Ahead of the Curves: Expertise - Don't Die of Ignorance
This discussion took place in June 2023 as part of the Ahead of the Curves series, in partnership with Stifel Europe. This phase of the programme focused on the subject of Experts and Expertise.
You can also listen to the podcast interviews which provoked this discussion here.
A large proportion of the UK’s economy and exports consists of the work of 'experts'. Our professional services sector has historically been strong and respected for its quality worldwide. But how much strategic thinking came down from the top about protecting and promoting these sectors post-2016?
The advent of artificial intelligence is likely to mean major change for these professions - those at the top of professional pyramids charge great sums for their services and if a convincing and reliable disruptor comes along there will be clients aplenty who are interested in giving them a try. However, if you’re engaged in a billion-dollar piece of M&A or some crucial litigation on which large sums rest, we are not at the point where a computer takes over and guides the proceedings. AI cannot read a room, it has no human experience in dealing with clients. Its widespread adoption would be far too risky.
But there’s no doubt that populist thought tends to do down those experts 'in the know'. Michael Gove struck a troubling chord when he announced “We’ve had enough of experts” during the EU referendum in the UK. We just have to see what happened to Dr Anthony Fauci in the States during the pandemic to see similar attitudes in the USA. Fauci is 83 years old with a lifetime of medicine behind him. From 1983 to 2002, he was one of the world's most frequently cited scientists in all scientific journals but that did little to protect him in the face of an onslaught of opprobrium.
People do not defer to men (or women) of medicine, the cloth or the law as they did a generation ago. That lack of respect for learning, science and facts over opinions can be troubling. The value of 'lived experience' and 'my truths' over considered facts can be very troubling. This doesn't mean, as Professor Angie Hobbs suggests below that we should be cowed by respect for expert authority: "Experts deserve respect but not complete deference. Claims of expertise should always be scrutinised and assessed and a real expert will be happy to be questioned and will hold their own."At its worst it’s quite backward – the whole point of the Renaissance and Enlightenment all those hundreds of years ago was to replace the falsity of superstition with reason and facts. One just hopes we’re not heading back into a dark age of unreason in which the likes of smart people like Galileo are denounced.
What makes an Expert and what about AI?
Eithne O'Leary kicked off the conversation: "Experts are there to help us make better and more informed decisions when we lack the knowledge or experience to do so ourselves. That’s a state of affairs with which I’m quite comfortable - If I don't know, I ask someone who might.
"However, in recent years, as social media has taken more space in society, everyone has an opinion and everyone has a platform from which to publish it.
"What has become less acceptable to suggest is that not all opinions are equally valid. I believe not all opinions on all subjects are of equal merit. We need to find a way to distinguish between what adds value and what adds confusion and adds strife. However, the world is complex and not always easily understood and we should be always in a position to be informed by other people and keep our minds open and not view that as a weakness.
"Our other concern is how the advent of AI will affect the role of the expert. For this country and for this economy where 80% of what we do is service-driven, we need to think more carefully about what’s happening to that expertise and what happens under a new economy which is more AI-driven. The tax base that were all involved in, is dependent upon the provision of services. This is not a topic that we discuss enough and we need to understand what is about to happen and what’s coming down the track".
Matthew Gwyther mentioned Nicholas Taleb’s aphorism about experts: An expert is someone who knows exactly what not to be wrong about.
Dr Tomas Chamorro-Premuzic began, "I have just pulled from the Economist book of quotations one by quantum physicist Niels Bohr: 'An expert is a man (or a woman, of course) who has made all the mistakes which can be made in a very narrow field'.
The second illustration doesn’t mention experts: specifically, it talks about technical people, see image by Scott Adams:
"I think the beauty of this example is that it highlights that as intelligence or the world of ideas becomes more complex and harder to grasp, expertise itself is likewise harder to judge. We can be easily distracted by style and not focus enough on substance.
The book I wrote: ‘Why do so many incompetent men become leaders and How to fix it’ talks about not just the rise of populism but the surplus of incompetent men in leadership roles and the baseline for poor leadership performance – we follow our gut/intuition and as the world becomes more complex we become lazier with less time to think. Add AI into the mix and you have a very explosive cocktail.
"In a way, populism is an antidote to ideology. An ideology is dogmatic, it’s not a very open system that invites reasoning or the assessment of facts. Populism is optimised for what the majority of people think. Generative AI is a great source to access the wisdom of the crowds or more often than not the ignorance of the crowds.
"AI is populist in a way, it gives us more of what we want and less of what we need. AI is gasoline to the fire. AI can inform and misinform with equal power.
"In the future experts won’t be able to bullshit as much, because when we say generative AI can hallucinate – that means bullshitting. There are a lot of people who make a very decent living bullshitting. There is the wonderful paradox of professional expertise: the more you get paid for what you do and the more prestigious your job is, the harder it is for others to work out whether you’re any good at it.
"We will be pushed by generative AI to demonstrate our true value. Law firms will still charge a lot of money and have fewer paralegals but if clients are smart they will use the generative AI and wonder why they’re paying so much".
Expertise, new technologies and hard truths
From a psychologist to an eminent paediatric plastic surgeon. Bruce Richard wondered about the movement towards moral relativism and a marketplace of ideas. "Strangely, even simple facts are now up for debate and we’ve lost a sense of absolute truth," he said.
"As far as my professional world is concerned, I try to help parents decide about operations for their child. Some of this process is based on my expertise but a heck of a lot of it is based on who I am as a person. Expertise should be based on outcomes. An expert should not have bad outcomes. Outcomes are largely measurable and measuring them should help you pick a competent expert from a bullshitter. How do you spot a bullshitting plastic surgeon? We have a lot of people who believe in falsity because they want to hear the story being fed to them.
"My world contains some poor and inexpert practice in terms of the motive for carrying out some operations. I was a programme director for training in plastic surgery for many years but I frequently tried to prevent some candidates from gaining entry to training if I sensed they had largely self-serving motives, for example, narcissists from becoming plastic surgeons. Such a weeding out for our political class might be a good idea.
"I think we need to explore, even in the world of business, that expertise is not just a fact or knowledge but the character of the person who professes it and to whom they hold themselves accountable. Emotional intelligence is required".
Nick Moore explained that as a qualified medical doctor and now a banker "the intellectual approach that I take with my banking clients is the same as mine from medical school; you take a history, you examine, you diagnose, you treat. Taking that properly rigorous intellectual approach is incredibly helpful in what I do now".
Wendy Jephson suggested that there is a “marked difference between an open expert and a closed expert. Closed experts believe they have all the answers and they get stuck in a way of doing things. An open expert sees the ways that things have been done before, understands the challenges ahead and listens and continues down the route of learning to adapt their expertise."Thinking about AI/Chat GPT, the reality is that it’s currently being used for small tasks and it will be used for small tasks for a long time. What I hear is that it’s about 70% accurate and that is not accurate enough although, of course, it will continue to get better."Those who possess expertise are the ones that are getting the most out of it because they can see where it’s in error and they can correct it. Critical analysis is a skill that people in the future are going to need. The building of expertise is the development of schemas – seeing things multiple times in multiple ways".
Professor Angie Hobbs pointed out that there’s nothing new in fretting about change and expertise.
"The ancient Greeks were always anguishing about new technologies – the character of Socrates in Plato’s ‘Phaedrus critiques the invention of the art of writing because it reduces people’s capacity to remember. And you can’t question a book or have a debate with a book.
"As with all new technologies you won’t understand the potential for harm and benefit unless you have a picture of what flourishing human life looks like. Chat GPT is happily of no use to my students. There are certain specific tedious tasks I can give it to do but it can’t tell me the big questions to ask about the nature of the good life.
"Experts deserve respect but not complete deference. Claims of expertise should always be scrutinised and assessed and a real expert will be happy to be questioned and will hold their own.
"But we need to think hard about the notion of popular beliefs and the fact that the majority of people believe something to be true doesn’t make it true. In terms of objective truth, popular opinion won’t get you anywhere but it is so important to listen to what people are experiencing. The truth is not a matter of opinion and the popular phrase at the moment - ‘my truth’ - is a reflection of my emotional state and my belief at the moment.
"Experts need to listen and I don’t think we’ve always listened enough. We have to accept a bit of the blame for what’s been going on. People have said things which we know are not true and we haven’t understood why people want to hear things that aren’t true – why is what is true not wanted? We must accept the need to consider people's feelings and beliefs, even if they require correction".
Vicky Pryce explained, "I think most economists said Brexit would be a bad idea, most of those saying the opposite were simply saying what most people wanted to hear. Hard truths aren’t easy to accept. When Michael Gove said that ‘we’ve had enough of experts’ he really did not mean any other type of experts but specifically economists because we were arguing very logically about what the impact would be when the Leave argument was conducted very emotionally.
"Economists have got a problem because we are expected to come up with very detailed, expert forecasts of exactly what will happen to inflation, for example, to the decimal point. It simply cannot be done. We should rethink how we do it – the BoE tried to give forecasts to very detailed decimal points and then have had to say we made a mistake".
Jane McCormick suggested, "Expertise can only be labelled as such if it has concrete results. It’s about real outcomes. However, good outcomes come when we combine expertise. When experts from different fields combine their knowledge, then we see real leaps forward.
"AI has a hell of a long way to go, what it can’t do is fill in the gaps. Expertise allows us to make a judgement in a situation where we have a lack of knowledge, which allows for the best outcome. AI can recognise a tree but it doesn’t understand tree-ness".
Madison Kominski added, "I think most ambitious and hard-working people hope to be an expert in something one day, but I think that’s a very high bar. I’m comfortable with the introduction of AI as a disruptor and changing the landscape because it means raising the bar. It may mean the shape of my career may be different but it is making you challenge yourself".
There has been much discussion about AI removing laborious grunt work, for example, that done by young associate lawyers.
However, Wendy argued that “doing the boring jobs has value. Medics and lawyers learn the anatomy of the human body and of contracts. So, as you do the boring job of proof-checking a contract, you are subtly learning the meaning and interactions of the clauses - you see how people can interpret things differently. This is part of what creates an experience. It's important that we are questioning whether, as we automate certain things, are we losing something that helps to create expertise as people move through their careers".
Angie concurred, "To get the most out of this tool you need to have some understanding of the boring tasks. It needs the correct data to be fed in".
Simon McDougall explained, "Many of my peers are looking at what data was this AI trained on and what the issues are there. Trying to work out who will be replaced by AI is tricky – in robotics, it turns out that humans are difficult to replace. I was chatting with an AI expert and looking at AI hallucinations where the tech does something wacky and this person said you need to remember that all generative AI is a hallucination. Guessing sequential words, leveraging a massive amount of training behind it. Only when the hallucination goes in a way we didn’t expect do we call it one".
Expertise in Government
Next, the group discussed the value of technocratic/ meritocratic government run by "experts" as opposed to democratic ones.
Vicky explained, "There have been technocratic governments in Greece and Italy, and in several governments in Europe people get appointed to Finance Minster or other cabinet posts straight out of industry".
Simon thought, "With CEOs and senior politicians the two key competencies they need are to listen to experts and weigh them up and make not bad decisions quickly and continually".
Eithne agreed, "Listening is very important but in terms of making decisions, people underestimate the energy involved in making decisions all day long – you must have some self-awareness in the process. You must be logical, reasonable and consistent. This would help enormously on a political stage".
Tomas explained, "As somebody born in Argentina, a country that is outperforming every country in the world in terms of perpetual decline, this has happened mostly through democratic systems that were unable to sift out parasitic narcissists. Britain has first-world problems, go to South America to see the sophistication of these parasitic leaders".
Angie agreed, "I understand what Tomas has said – I want to hang on to democracy for a little longer. My argument for democracy wouldn’t be based on a belief that a majority view is true or morally right but it’s a human rights view that is based on the idea that each person has a right to a voice and right to be heard. How can we get a more expertly informed democracy? My huge worry is how we have a truthfully informed electorate in the face of the algorithms".
Eithne posed, "Social media doesn’t seem to have any requirements for telling the truth. One of my worst phrases of the moment is ‘lived experience’ as it seems to be an excuse to validate almost anything".
Wendy suggested, “We should consider who the audience is when you're presenting expert information and what we're expecting them to do with it. Taking Brexit as an example - experts were talking about all the consequences we are now seeing but there was a "bait and switch" that happened. When you have complex decisions to make about things outside your expertise people often switch the hard question for one they feel they can answer. For example: think about all the many potential consequences of Brexit and vote on what was switched for: do you want to improve our immigration control? The latter is much simpler to think about and calls for a much simpler yes/no answer. Sometimes people don't stop to consider complex information until they have taken a quick decision and are living with the consequences. This is exactly recognised by financial law which gives everyone a 14-day cooling-off period when they take out a credit card - knowing people might regret a decision that can have financial implications on their lives and allows a change of mind. It's extraordinary the same principle wasn't applied to something as significant as the Brexit vote. As we think about AI and with the knowledge that they can exacerbate this risk we at least know this a bit better and can hopefully regulate accordingly".
Vicky reminded us that "economists have always assumed that people behave rationally when they make their forecasts and then the politicians get in the way and mess this up. There is a deep issue that we have great difficulty forecasting how the consumer will behave. There are times when interest rates go up and the consumer behaves in a way that you just couldn’t have predicted. Using an AI system here based on historical data simply won’t work".
Simon added, "A big point to end on. I challenge what content is currently regulated on the internet and what is not. Copyrighted music on Facebook is picked up by the algorithm very quickly; child sexual abuse material on Reddit is picked up very fast by that algorithm – going back to when Google and Facebook would stand on stage and say they are neutral but there has always been selection in place. The ongoing dynamic, the online safety bill and the regulator that will follow will look at misinformation".
Contributors to the discussion included:
Dr Tomas Chamorro-Premuzic, Author: I, Human: AI, Automation, and the Quest to Reclaim What Makes Us Unique; Chief Innovation Officer, ManpowerGroup
Matthew Gwyther, Partner, Jericho Chambers
Angie Hobbs, Professor of the Public Understanding of Philosophy, University of Sheffield
Wendy Jephson, Founder, LetsThink
Madison Kominski, Investment Banking Analyst, Stifel
Jane Lawrie, Global Head of Corporate Affairs, KPMG
Jane McCormick, former Head of Global Tax, KPMG
Simon McDougall, CCO, ZoomInfo
Nick Moore, Managing Director, Stifel
Eithne OLeary, President, Stifel Europe
Vicky Pryce, chief economic adviser and board member, CEBR
Bruce Richard, Paediatric Plastic Surgeon
