafa Suleyman, recently told me.

Despite the challenges, Ben Wood, chief analyst at tech research firm CCS Insight, predicts more personalized AI services in 2025. "The output will be continuously updated by leveraging an evolving source of data such as emails, messages, documents and social media interactions. This will enable AI services to be specifically tailored to an individual's communication style, needs and preferences," he says. But Mr. Wood also acknowledges that giving AI unfettered access to your personal information will be a big step. "Trust is paramount," says Mr. Wood.

As more money pours into AI, there will be a need to build more data centers. Training and running AI requires vast amounts of computing power, and it is best done using the latest computer chips and servers. In the next five years, the biggest data users, including Google, Microsoft, and Meta, are likely to invest as much as $1 trillion in data centers. In Europe alone, data center capacity is expected to grow by an average of 9% each year between 2024 and 2028, according to real estate services firm Savills.

However, these new facilities are unlikely to be built in current data center hubs like London, Frankfurt, and Amsterdam. High property prices in those cities – Savills says land prices in London can be as much as £17m per acre – coupled with strained power supplies, mean that developers will be looking elsewhere. In the UK, cities like Cambridge, Manchester and Birmingham are likely to be the locations for the next wave of data center construction. In Europe, Prague, Genoa, Munich, Dusseldorf and Milan are also likely to be considered.

At the heart of these new data centers will be the latest computer chips from Nvidia, which dominates the AI chip market. The Blackwell chip, released in March 2024, is expected to start shipping in volume in 2025. Vivek Arya, a senior semiconductor analyst at Bank of America Securities, says the new chip should allow tech companies to train AI four times faster than with existing computer chips and run AI 30 times faster. Nvidia's biggest customers, Microsoft, Amazon, Meta and Coreweave, are likely to be first in line to get their hands on the technology.

But other customers may struggle to get hold of the superchip, with Mr. Arya saying "supply will be constrained in 2025".