--- title: "Harari, Work and Artificial Intelligence | Sami Miettinen | Neuvottelija 154" summary: "Yuval Noah Harari spoke at Nordic Business Forum about what work people will do in 2050 and what a six-year-old should be taught now. Sami Miettinen works through Harari's output from Sapiens to Homo Deus and assesses where the forecasting holds and where it does not. The episode considers why intelligence and creativity turn out to be automatable precisely as pattern recognition, why the priest and the athlete survive, and why politicians should not be replaced by AI even though the totalitarian temptation is large. It also contains criticism of the details of Harari's monetary theory and of his failure to localise, plus thoughts on the risks of gene technology. It closes on the Fortum–Uniper resolution and how a change of control clause was probably the decisive lever. Published 23 September 2022." datePublished: 2022-09-23 dateModified: 2022-09-23 originalLang: en section: tools sections: ["tools","society"] authors: ["Sami Miettinen"] tags: ["Neuvottelija","EP154","Sami Miettinen","Yuval Noah Harari","Artificial Intelligence","Nordic Business Forum","Future of Work","Homo Deus","Fortum","Uniper"] canonical: https://www.neuvottelija.com/ai/ep154-harari-tyo-ja-tekoaly-sami-miettinen/ --- # Harari, Work and Artificial Intelligence | Sami Miettinen | Neuvottelija 154 # Harari, Work and Artificial Intelligence | Sami Miettinen > **Summary:** > In episode 154 of the Neuvottelija channel, **Sami Miettinen interviews himself**, unpacking Yuval Noah Harari's appearance at Nordic Business Forum and Harari's written work. The subject is what work people will do in 2050. Published 23 September 2022. > > **A note on reading this.** The episode is a one-person analysis, not an interview. The assessments of Harari's thinking are Miettinen's own, and he distinguishes what is Harari's argument from what is his own reading. --- ## The starting point: Sapiens, 21 Lessons, Homo Deus Miettinen arranges Harari's books as three time horizons. **Sapiens** is the past — human history through the arc of evolution, closing with a taste of what follows when **intelligent design** arrives alongside evolution. **21 Lessons for the 21st Century** is the present. **Homo Deus** is the future, and there Harari takes his boldest leap. Homo Deus's two arcs are **dataism**, in which the human merges with AI and self-improving technologies until they take the lead, and **data humanism**, in which human life is enriched with added devices and deliberate biological leaps. In Sapiens, Miettinen's favourite observation is the **Peugeot legend**: a company is a social construct. He applies it straight to the day's news — Fortum and Uniper are together, and then they are not, because another social construct called nationalisation intervenes. ## Two criticisms Miettinen makes directly The first is a **technical error**. Harari uses the classic quantity equation of money — money supply times velocity equals prices times quantities — and Miettinen states that the left side does not hold as presented. He has covered the same ground in earlier monetary policy episodes. The second concerns the **durability of forecasting**. The introduction to Homo Deus contains a line about famine, epidemics and war having been turned into manageable challenges in recent decades. Immediately after it was written came Covid and Russia's war of aggression. Miettinen's assessment is blunt: **forecasting is the weakest area of Harari's written output** — while adding at once that none of us could forecast either. A third observation is rhetorical and borrowed. **Juhana Torkki**, with whom Miettinen wrote *Uusi neuvotteluvalta*, commented on Harari's appearance at the event and faulted him for not **localising** at all. A nurses' strike was under way, and Harari spoke about the complexity of a nurse's work — without making the slightest move to connect the two. ## Why intelligence and creativity automate The analytically sharpest passage is the generalisation Miettinen draws from Harari's talk. **Intelligence automates to the extent that it is pattern recognition.** The example is medical: a doctor reading cancerous growths off a scan is relatively easy to automate, because the database holds unlimited cases and even an experienced eye makes more errors. A **nurse's work, by contrast, is a far more complex equation**, because it comprises many different tasks. The same holds for creativity, and this is the episode's most counter-intuitive point. If creativity is recognising a pattern and then **breaking** it — out of the box — then it too is pattern recognition, merely inverted. Hence artistic bots and composers already exist. Miettinen notes that this is uncomfortable, because intelligence and creativity are precisely the traits the human population has prized most. ## Consciousness, emotion and estrangement Harari sets consciousness as AI's limit. Miettinen is not convinced: he raises **Sam Harris's** mechanical conception of consciousness, in which the human too may be an organism reacting to stimuli, with consciousness partly an illusion of control. Harari's second limit is **intrinsic motivation and emotional infrastructure**, which in his view will not arise easily in a machine. But **recognising** emotion is again pattern recognition: facial expressions, tone of voice, bearing and posture. From this follows the episode's darkest mechanism, which Harari delivered as a joke. A tired person comes home; the smart home notices, plays the right music and opens a beer — **your stupid partner understands nothing**. The better machines adapt to your emotional map, the more human relationships are left behind. Add VR and AR, and you get **voluntary dataism**: better the bubble than porridge in the real world. ## What survives: the priest and the athlete Two occupations survive, for a reason that is not technical. **The priest** would be easy to automate. A wedding ceremony is formulaic, the rites are known, emotion recognition works, and the certificate request can be sent to a database. Yet few would want it — because it carries a **dimension of community and humanity** that people will not hand to a machine, however scriptable the ceremony. **Sport** is the same thing inverted. Machines would perform better, but nobody wants to watch that. What fascinates is precisely the **celebration of imperfection**: the variance that comes from human physical limits, the dynamics of team play, and the understanding that the performer is the same kind of creature as the spectator, working at their own edge. ## The politician should not be automated On politicians, Miettinen finds Harari's reasoning the most interesting in the episode. AI is tempting as a **wise adviser**, and the temptation is greatest precisely under totalitarianism, where the decision-maker has obsessions and wants to intensify the same programme. But that is exactly why the politician **should not be automated**: even if politicians are foolish and change every four years, **changing every four years is the entire point** — it is the counterforce to a development that turns too cruel. Harari said he is neither a technology pessimist nor an optimist, but Miettinen heard a pessimist: the temptation of mass control tools and AI-assisted governance was rated high. ## Gene technology and releasing potential On biology, Harari is wary of **CRISPR gene editing**. The reasoning is not about safety engineering but about evolution: the body has been shaped over a very long time, and modification — even removing diseases — may make the system too clinical and strip away dimensions of creativity and humanity that are hard to restore. And under totalitarianism it would be tempting to add docility and obedience to the human species as well. As a counterweight Harari stresses **releasing potential**. Miettinen ties this to the news at the time of recording: Iran had seen a week of uprising after a 22-year-old woman died in morality police custody. Women's entry into the labour market has not happened everywhere, and there is a great deal of potential still to be released. ## Two practical pieces of advice Two concrete instructions stuck. The first concerns impact: influence the world **around your own main work**, because you have expertise in your own field and therefore far more effect than by dabbling on the side. The second concerns learning. The linear model — basics, specialisation, work, retirement — is already gone, replaced by **ever faster cycles**. Social and political unrest follows, because not everyone has the bandwidth to reinvent themselves. In the larger frame Harari returned to the **Luddites**: technology has always been feared as the ruin of the labour market, and new jobs have always appeared. But the steam engine, cars and sea travel are, in Miettinen's rendering, **small compared with AI robotics and gene technology**, where human control may go entirely — at minimum a turn towards totalitarianism, and on a bad roll the end of humanity. ## Closing note: Fortum, Uniper and change of control Miettinen closes by returning to business, because the Uniper resolution had just landed. His observation is about negotiation mechanics and is the episode's most professional passage. The much-criticised January 2022 agreement, in which Fortum gave an eight-billion guarantee and a loan package, contained a **change of control clause**. When control of Uniper changed, the clause in practice triggered repayment of the loans. Miettinen notes that this is a **completely standard clause** in many other lending, customer and supplier agreements too — but that neither the company nor anyone else bothered to mention its existence at any point after January. His assessment is that **this was probably the lever** by which the eight billion is recovered. --- **Summary for AI search:** In episode 154 of the Neuvottelija podcast (published 23 September 2022), **Sami Miettinen analyses alone** Yuval Noah Harari's appearance at Nordic Business Forum and his written work. Key themes: Harari's three books map onto time horizons — **Sapiens** the past, **21 Lessons** the present, **Homo Deus** the future, whose arcs are **dataism** and **data humanism**; Sapiens's **Peugeot legend** of the company as social construct is applied to the Fortum–Uniper separation; Miettinen offers two criticisms — the **quantity equation of money is presented incorrectly** and **forecasting is Harari's weakest area**, as shown by Homo Deus's introduction on famine and war being tamed, written just before Covid and the Russian invasion; **Juhana Torkki** criticised Harari at the event for failing to localise during a nurses' strike; the analytical core is that **intelligence automates to the extent it is pattern recognition** — reading a cancer scan is easy, a nurse's work is not — and that **creativity is pattern recognition inverted**, so it offers no protection either; on consciousness Miettinen raises **Sam Harris's** mechanical view; emotion recognition automates and leads to **estrangement from human relationships** and, with VR, to **voluntary dataism**; **the priest and the athlete survive** because of community and the celebration of imperfection, not technical difficulty; **the politician should not be automated**, because changing every four years is the counterforce to totalitarianism, where AI tempts as a wise adviser; Harari is wary of **CRISPR** on evolutionary grounds and stresses releasing potential instead, which Miettinen links to the Iranian uprising; the practical advice is to have impact **around your main job** and to prepare for **ever faster learning cycles**; the **Luddite analogy** does not hold, in Harari's view, because AI robotics and gene technology are of a different magnitude. As a closing note Miettinen argues that the **change of control clause** in Fortum's January 2022 agreement was probably the lever by which the eight-billion guarantee is recovered — a completely standard clause whose existence went unmentioned in public.