Tech Forecasts 25: AI Warfare and Crypto Challenges

Edited by Mal Fletcher

The technologies shaping our immediate future are not distant dreams, but burgeoning realities. They're already taking root in laboratories and startups the world over.

This is part two of the "Tech Forecasts 25: Human vs Machine" report. It delves into some of the cutting-edge advancements poised to redefine our world very shortly, offering both a beacon of hope and a clarion call for vigilance.

In the event of a US-brokered deal regarding Ukraine, 2025 will bring an increase in cyberwarfare activities initiated by countries like Russia, China, North Korea and Iran.

These nations are already engaged in cyber attacks against the West, though thankfully not yet at the scale of shutting down core infrastructure on a large scale. This year we will hear more about unmanned submarines, underwater drones, cutting underwater cables and AI-powered malware interfering with elections, financial institutions, mobile networks and business activities.

Meanwhile, significant disinformation campaigns and reputational attacks will be launched against institutions and prominent public figures in free nations.

At the same time, AI might offer us new means of protecting against cyber attacks. This will be particularly true with AI that's powered by quantum computers. Some military experts expect to see the first successful quantum-based cyber attack by 2028. China is already investing heavily in this, aiming to achieve quantum supremacy soon.

Meanwhile, AI will drive the growing field of cyber espionage. The UN expects a 300 per cent increase in cyber espionage in the next five years. China is leveraging its huge technology resources to gain an advantage in this area. Arguably, it already has achieved this in the area of corporate spying.

Of course, artificial intelligence, quantum and blockchain technologies might offer some security in an increasingly dangerous world. Digital shields, powered by machine learning, might well adapt to new types of attacks. Countries like Israel and the US will be at the forefront of a cyber defensive revolution. The push for cyber defences will likely kick off a new type of AI arms race.

Machine learning will also help security agencies and businesses build more foolproof digital security systems. Yet no matter how sophisticated the technology, we will still need trained human eyes and ears to help spot inconsistencies and fraudulent activity online. In 2025 almost all machine intelligence systems will benefit from having humans in the loop.

The line between war and peace might well become increasingly blurred in the digital age. Free societies might need to adapt to a new normal in which a fairly persistent state of low-level conflict exists, with intermittent threats to services and infrastructure. Digital communications and banking services would be among the first targeted in major cyber attacks.

For this reason, among others, we must exercise caution when it comes to Central Bank Digital Currencies (CBDCs). We should not be drawn into a complete abandonment of cash, which carries no personal data.

Thankfully, we can expect governments to discover the need for data literacy programmes in schools, universities and workspaces. Training in critical and analytical thinking skills helps people understand how their data is collected and used online - and who controls that process. It also helps us identify scams, deepfakes and false narratives. 

Technology alone won't save us from cyber threats or criminal activities. Human minds will need to be trained to correctly interpret what they see and hear.

AI-powered cyber attacks could significantly impact the economy, not just in terms of its metrics, but its very form and nature.
In 2025, some pundits will advocate a switch to digital currencies as a remedy for cost-of-living pressures and other financial crises.

Yet for all their promise of blockchain-based security, easy access and quick investment profits, private enterprise cybercurrencies are subject to wild swings in value. Their worth at any given moment is determined largely by the vagaries of human whim. History tells us again and again that the wisdom of the crowd is not always either wise or reliable.

In 2025 we'll hear that the remedy to this volatility is to launch a Central Bank Digital Currency (CBDC). This is the government-owned, controlled and backed version of Bitcoin et al, offering the supposed benefits of digital currency without the fluctuations in value. The idea of a digital pound has been floated by the Bank of England.

CBDC advocates often speak of it as the first step in launching a Universal Basic Income (UBI). This too will be a subject of conversation in 2025, especially as we move to higher levels of jobs automation.

The idea is that the government issues every adult individual with a base monthly income, perhaps several thousand pounds, whether they work or not. The individual is then free to augment the wage with work if they wish to - and if they can find it.

There are significant holes in this arrangement, which are beyond the scope of this report. I can offer this, though: while automation will certainly change the job market, there is no evidence that it will simultaneously block the introduction of new types of work for humans.

There are precedents in history. For example, the introduction of the printing press forever changed the face of publishing and gave birth to revolutions in social structure such as the Reformation and modern democracies. Along the way, it reduced demand for caligraphers, for example, but it created many new skilled occupations such as typesetting.

Moreover, human beings thrive on work and productivity in labour provides part of our sense of worth and identity. A UBI arguably robs people of the drive to work or set up businesses and other enterprises. It makes the individual too reliant on the state as one's provider, removing the freedom to succeed on one's own merits. Even so, discussions about a possible UBI will continue this year.

CBDCs make us over-reliant on digital systems. They set us up for disaster when a nation or sector of industry undergoes cyber attacks. In recent years we've seen several examples of outages that have prevented people from accessing their money online.

Going fully cashless would mean that we're at the mercy of digital connectivity for the very basics of life. Cash may be messy and inconvenient at times, but it has advantages. It doesn't carry with it personal data; its weightiness helps us understand how much we're spending; and we can still pay with cash even in the midst of data outages, natural disasters or cyber-attacks.

In 2025 governments need to do more to prevent a total elimination of cash in the economy.

In 2025, more hospitals will experiment with AI-powered surgical robots. Whilst not replacing human surgeons, bots can analyse important medical data, in real-time during surgery, much faster than humans. What's more, studies show that robots can perform certain types of surgery more precisely than human surgeons.

Schools will begin to investigate using AI-powered robotic tutors, which can tailor the teaching experience to individual student needs. These will become indispensable teaching aids, especially for students who experience learning difficulties.

Meanwhile, many teachers will want to incorporate AI-driven virtual reality experiences, perhaps especially for subjects like history and languages. Immersion in a virtual 3D environment will help students understand subjects more quickly.

In the coming year, we will continue to see developments in the field of nanorobotics. Nanobots are microscope machines, built from the atomic level up, which can be loaded with chemicals and injected into the bloodstream to identify and destroy harmful cells while leaving healthy cells intact. MIT has already developed nanobots smaller than human cells. These could revolutionize drug delivery within the human body.

We can also expect to see more so-called soft robots. Their design is inspired by nature. They are more flexible and compliant than standard robots. In hospitals and rehabilitation centres, they can adapt to a patient's movements, providing gentle support during therapy. Soft robotic exoskeleton suits can improve mobility in stroke patients by up to 20 per cent.

Collaborative robots (or cobots) will become our helpers in the workspace. The number of robots involved in industry is expected to grow by 12 per cent annually from here, with cobots leading the charge.

And then there are autonomous robots. The University of Oxford predicts that within five years the average British household will own at least three autonomous robots for mundane tasks. 

The development of autonomous robots will owe much to the production of AI agents. Agents are AI platforms that can automatically access and utilise many other tools without human prompting and use them to make autonomous decisions. Essentially they will allow us to set an overall goal and leave the machine to find the best ways to achieve it, from start to finish. This will no doubt rescue us from many time-consuming mundane tasks. 

AI agents will also become a major selling point for humanoid cobots, especially in the workspace. However, this level of automation will require that we pay special attention to human skills we do not wish to lose. We will also need to understand and mitigate the mental health impact of humanlike machines. 

One such impact is the famous “uncanny valley effect”. When we first interact with a machine that looks and acts like a human, our first reaction is mostly positive, if only because of the novelty. However, as the machine’s capabilities become more and more human-like, we feel a growing unease. That uncanny feeling can lead to anxiety and other mental health issues. As we grow more accustomed to the machine, our levels of comfort might increase again, but it seems inevitable that, for at least some humans, the uncanny valley will become a plane, with long-term consequences. 

Robot designers will need to study the human impact of their creations very carefully, as will governments. We need regional and international agreements on ethical standards for AI and robots, and regulations to back these up. 

The rapidly developing world of robotics also raises other important questions. How do we ensure that AI-powered robots make ethical decisions in complex situations? As robots become more integrated into our lives how can we trust them to keep our data private? Will collaborative robots be buffered against cyber attacks? How will we ensure that they remain subservient to human agency? In the age of robots, safeguarding will take on a whole new meaning.

Thankfully, major international organisations, such as the Global Initiative on Ethics of Autonomous and Intelligence Systems (IEEE), are already publishing design guidelines for AI robotics. They emphasise transparency, accountability, and the importance of human values in robotic systems. They will have their work cut out for them.

The year ahead will be one of breakthroughs with human implants.

The next generation of cochlear implants, one of the most prominent types of implants, will use AI algorithms to process and transmit sound. This will produce something very close to natural hearing experiences.

Meanwhile, new retinal implants will restore functional vision to people with certain types of blindness. They'll incorporate nanotechnology to create less invasive devices with more precise vision.

We're about to witness breakthroughs in smart prosthetics, too. These will see a merging of man and machine, which allows users to feel texture, pressure and temperature.

European researchers are developing biodegradable materials for temporary prosthetics, to reduce the need for multiple surgeries, especially in growing children.

Another type of implant we will hear more about is the spinal cord stimulator, which is used to treat chronic pain. By 2027, we could see AI-enhanced spinal stimulators that automatically adjust the level of stimulation they provide based on the user's movements and pain levels. They could provide much more effective pain management.

Of course, the type of implant we hear most about is neural implants, brain-computer interfaces that read and interpret signals directly from the brain. In the next couple of years, companies like Neuralink will produce neural implants as small as a grain of sand, using nanotechnology. This will dramatically reduce the invasiveness of implantation procedures.

We can expect to hear about implants created using biodegradable materials, which can dissolve safely in the body after they've served their purpose. This would reduce the number of surgeries patients need to endure.

The UK is positioning itself for a leading role in implant technology, bridging the gap between research and practical applications.

As with any rapidly developing technology, there are many ethical questions to be considered here, not least data security. As we rely more on implants, they and we will become potential targets for cyber attacks, through hacking and unauthorized access.

There are also questions about consent: how do we ensure that people have given informed consent to receive an implant? How do we ensure that the availability or cost of implants does not create a new technology gap, between those who have access to the technology and those who don't?

READ THE FULL REPORT