Man or Mythos: AI, costs and beyond

Andrew Hogan’s new book Andrew and the Marvellous Analytical Engine arrives at a propitious moment. Grayson Perry has just returned from Silicon Valley — two episodes of primetime television that include meeting a woman who married her AI companion, a mind-reading startup harvesting neural data, and tech billionaires discussing the transformation of human life with the breezy confidence of men who have never lost a job. Perry is a thoughtful and good-natured presenter. He seems egoless, as one reviewer noted; but the series’ animating anxiety is hard to miss. The most well-known people building the future seem completely unbothered by what it will do to everyone else.

Spike Jonze got there first, and less politely. I had wondered before watching it recently whether his 2013 film Her (prescient in ways that continue to resound), was about AI and economic displacement, but not at all. It is about what artificial intelligence does to human intimacy and relations. For lawyers, a profession built on human judgement and trust, that may be the more pointed concern. And yet the film, like the Grayson Perry series, ends somewhere closer to enlightenment than despair.

Andrew has a different temperament entirely. His account of artificial intelligence and legal practice is measured, highly practical and well-earned. He has used the tools. He understands the professional duties of confidentiality that practitioners navigate at their peril when feeding client data into AI platforms, and his treatment of that terrain is worth the cover price alone. His optimism about the enduring value of human judgement is not complacency dressed as wisdom. It is the considered view of someone who has done the analysis and is prepared to show his working. We could call it the grey-haired exception. For Andrew the irreducible premium on experience and the ability to read a room survives largely intact.

But there is a question the book doesn’t press. Andrew’s vision of practice in 2030 is populated by solicitors and barristers who deploy AI-generated briefings and cross-examination plans as tools, refining and adapting them with the depth and subtlety that machines cannot provide. Juniors will still matter, he assures us; their training will simply be more deliberate.

What the book does not quite confront is how those juniors will become the grey-haired survivors if the work they used to start their careers on — the chronologies, the disclosure reviews, the first drafts built from scratch — has been automated away. The career pipeline is not a peripheral concern. For knowledge-based professions it is the most uncomfortable question the technology raises.

The Master of the Rolls, Sir Geoffrey Vos, addressed the Association of Law Teachers at Exeter on 16 April, and was rather more direct: reduced recruitment is already happening, he observed, and where firms are recruiting, they are looking for different skill sets. The career ladder problem is not a prediction. It is a clear and present threat.

For those of us in costs practice, there are also opportunities. The direction of travel in large and complex litigation is toward costs of a scale that would previously have tested any process of oversight. Diesel Emissions litigation demonstrated that there is, in practice, no ceiling on budgetable costs. The question that follows is whether very large costs can be assessed with comparable rigour, or whether the process becomes its own war of attrition. Sebastian Holdings v Deutsche Bank gave a cautionary answer that to many will stand as aversion therapy: a detailed assessment that ran for years and at a huge cost.

AI-enhanced analysis may be the bridge the costs courts need. It should enable forensic engagement with costs of a scale and complexity that would otherwise defeat any proportionate process. The MR, in the same Exeter lecture, identified personal injury damages assessment as an essentially algorithmic exercise once a machine has access to all relevant authority. That logic extends to costs. That is not merely an efficiency gain. It is a structural change in what justice in the costs phase can look like.

Grayson Perry ends his series getting quietly emotional about the case for optimism. I find I do too. Among his more memorable interviewees is Jack Clark, co-founder of Anthropic — the company whose AI chatbot, Claude, is name-checked by Sir Geoffrey Vos as one of the tools already doing the work lawyers used to do first. Clark is bracingly honest: bad times are coming, climate, demographic breakdown, the usual roster of civilisational anxiety. But AI, in the right hands, can bring a little joy in too. There will be good actors and bad actors, and we will have to cope with that. As Perry correctly observes, the toothpaste is out of the tube.

It is not a comfortable conclusion, but it is a balanced one. Perry also quotes a tech journalist’s observation that the typical Silicon Valley founder is technically brilliant and emotionally stunted — too rarely in contact with real people, too dominated by the wrong kind of male energy — and the observation lands because it identifies where the risks actually sit. And then, not entirely without menace, Perry offers his own verdict: AI will wipe out all the mediocre intellectuals. As someone who has spent a career struggling to stay on the right side of that line, I take the point. Andrew Hogan takes it too, and his book offers a toolkit for doing something about it.

At Practico, we are forming our own version of that toolkit and applying it: both to how we work individually, and to the collaborative layer that Andrew does not quite reach. A specialist firm of our size can do this in ways that once required the resources of much larger organisations. The democratisation of technical development makes that possible. I remain hopeful that some grey hair helps too.

Andrew and the Marvellous Analytical Engine by Andrew Hogan. Details of where to obtain the book are at themarvellousanalyticalengine.ai. Grayson Perry Has Seen the Future is currently airing on Channel 4. Her (Spike Jonze, 2013) is available to rent on Amazon Prime.

‍