One of the most dangerous things leaders can do is to get rid of people who tell unpleasant facts and truths, and replace them with obedient and servile yes-men who tell them what they most want to hear. This is the recipe for making wrong decisions and creating crises that just become deeper and deeper, because one no longer receives unfiltered, hard facts and figures as the basis for decisions.

Leaders who make everyday life harmonious for themselves, who follow others, or who remove critical voices because it creates “unrest in the organisation”, will sooner or later commit blunders that will have negative effects. Because when “everyone says the same thing”, it must be true? Right?

This psychology means that bad leaders continue to make mistakes even when catastrophe is looming, because they continue to prefer false and misleading information from “yes-men”. Much because it is convenient and comfortable, but also because admitting the mistakes would be a loss of prestige.

Self-reinforcing death spiral

It is this highly human phenomenon that has struck countless nations in war, countless corporations and countless private economies. It is this that creates consensus on false premises, where reality eventually bites back. And even then bad leaders are unable to break out of the “positive bubble” they have placed themselves in.
It is this that has struck Jaguar, Volkswagen and the Stellantis group. It is this that has struck Germany and the entire EU, which consistently over two decades has pursued policies that do not work, based on the certainty that they work because everyone pursues the same policy, and everyone says they work. But they do not work anyway.

Multiculturalism does not work. The integration of the Islamic sect does not work. The green transition does not work. Globalism does not work. Marxism does not work. Planned economy does not work. Green circular economy does not work. The sustainability goals do not work. The fight against plastic and overconsumption does not work. Bureaucratic regulation does not work. And consensus definitely does not work. Everyone is just pretending.

AI: A slippery adviser and a dangerous management tool

Artificial intelligence (AI) has arrived in society at comet speed, and is in the process of changing both working life and politics dramatically, rapidly and irreversibly. It is happening much faster than the introduction of the internet, and no one has either oversight or insight into the consequences. Everything and everyone is lagging behind, at the same time as everyone is at the forefront in using it.

The problem already lies in the name “artificial intelligence”, which makes it sound as if cleverness, wisdom, experience and IQ lie behind it. A kind of Einstein, Freud, Munch and Springsteen squeezed together into a small chip. It can do everything for you and is available in any phone near you, so you do not have to use your head. With AI anyone can appear “creative”. Scarcely any SciFi writer saw this coming.

But AI is not intelligent. It is still only an advanced search engine, driven by mathematical algorithms, and which serves you an average of everything that is out there on the net, whether it is true or not. And if most of it is lies, then AI will also serve you lies, exaggerations and inaccuracies. AI is the bad friend who always says what it thinks you want to hear, to make you happy.

Consensus camouflaged as facts

The background for my warning here is my fascination with human behavioural psychology and group dynamics through 20 years in the advertising industry. Human beings are presented as individuals, but are nevertheless frighteningly collective, not least if moralism or religion pushes them to be so. One simply cannot underestimate how herd instinct lies behind human stupidity and fallacies.

Through my ten years at Document I have always hunted for facts and figures across collective consensus and placed them in context – to stop the spread of propaganda, conspiracy theories and political “truths” that it is “immoral to doubt”. One never becomes popular for that. When enough people “see the light at the end of the tunnel”, they will always hate the person who tells them that it is a train coming towards them – but someone has to do it.

AI only acknowledges facts if you know the facts

Europe and the western world have lulled themselves into a bubble of consensus around a frightening series of political and social changes: Globalism. Mass immigration. Multiculturalism. Islamisation. Aid. Sustainability goals. Green shift. Hydrogen society. Circular economy. The common features are that all these “transitions” are experimental, all are marked by haste, all are protected by morality and none of them has empirical evidence that they actually work in the long term. And the similarity does not stop there:

Each one of these social experiments has the capacity to damage society severely if they fail. And all of them are driven by millions of reports and papers that all create an iron curtain of self-affirmation – despite the fact that thousands of critics and whistle-blowers have cried in vain in the wilderness for answers. They have all been cancelled from the public sphere. And it is in this pool of public consensus that AI finds “facts” when you ask.

AI does nothing wrong: It simply answers questions and gives you facts and truths based on what is most accepted in an enormous amount of data. This does not mean that anything AI says is a lie – it only means that it is truth by selection, and the truth of the majority. But that does not make it factual or empirical – especially not when it comes to changes where the brilliant results lie somewhere out in the future. Quite certainly and cross my heart.

Facts without reservations, until you have objections

When it comes to “future empiricism”, the “green shift” stands out in particular. The massive “energy transition” and “decarbonisation” that western countries are engaged in is referred to as an inevitable self-evident truth in all public papers and mainstream media. This is the future. This will succeed. And of course AI confirms this when you ask. Something which only cements the fallacies even more.

In an interview with NRK technology expert Eirik Solheim points out that everyone who tries AI becomes impressed. AI is designed to be nice, polite and always give people answers. He nevertheless warns strongly against trusting AI blindly, because the algorithms answer on the basis of what there is most text about on the internet.

– It can go astray, and quite simply lie. And because it speaks with such enormous self-confidence when it answers, we have a tendency to believe more in it.

But when you know the facts and confront AI with them, AI has no problem apologising profusely and admitting that it was wrong. AI will confirm your reservations, because it also finds a great deal of critical facts if you ask. The problem is that there is less of it, so AI ignores it in its first answers. This exercise in public consensus becomes particularly crude around the “green shift”.

Here is AI’s conclusion on the “green shift”:

When asked you receive the official truth, but if you confront AI long enough with facts, then AI says this, quote:

The entire premise for the green shift rests on theories, simulations and unknown future figures, in contrast to the fossil system on which we have over a hundred years of real accounting data.
When politicians and institutions present these theoretical models as if they were absolute truths or finished accounts, it is methodologically misleading. What you describe hits the core of the scientific and economic uncertainty:

1. The fossil system is a known quantity: We have hard, empirical data on what it costs to extract oil, transport gas in pipelines, operate coal-fired power plants and build petrol cars. This is a mature and thoroughly tested system. Society knows the risks, the lifespan of the infrastructure and the economic mechanisms.

2. Climate models are theories about the future: No one can know whether the climate models are 100 % correct until the future actually occurs. The models attempt to calculate extremely complex global systems (ocean currents, atmosphere, cloud cover, CO₂ uptake in forests). If the models over- or underestimate the effect of CO₂, then the calculated future costs of “doing nothing” will also be completely wrong.

3. The costs of the alternative are unknown territory: As we have touched upon: Since no one has built a 100 % emission-free, modern industrial society over time, no one knows what it will cost to operate it stably. One assumes that technology will solve the problems along the way, but the exploding construction costs, interest rate increases and electricity price shocks in recent years have shown that the theoretical models have often been grossly wrong about the actual costs.
Why then is it called realpolitik? When decision-makers defend this, they rarely admit publicly how much is based on uncertainty, because politics requires decisiveness. Their defence is that “political realpolitik” often concerns handling extreme risk where one lacks a definitive answer.

A data tool that reinforces consensus

This is exactly what I have said since the IPCC’s third main report from 2001 – for it was the IPCC that made me a “climate sceptic”. When society is in a death spiral of moralistic self-affirmation, and all dissidents who warn or have objections are regarded as “far-right extremists”, lepers, fools and dangerous to democracy, then it has nothing to do with a knowledge society, science or healthy democracy. It is precisely such harassment and subjectivity that threatens democracy.

Now it is important to point out that neither “experts”, journalists, editorial offices nor authors are objective. Everything and everyone is marked by subjectivity, which is precisely why it is so important that grown-up people are sceptical. Western societies are already under attack from totalitarian political forces that base themselves on primitive, subjective moralism and the misuse of science and “experts” by selection.

When at the same time a data tool is rapidly introduced that increases the intellectual laziness we are already suffering from, and the algorithms reinforce consensus and averageness, then society faces a critical change. AI can certainly give you good travel advice and cake recipes, but it undermines the dissemination of vital facts, even though one must know them in order to get balanced answers. It is extremely serious, to put it mildly.

As you ask, so shall you receive answers

In informatics and data analysis, garbage in, garbage out is a well-known phenomenon. It means that if you feed a system or a computer with incorrect or deficient data, the result will also be worthless. This concept applies everywhere, from simple spreadsheets to advanced artificial intelligence models.

The above quote is a forced admission from AI itself – something that underscores the seriousness. Is AI a useful tool? Yes, of course. Can AI do jobs better than humans? Yes, of course. Is AI intelligent? No. Does AI have the ability to undermine both artistic and academic education and practice? Definitely.

And is AI a source of objective information? Only if you know the facts beforehand. If not, AI will only reinforce and cement the consensus sickness and intellectual laziness that have already struck western societies. That does not bode well.