As the famous saying goes, necessity is the mother of invention, and no period creates so fertile a ground for invention as war. The Second World War is a prime example, with a wave of technological innovation that birthed the world’s first electronic computer at Bletchley Park, the atomic bomb at Los Alamos, jet engines and superglue. What was also in abundant supply was a litany of heroes: men and women who rose to the demands of history to leave a legacy of sacrifice and success.

One of the lesser known of these was a German pastor, Dietrich Bonhoeffer, a devout Lutheran who campaigned tirelessly against the Nazi regime. His relentless protests against the extermination of the Jewish population ultimately led to his own internment in a Nazi concentration camp, and eventually, his murder. During his imprisonment, Bonhoeffer wrote extensively on several topics, including theology, but is well known for a particular letter he penned to a fellow clergyman, called his ‘Theory of Stupidity’.

By stupidity, he did not mean low intelligence in others; rather, his theory focussed on irrational behaviour, reflecting that actions and decisions are often based not on raw brainpower, but instead on a choice to surrender critical thinking to external forces:

“Against stupidity we are defenceless. Neither protests nor the use of force accomplish anything here; reasons fall on deaf ears;

…when facts are irrefutable, they are just pushed aside as inconsequential, as incidental. …under certain circumstances, people are made stupid or allow this to happen to them.”

Dietrich Bonhoeffer, ‘After Ten Years’, Letters and Papers from Prison

Whilst Bonhoeffer was talking of perhaps weightier matters, I believe that his observation about the nature of human irrationality has profound implications for the determined value investor – implications that are especially relevant today in the age of artificial intelligence (AI).

All I gotta do is act rationally

As value investors, we don’t spend our days trying to be much cleverer or quicker than other investors. Rather, we focus on what we can genuinely do best: be patient, take our time, remain grounded in the historical facts and – above all else – behave rationally. The bread and butter of a value investor’s opportunity set comes from an ability to remain rational, when others around you lose their sense of rationality.

Whilst it may be uncomfortable for finance academics to hear, participants in the stock market behave irrationally every day: they buy or sell securities – thus, pushing their prices up or down – for a whole host of reasons, some sane, and some less so. Occasionally, they sell pieces of a company at prices that reflect an absurdly low valuation for the business in question, driven by fear, risk constraints, or a blinkered focus on the short-term. When this happens, long-term value investors such as us – who view listed shares as part-ownership stakes in genuine businesses and price these shares accordingly – can stand by to take advantage.

Take, for example, our good friends in multi-strategy hedge funds, or ‘pod shops’: they are driven by their need for rapid portfolio turnover in order to eke out daily or weekly profits, all the while risk managing their positions with a fine-tooth comb to remain ‘market neutral’. To a large degree, they have had success in doing so: their approach cannot be dismissed as irrational from the perspective of one trying to squeeze steady, uncorrelated returns out of a market, with a willingness to employ huge leverage to amplify slight price movements. This activity is what they are being paid[1] to do by people who, perhaps, need constant access to their funds, and an assurance that they will never lose more than 2% of their capital. For others, however – who allocate their endowments or funds with a patient, long-term mandate – such frenetic trading activity could certainly be deemed irrational.

As long-term investors, we are quite simply playing a very different game to the multi-strategy hedge funds, thinking beyond the minute ups and downs of the everyday, and doggedly focussing on the intrinsic business value of a company. Where we think that we can buy a stake in that company for a material discount to that business value – a discount we term the margin of safety – we are willing to do so, owning a large chunk of a company for years on end whilst value compounds. Pod shops are likely to sell us shares on any given day at a large discount to their true worth, because they have no interest or ability, given the nature of their strategies, to hold onto their investments until true business value is realised. In that scenario, whilst it may make sense to the pod shop to sell to us, from our perspective – a through-the-cycle business owner looking to pay cheap prices – that decision is irrational, making us happy buyers and, hopefully, earning us an outsized return.

On the other side of the behavioural ledger – and no less profitable for the value investor – are those market participants who make irrational long-term decisions not because they have a different strategy, but simply because they are allowing their emotions to dictate their behaviour, rather than their reason. As per Bonhoeffer’s Theory of Stupidity, this rarely has anything to do with intelligence; rather, it is a willingness to suspend critical thinking in order to follow through with a particular aim, or to bend to the strong emotional impulses to which we are all subject.

A recent and rather public spat illustrates this point well: the co-founder and CEO of Palantir, Alex Karp, delivered a heated diatribe against an investor (yes, a value investor) who had publicly disclosed a large short position against the company. On a segment on CNBC, Karp said:

“When I hear short sellers attacking what I believe is clearly the most important software company in America and therefore the world in terms of our impact, simply to make money, and calling the AI revolution into question…. It is super triggering, they can pick on any company in the world, they have to pick on the one that actually helps people… why do they have to go after us?

Honestly I think what is going on here is market manipulation… I do think this behaviour is egregious, and I’m going to be dancing around when it is proven wrong… it’s bats#!t crazy!”

CNBC, 4th November 2025

Now, Alex Karp is a clever man by any measure: he has a law degree from Stanford University, a PhD in social theory, and is the co-founder of a large and successful company; hardly the hallmarks of an idiot. Yet, his reaction to Michael Burry’s short position is quite clearly driven by emotion, rather than intellect. Burry is arguably one of the most famously successful short sellers in history, known for his quiet, analytical approach, rather than any predilection for loud, aggressive short campaigns. The points that Burry was making at the time, and continues to make, were simply that Palantir was wildly overvalued, and burning through precious company stock to overcompensate its management (including Karp himself).

Whatever Karp might think about these things, his reaction did not address or consider them: rather, Karp – an intelligent man – attacked Burry, questioning his motivations[2], baselessly alleging illegality, and utterly missing Burry’s point. It is hard to argue that there should at least be a possibility of having a reasonable, rational short thesis on a company valued at 455x earnings, as it was then (now a mere 150x)[3]. But Karp was not approaching the issue rationally, falling instead into Bonhoeffer’s trap, where “reasons fall on deaf ears; facts that contradict one’s prejudgment simply need not be believed”.

This episode is, for us, symptomatic of why value investing has endured successfully for so long, and, in our view, will continue to do so: we are not competing with the rational, but instead taking the opportunities afforded to us by actors making irrational long-term decisions.

We are quite satisfied that most of the time, most of the market will be sane, buying and selling portions of companies for a reasonable reflection of what they are worth. But sometimes – and we only need it to be sometimes – we will have the chance to buy cheap shares from someone who has no rational reason to believe that the long-term value of the business is impaired: perhaps they simply think that next quarter’s earnings will be a penny short of expectations, or they are fed up of owning a boring business struggling with a cyclical downturn. Maybe they just hate the management team, who turned them down for a round of golf. Whatever the cause of the irrational decision, if the price is right – and if the business passes our inspection – we will stand ready as willing buyers.

No panic, no problem

Which brings us to the topic of AI. Many have fretted about the impact on jobs, and we don’t have any insightful answers to offer on that front. Where we feel more confident, however, is in assessing the potential impact of AI on value investing – and whether or not an approach that has persisted for decades will endure.

Consider the standard fear: AI tools, from Claude to Grok to Gemini, are super-efficient at many tasks[4], eliminating the need for ‘knowledge workers’[5]. In theory, there is no reason that this won’t hold for many parts of the stock market: there are indeed some traders who pore over opaque markets, and profit from minute inefficiencies, buying and selling on small micro-signals based mostly on algorithms. In our view, they should indeed be worried about their ability to continue to earn their keep; after all, why can an AI tool not interpret the same small mathematical signals, spot obvious arbitrage, and close price gaps with speed and precision[6]? Over the course of financial history, mathematical arbitrage opportunities have generally closed as inefficiencies are addressed, and technology improves to close the gap.

Value investing, though, is a different beast, as we have described above: not only is it slow and relatively simple in its thinking – buying companies for less than their intrinsic value – it is, by nature, focussed on irrationality. The key constraint isn’t brains, awareness, or computing power, but temperament and judgement. AI has no temperament, it simply executes on a specific mission within a specified set of parameters, mimicking the wholesale patterns it sees in its training set – and, as an investor recently remarked to me, “AI is super smart, right up until the point of judgement”.

If an army of super-rational AI bots enters the marketplace, trading furiously to eke out market-neutral profits, this has no impact whatsoever on us; after all, all we seek to do is buy from those sellers who are too panicked or short-term-focussed to consider the true, long-term value of a business, and reflect that in share prices. We earn our keep by staying calm when others panic, thinking more than three months ahead. If AI bots don’t panic, then we simply won’t be doing business with them, and we will gravitate towards those areas of the market where the poor sentiment and negative emotions have depressed valuations beyond the point of reason. The fact that prices are set on the margins – by the most anxious seller, or the most excited buyer – helps us in this regard, amplifying the extremes of negative emotion, and affording us select opportunities to deploy capital in a prudent, time-tested manner.

It is this very feature – always skating to where the puck is the cheapest – that keeps value investing interesting and profitable, and its future evergreen. When what we buy goes back into fashion, and grows in price, we sell it – and go back to hunting out in the bargain-bin of the marketplace. It is not most of the market being right that matters to us, but a select few, every so often, being wrong about the long term. Indeed, when I made my pilgrimage to Omaha, I was fortunate enough to hear famed value investor Warren Buffett clarifying this point in person, in his typically straightforward fashion:

“New things coming along don’t take away opportunities: what gives you the opportunities is other people doing dumb things”

Warren Buffett, May 2023

So, for now, we embrace AI – and continue to celebrate human irrationality.