Artificial intelligence: could it be magic?

Arthur C Clarke once said, ‘Any sufficiently advanced technology is indistinguishable from magic,’ in one of his futurist papers. Artificial intelligence (AI) certainly seems to have an element of the magic wand about it.

There is little doubt that the technology will be transformative. How to interpret its work product, in all its guises, will be a big ask. Where do we start?

AI has already landed

The risk profiling of AI is well underway, with reports on the risk and benefits to the legal profession published on an almost daily basis. Even if lawyers are reticent in their use of generative AI, clients may have already boarded the train and left the station. We are entering a time when the computer says not only ‘no’ but a whole lot more. Taking a witness statement in 2028 could be interesting, not to mention cross-examinations. The question as to ‘who’ made a particular decision might well be reframed as ‘which AI package’ did so.

Developing interdisciplinary lawyers

The interdisciplinary nature of legal work will be one of the most important aspects of working life in the future. Law firms are complex places and employ people with a variety of skills. This is not just idle musing. In October 2023, the Royal Society ran a conference exploring careers at the interface of science and the law.

This kind of expertise and, perhaps, dual qualification in whatever LegalTech becomes, is best developed in situ a lot of the time, rather than bought in or bolted on. Look around first, not just at the software, but the people within your organisations. What do they do and how. Who would want this job?

Challenging law schools

There will be a change, over time, to the training of lawyers. Even without ChatGPT and other LLMs, we have been on this road for some time. A book like How to Use a Law Library seems quaint.

However, there is one area of academic research that has started to move centre stage, especially for students who have had placements in finance and industry: research methods beyond legal research. This is a missing cog in legal education that we should consider in detail now and the management of primary data. At present, this kind of education is the province of, mostly, post-graduate research in qualitative and quantitative research the harder edged data interrogation and inference skills, working with primary data.

The Betamax effect?

What to buy? If we find ourselves just using 30% of an application, if that, it is probably not the greatest buy. Ubiquity and usefulness are not always the same thing.

However, the selection of the type of AI software could become a matter of contention when push comes to shove in the context of a negligence claim in future. The points made by Sir Geoffrey Vos in a lecture to the Professional Negligence Bar Association in May 2024 should be required reading before venturing into the area for the first time.

The place of intuition and expertise

It is easy to get caught up in the breathless excitement of all things AI. However, it cannot do everything.

Interpretative skills and creative flair are likely to be more important when ‘some machine’ is dealing with the more mundane aspects of legal research with little or no attentional error.

However, if you are about to fall into a great despond, there is solace to be found. Hubert L Dreyfus and Stuart E Dreyfus’ Mind over Machine, written in 1988 during the infancy of computing, has much to say about the importance of human intuition, creativity and expertise. ‘Computers are more precise and more predicable than we, but precision and predictability is not what human intelligence is about.’

Law adapts. The scriveners and the copyists disappeared over a century ago.

AI is not a magic wand. We are not going to switch the AI on and the lights off.

Copyright & permissions