Artificially Intelligent Psychology explores the profound interplay between human cognition, language, and artificial intelligence, particularly within organizational contexts. The book argues that language itself is humanity’s first artificial intelligence—a predictive tool that extend
s the mind beyond the brain. As AI systems increasingly mimic and manipulate language, understanding their nature becomes essential for co-evolving with them.
The narrative unfolds through a conceptual journey, beginning with the historical and psychological roots of tool use, emphasizing how humans have always shaped and been shaped by technology. It distinguishes between predictable systems (like machines) and the inherently unpredictable nature of human beings, framing intelligence as a mapping process—where both brains and machines create compressed representations of reality.
Central to the book is the idea that organizations function as predictive systems, and that AI can assist with problems and secrets, but not mysteries. Trust, shared situational awareness, and the limitations of language models are recurring themes. The text critiques psychological constructs, warns against category errors (confusing maps with reality), and highlights the dangers of over-relying on AI-generated language.
Ultimately, the book invites readers to rethink intelligence—not as a static trait, but as a dynamic, socially embedded process. It encourages the design of new conceptual tools and organizational practices that embrace unpredictability, storytelling, and human agency. The epilogue suggests that the real challenge is not the rise of AI, but learning to see clearly with our extended cognitive tools. The English version of this book is available through this website (unfortunately in Norwegian although the book itself is in English!):
Here is the table of contents, together with a reading guide to understand the journey I am inviting to:
Table of Contents & Travel Guide
- Robots in Organisations, Management and Psychology
Human minds have always evolved through interaction with technology. Tools are extensions of our intentions ‘by other means’, and one such tool is what we now call ‘artificial intelligence’. To co-evolve with our new talking tools, we must learn to understand both the tools and ourselves more clearly.
- Technologies Are Predictable. Humans Are Designed as Unpredictable
We use technology to shape the future in our image. In that regard, our concept of ‘intelligence’ carries a strange built-in flaw: we tend to assume that the world is predictable and thus knowable. Undeniably, predictability benefits the hunter who wants an efficient weapon. Just as certainly, to be unpredictable is essential for the prey, as they must learn to stay hidden.
- Three Fundamental Predictive Tasks for Intelligence of All Kinds
Organisations are advanced hunting tools. Operating them presents humans with three very different challenges in prediction: Problems, Secrets and Mysteries. Crucially, AI can only assist with two.
- Science and Artificial Intelligence Are Rooted in Trust
Language emerged as a tool for people to develop mutual situational awareness. This is where science—and belief in intelligence—was born.
Trust and shared situational awareness
Everyday science and academic science
Artificially intelligent?
- All Intelligent Systems Are Types of Maps
Brains create maps of reality, and language is just one type of map among many. A key feature of a good map is its minute size. It must be much smaller than the terrain it describes.
A brief exercise
A good map is a small map
Good maps are predictions
What AI Has in Common with Ticks and Used Car Dealers
Representations as predictions of actions
Representations and hallucinations
Language and competence without comprehension
Shared intelligence
- Our First Artificial Intelligence Was Language
‘Extended mind’ lets you become more intelligent through the use of gadgets outside your body. You never think with your brain alone.
Cultural tools
Arithmetic tools
External thinking tools: writing
Extended mind
- Mapping Maps: Basic Information Theory
A map is a code for a message. The systematic study of codes, messages and media is called ‘Information Theory’. Essentially, the news is the same on both FM and DAB radio.
Information Theory I: message, content and intelligence
Information Theory II: codebooks
Information, codes and media
- The Foundation of Artificial Intelligence: Mapping Codes
When a map becomes a mathematical code, you can therefore calculate new maps. That is how you get ‘language that speaks itself’.
Redundancy in messages
Probable and improbable messages
Plausible sentences without meaning
- Calculating Meaning
Language models do not guess the next word—they predict the next bit of meaning.
Semantic spaces
Meaning as statistical geometry
- On the Threshold of Artificial Intelligence
We struggle more with language than we would like to admit. Your brain is packed with intentions of all kinds, at all levels. In short, you have agency. You are a self-determined agent.
When will we believe in AI?
The brain exists for survival, not IQ: intention and agency
Ever-larger language models
A self-speaking language
- From Language Model to Quasi-Agent: What is an Intelligent Alarm Clock?
Making machines ‘intelligent’ risks making them less reliable.
God doesn’t write prompts
Would you want an intelligent alarm clock?
Would you want an artificially intelligent partner?
- Organisation, Work, and Management as Practical Predictions
Universities can now let a language model design their academic programmes. This poses the question – what do we really need from education?
Capturing the future
Management Studies as Disciplines of Prediction
- Predicting What We Already Know is of Little Value
Publishing the multiplication table adds little if people can already compute it. There is a lot we ‘know without knowing that we know’, much of which can be figured out if we just pause to think. Paradoxically, we often believe we need this knowledge, even if we do not. This is a hard blow to much research in organisational psychology.
Predicting ‘leadership’
All we know if we just think about it
The need for research and empirical knowledge
- Abstract Concepts Are Also Maps, But How Real Are They?
Many scientific ‘constructs’ have a tenuous relationship with reality yet remain useful. They are colloquially known as ‘social constructs’, and much of human behaviour depends on us taking them seriously, as there are scientific conventions for defining them. Language models can handle such ‘constructs’ just fine, no matter how ‘made-up’ they are.
Gnomes
Money
Constructs
- In Language, All Ideas Are Equally Good: The Very Real Danger of Confusing the Menu With the Food
To language models, all ideas appear equally ‘real’. They can’t tell a map from its terrain. This means that there is a real danger that they might eat the menu instead of the meal. In science, this is called a category error.
- Language Is Constantly Evolving. We Can Create New Concepts As We Need Them
Fortunately, language is a flexible tool. We can create new linguistic tools and concepts to grasp new phenomena as they arise.
Organising the world
Vocabulary explosion
Sensemaking: about organisation, leadership, and other inventions
A ‘buzzword’
- How Solid Are Psychology’s Own Concepts?
Psychology is especially vulnerable to our tendency to mistake the map for the territory. What do we really have inside our heads—and how does that affect our ability to predict human behaviour?
Sensemaking in psychology
Inspiration as spirit summoning
Diagnoses and self-fulfilling prophecies
From words via constructs to ghost: The 42% solution
- How the Brain Models the World
It’s time to leave behind constructs, statistics, and philosophical mazes. What is the brain actually doing?
Astronomical data between your ears
Energy use in brains and artificial intelligence
- How the Brain Perceives, Understands, and Produces Language
Language involves many different codes, which are represented in many different ways in the brain. Here are some of them:
The Circuitous Route of Language Through the Organism
From sound to word
From words to meaning
When things do not mean what is said
From meaning back to speech
Language is for soliciting alliances
- The Pace of Change and Learning: Intelligence in DNA, Nerve Cells, and Language
What is information, really? And how robust is the kind of information we use in our brains and languages to navigate organisations?
Where information comes from
Wisdom from the kitchen
The wisdom in jelly and smoke
The Cognitive Ossification Through Language
- Deep Knowledge is a Compressed Map
We do not need to understand every piece of technology in this world to use it. But users of a particular technology must understand it well enough for their own purposes. That raises the question: what should knowledge actually look like?
You in a nutshell
Deep knowledge about artificial intelligence
- Organisations as Artificial Intelligence
How much help are language models actually to organisational psychology? What is control, how does that relate to problems, secrets and mysteries, and how do we move from words to action?
Introduction to the Role of Text Robots in Organisational Psychology
The catalogue of problems
Optimisation is not control
Examples of calculated organisational psychology
Method and Technology as Organisational Innovation I: The unpredictable
Method and Technology as Organisational Innovation II: The predictable
Method and Technology as Organisational Innovation III: Getting Out of the Quagmire
Being Cool is just getting easier: Psychology as Design
- In Reality, Life Is a Tale of Mysteries
Most of real life happens outside the predictable rhythms of formulas and algorithms. We must understand the limits of prediction—and the value of inventing our own worlds. We do that by telling stories.
Introduction to Reality
The limits of mapping
A space of intentions
Stories are combinations of agents with goals
Myths, constructs, and pickup lines
Is the Language Robot a Hero, a Villain, a Friend, or a Traitor?
- Epilogue: The Way Out Is a Loop
It was not such a big deal when we learned to read and write. Subsequently, it might not be such a big deal that machines now help us to read and write in a new way. What does matter is that we learn to use our extended cognitive abilities to figure out together what we are really looking at, once we have become used to the view.