Like most people, there have been various points in my life where I have come into contact with the legal system. Whether buying a house or a car, dealing with matters related to home life, or handling business issues ranging from employment to intellectual property law, I have often had to rely on a solicitor or lawyer to interpret the law and advise me on the situation at hand. With AI beginning to cast a shadow over the legal field, I wonder whether there will come a time when I can simply describe a situation and receive clear, effective legal guidance in return. To explore this question, I decided to ask AI. The following is its response.

Me:   Why is the UK legal system so complicated that it requires expertise to provide advice?

AI:   UK legal system is complex because it has evolved over centuries, combining thousands of statutes with court decisions that interpret them. Legal systems also need to handle many different situations and exceptions, so simple rules often become detailed and technical to avoid unfair or inconsistent outcomes. As a result, understanding what the legal system actually means in a specific case often requires specialist knowledge. Some complexity is unavoidable in a modern society, but much of it also comes from layers of historical amendments and legal precedent accumulating over time.

Me:   As your powers increase will there be a point where you take over from the experts?

AI:   As AI becomes more capable, it may perform much of the technical work currently done by experts, especially in fields that involve analysing information and applying rules. However, experts don’t just provide knowledge—they also take responsibility, exercise judgment, and are accountable for decisions. So AI may replace many expert tasks, but whether it replaces experts themselves depends as much on trust, legal system, and accountability as on technology.

Me:   I have heard a lot about the next stage in AI development - Agentica - what is it and how could it be applied to the legal system?

AI:   “Agentica” refers to more advanced AI systems that can plan, use tools, and take actions over time rather than just answering questions. In legal system, such systems could handle tasks like legal research, drafting documents, managing cases, and even suggesting strategies or outcomes. This could make self-representation and access to legal help much easier, especially for simple cases. However, humans would likely still be needed for accountability, courtroom advocacy, and high-stakes judgment.

Me:   Put forward a case for applying you to the legal system, in particular for the general public.

AI:   The case for applying AI to legal system is mainly about improving access, cost, and understanding for the general public. AI could translate complex UK legal rules into plain English, guide people through procedures like claims or appeals, and help draft documents without needing expensive legal support. It could make legal help more widely available, especially for simple disputes where hiring a solicitor is often too costly. Overall, the goal is to reduce barriers to justice and make it easier for ordinary people to understand and use the legal system themselves.

Me:   Looking into a crystal ball how long before the ideas that we have discussed will become reality?

AI:   Here is the summary with possible stages of development realistic timescales added. The dates are approximate and not predictions because adoption will depend on technology, regulation, public trust and how quickly legal institutions adapt. The transition to an AI-enabled legal system should be gradual rather than sudden.

Over the next 1–3 years, AI will increasingly be used as an assistant for tasks such as legal research, summarising documents, drafting routine correspondence and explaining legal information in plain English. Lawyers and judges will remain responsible for all important decisions.

Over the next 3–7 years, AI is likely to become a more active adviser, helping analyse cases, identify relevant precedents, predict likely outcomes and suggest possible solutions. Human professionals will still provide oversight and make the final decisions.

Over the next 5–10 years, some routine legal processes may become largely automated, particularly where the rules are clear, such as simple claims, standard applications and administrative decisions. People should still have the ability to appeal to a human decision-maker.

Beyond 10 years, AI could transform access to justice by providing everyone with a personal legal assistant capable of explaining rights, preparing documents and helping people navigate legal processes. Lawyers would continue to play a vital role in complex cases, but routine legal support could become available to a much wider group of people.

Throughout this transition, the key principles should remain the same: AI should support rather than replace human judgement, decisions must remain accountable, and the legal system should become simpler and more accessible. The goal is not to remove lawyers, but to create a system where technology helps people understand and use the legal system more effectively.

Me:   Do you have any final thoughts?

AI:   The future of the legal system is not AI or people — it is AI helping people deliver better justice.

Many species reintroductions in the UK, including the Red Kite, Eurasian Beaver, and White-tailed Eagle, have shown promising signs of restoring biodiversity and rebuilding damaged ecosystems. Yet nature is a complex network of interactions, and even successful reintroductions can produce unexpected effects on other species. Conservation decisions are often based on limited data and observations made after changes have already occurred. What is needed is a new approach.

To understand how ecosystems behave, scientists often begin with simple models of interaction. One of the most common is the relationship between foxes and rabbits.

When rabbit populations are high, foxes thrive. Food is abundant, survival improves, and their numbers begin to rise. But this success does not last indefinitely. As foxes become more numerous, the pressure on rabbits increases, and their population begins to fall. With fewer rabbits available, foxes in turn begin to decline, and the cycle gradually resets.

At first glance, this pattern suggests a predictable rhythm in nature—almost like a natural balance that repeats over time.

But real ecosystems are not this simple.

Foxes do not depend on a single food source. They hunt rodents, birds, and amphibians, and they also scavenge carrion. They may even consume fruit and berries when available. At the same time, their survival is shaped by disease, habitat change, climate, and human activity.

As more species and interactions are added, the system stops behaving like a simple cycle and becomes a complex web of dependencies.

And in systems this interconnected, predicting the long-term effects of a single conservation decision becomes extremely difficult.

This is where AI could play a transformative role. By analysing vast amounts of ecological data, AI can begin to model the intricate relationships within ecosystems and predict the likely outcomes of conservation decisions before they are made.

But its power increases significantly when combined with the idea of a digital twin.

A digital twin of an ecosystem would be a living virtual representation of the natural world. It would be continuously updated using data from satellites, drones, camera traps, weather stations, acoustic sensors, and field surveys. Together, this stream of information would allow the AI to build a detailed picture of the environment—tracking populations of plants and animals, mapping their interactions, and monitoring changes in habitat and climate over time.

Rather than relying only on historical data, the system could then explore the future. It could simulate thousands of possible outcomes, allowing conservation managers to test different actions within the virtual ecosystem before any intervention takes place in the real world. The reintroduction of a predator, the restoration of a wetland, the planting of a woodland, or the removal of an invasive species—all could be explored safely in advance.

Crucially, the digital twin would not be static. As new data flows in from the real ecosystem, the model would be continuously refined. Each real-world outcome would feed back into the system, improving its accuracy and strengthening its predictions. Over time, this would create a dynamic feedback loop in which the AI becomes increasingly capable of understanding ecological complexity.

If we can build a digital twin of our ecosystems and use AI to explore the consequences of our actions before we take them we may finally begin to shift from reacting to environmental change to anticipating it.

The question is no longer whether such systems are possible. It is whether we can develop and deploy them quickly enough to make a difference before biodiversity loss reaches a point of no return.

My inbox is full of campaign emails asking for donations, volunteers, or other forms of support. The problem is that most of them fail to inspire action. They may be worthy causes, but their communications rarely make a convincing case for me to set aside a few hours or reach for my credit card. To separate the campaigns that truly deserve support from those that don’t, I have developed “The Supermarket Test.”

We are all busy juggling competing demands and priorities. As a result, campaign communications must be clear, concise, and immediately relevant to the interests of the audience they are trying to reach. But relevance alone is not enough. Effective communication must also inspire people to act. It must create a sense of urgency, importance, or personal connection that moves someone from simply reading the message to supporting the cause.

Picture a busy supermarket. Shoppers push overflowing trolleys through crowded aisles, shopping lists in hand, scanning shelves for elusive ingredients and trying to complete their weekly shop as quickly as possible. Now imagine approaching one of those shoppers and asking for a few minutes of their time to discuss an important issue that affects them. Your challenge is not simply to inform them about a campaign, but to persuade them to support it. In those few brief moments, you must explain why the issue matters, how it connects to a wider movement, why they should care, and what action you would like them to take. This is what I call the Supermarket Test.

A campaign passes the Supermarket Test if it can persuade someone to change their behaviour or take action in the middle of a busy, distraction-filled environment. If the message is too complicated, too abstract, or fails to connect with people’s immediate concerns, it will be ignored.

Consider a campaign aimed at improving public health by reducing the consumption of ultra-processed foods. You approach a shopper in a supermarket and, after gaining their attention, briefly explain the health risks associated with these products. If, after that short conversation, they put the item back on the shelf, the campaign has passed the Supermarket Test. It may seem like a high bar, but it demonstrates that the message was clear, relevant, persuasive, and capable of motivating immediate action. If I attempted this in my local supermarket, I have little doubt that security would soon be showing me the way out. Nevertheless, the principle still holds. The Supermarket Test can be applied wherever people are busy and their attention is in short supply.

Next time a campaign lands in your inbox, or you’re approached on the street by a fundraiser, ask yourself whether it passes The Supermarket Test. If it does, get involved. If it doesn’t, politely decline and move on.

AI is taking over many aspects of our lives, including creating works of art. Can this be true? Let’s go and knock on Alan Turing’s door and ask him.

Alan Turing, developed a test to answer the question: Can machines think? The test, called the Imitation Game, replaces the question with a practical test of intelligent behaviour. It involves a human judge who communicates (usually via text) with another human and a machine behind a wall. The judge asks questions and receives answers, and shape must determine which is the machine and which is the human. Although Turing’s test was about intelligent behaviour the same approach can be taken to test creativity. So let’s set up an experiment where you the reader are the judge, and you have to decide which of the following poems were written by AI.

Let’s start. At the end of each poem ask yourself: could this have been written by AI or a human? Here is the first one:

The poem captures the effect of Spring on the poet. Now for the next one:

This poem describes how the changes in the flora and fauna send signs that Spring has started. Finally,

Before you read on - stop for a few minutes and make your decision.

Which one was generated by AI? The first poem was written by Tom B, Year 6, who is showing early signs of being a fine poet. I hope that he is still writing. The last poem was by E E Cummings. He uses both form and format to depict spring as a gentle, precise, and creative force that subtly transforms the world, much like a hand rearranging a window display. The second poem was generated by AI, and its comes down to you and what it means to you. But I feel that the second poem lacks any depth compared to the other two.

It is nearly impossible to define what turns a few words laid out in a certain way into a poem. But a good poem should invoke an emotional response. The AI-generated poem put together words that describe spring; we don’t experience a moment in Spring that the others produce.

You may have picked the AI-generated poem; maybe not. But I hope that you agree that AI has a long way to go before it is labelled ‘creative’. If you are a budding poet, then keep writing; you will be ahead of AI for many years.

It is easy to look at the weather and conclude that climate change is happening. It seems to be getting either wetter or drier for longer periods of time. But drawing that conclusion would cause the philosopher Aristotle to shake his head in disapproval.

Let’s look at an example to find out what is frustrating Aristotle. We know that if it rains, the pavement gets wet. We walk outside one day and see that the pavement is wet; therefore, we conclude that it has rained. But this is where the argument goes wrong. The ground might be wet because we live next door to a keen gardener and their sprinkler was on, or another neighbour has washed their car, or worse, a water pipe leaked; there are any number of other reasons.

If we take a similar approach with linking climate change and the weather. We could follow the following reasoning:

• If greenhouse-gas–driven climate change is happening, we will see more extreme weather,
• We see extreme weather,
• Therefore, greenhouse-gas–driven climate change is happening.

But there could be other reasons. The climate can be influenced by solar changes, volcanic eruptions, ocean cycles, orbital variations and land-use changes. Climate scientists are well aware of this and take a different approach to avoid the mistaken argument.

Returning to the example. We walk outside and see that the pavement is wet. The possible causes are that it has rained, the neighbour’s sprinkler was on, or someone has washed their car. Rather than jumping to a conclusion, we should ask, “How likely is each possible cause?” Before seeing the pavement, our expectations are that it is fairly likely that it had rained, less likely that a sprinkler was on, and somewhere in between that a car had been washed. Then ask: How well does each cause explain the wet pavement? If it rained, a wet pavement is very likely. If a sprinkler were on, it would also be quite likely. If someone washed a car, maybe only part of the pavement is wet.

After seeing the wet pavement, you might reason; rain is still the most likely, because we live in a wet part of the country, and therefore it was already fairly likely and strongly explains the wet pavement. A sprinkler becomes possible, especially if you know that your neighbour is a keen gardener. Car washing becomes less likely unless we see other clues (like soap or a hose). This approach is not asking, “What could cause this?” but asking, “Which cause is most likely given both my prior knowledge and the evidence?”

Climate science does not rely on one heatwave, one storm or one flood to draw conclusions about the effect of climate change on the weather. Instead, they use: long-term statistical patterns, multiple independent lines of evidence, and quantitative physical models. They combine all of this evidence to give a probabilities that climate change is influneced by human activity compared to other effects. For example, since 2017 about 93% of extreme heat events studied showed that climate change made them more likely or more severe. For the 126 rainfall or flooding events studied, 56% found human activity had made the event more likely or more severe, for the 81 drought events it’s 68%.

To cut through the noise from the socials about weather and climate change why don’t you look at the global studies at Carbon Brief, where accurate science is carried out, and then make your own mind up. That would get a thumbs up from Aristotle.