Roald Dahl’s AI Prophecy: ChatGPT’s Unexpected Roots

Will AI Lead to Job Losses and Transform Capitalism?

As I was rereading Roald Dahl’s "The Great Automatic Grammatizator ", I realized something startling: Dahl had envisioned the concept of ChatGPT long before its time. This short story, published in 1954—the same year IBM introduced the first mass-produced computer, the 650 magnetic drum calculator—features a young engineer who has just played a pivotal role in the invention of a “great automatic computing engine” for the government.

The Engineer's Dilemma

The engineer, who in his free time is a writer, finds his work consistently rejected because it does not meet the standards of the magazines.

“In the last ten years I’ve written hundreds, literally hundreds of short stories. Five hundred and sixty-six, to be precise. Approximately one a week.” “Good heavens, man! What on earth did you do that for?” “All I know, sir, is I have the urge.” “What sort of urge?” “The creative urge, Mr Bohlen.”

He starts to become obsessed by the idea to use the same technology to fuel his ambition to finally get published. Although he first realises that a machine is not capable of original thought, he then understands that the rules of grammar can be used to program the computer. He also notices that most published articles follow a certain style and formula, according to the taste of the audience.

“Give it the verbs, the nouns, the adjectives, the pronouns, store them into the memory section as vocabulary, and arrange to have them extracted as required. Then feed it with plots and leave it to write the sentences.”

Parallels with Modern AI

With this approach, the machine operator can create an infinite number of articles and books, adjusting for variables such as writing style, story formulas, tone of voice, characters, literary style, plots, and the amount of passion. We might call these prompts. Sounds familiar? I thought so too.

This technological concept is strikingly similar to modern AI systems like ChatGPT. Beyond the technical parallels, Dahl’s story also echoes the societal concerns and developments we see today as we enter the AI era. Primarily, it reflects the anxiety about the evolution of technology and its potential impact on work, especially among the creative class—similar to the concerns we see today among artists affected by AI.

A Cynical Critique

Additionally, the story can be interpreted as a cynical critique of how the public consumes algorithmically generated content over original, creative, human-created works, with publishers catering to this demand. This criticism mirrors current debates over generic AI content and social media algorithms.

The Engineer’s Success and Ethical Dilemma

The narrative progresses as the writer and his business partner quickly become the most published authors in the country, even using pseudonyms to publish more. In his new role, our protagonist becomes obsessed with maximizing revenue through commercial content. This leads to an intriguing development where the new tech millionaire offers established writers a deal they can’t refuse: lifetime pay to cease writing and allow the machine to write under their names instead. Would that work? The engineer, focusing on money and productivity, had forgotten about the creative urge:

“They’ll never agree.” “You don’t know writers, Mr Bohlen. You watch and see.” “What about the creative urge, Knipe?” “It’s bunk! All they’re really interested in is the money—just like everybody else.”

Modern Reflections

Seventy years later, in the early days of Gen AI, we find ourselves pondering similar issues. AI’s impact on automation, job displacement, the value of human creativity, productivity, and the future of capitalism are all pressing concerns. Dahl’s story brilliantly prefigures these contemporary debates. Just as in "The Great Automatic Grammatizator," AI has the potential to increase productivity substantially, improving employees’ surplus value or replacing jobs altogether. As a society, we need to understand what that means and how we will distribute that productivity surplus in a way all citizens benefit. While some companies have adopted a four-day workweek to more equitably redistribute the surplus value generated by AI, they are exceptions.

Subscribe and receive more content like this

Most of us don’t question the “logic” that the productivity gain goes directly to the company—stuck as we are in the 40-hour workweek model. That 40-hour workweek is based on the situation with (manual) labor as a means of production, owned by capital. Labor was turned into a product, that was also owned by capital, and then sold to the market.

The Rise of the Knowledge Worker

The rise of the knowledge worker (which started in with the release of the IBM 650) brought a mayor shift: suddenly the knowledge worker had became the owner of the means of production - and was also producing knowledge (Drucker, 1960). The latter, the value surplus, for example: a software tool, is owned by capital and sold to the market. We can see this ownership of output reflected in labor law: copyrights on research, code, and anything else produced during ‘working hours’ are owned by the employer, not by the employee.

The Question of Ownership and Redistribution

In comes Artificial Intelligence—a new means of production. Neither owned by the employer, nor by the employee.

The marketer, owner of their knowledge as means of production, using that same knowledge to prompt an e-book, that is written by AI, a book, then owned by capital and sold to the market. However, while capital owns the profit of the book, they do not own the copyright. In fact, no one does. There is no copyright on AI-generated output unless adjustments are made.

So why does capital own the benefit of its surplus value? If we are more productive because of AI—why aren’t we moving to compressed working days for better work-life balance, or why aren’t we increasing salaries? And if productivity is just another word for unemployment, what will happen to employees if AI starts replacing our jobs?

“Yet, as labour is redeployed and the potential for surplus value increased, questions of redistribution are ignored and importantly for us, the concept of objectified labour is underplayed”- (Hughes et all, 2019).

Exploring the Future

To answer that: I dived into this paper: The world of work and the crisis of capitalism: Marx and the Fourth Industrial Revolution,

Our capitalist system is based on two main drivers: maximizing value and society’s ability to consume. There is a possibility that in the near future, jobs are replaced by AI at scale and at a rate that will leave many without income. Yes, new jobs will emerge, but can they be done by the same people, will there be enough work and can we up-skill and re-skill fast enough?

“While the pursuit of profit is primary, the destruction of the ability to consume is a constant threat produced simultaneously. There is no space for social responsibility in this equation, meaning technology and automation may well destroy jobs at a much quicker rate than capital can create new employment” - (Hughes et all, 2019).

The Main Question: Sustainability of Capitalism

The main question is: can capital create enough new employment fast enough to sustain itself? Because if not, that might be the end of capitalism as we know it. As quoted in the same paper:

“Think about all of the gadgets that it creates, the technologies. Just very briefly imagine for a moment that this technological innovation, artificial intelligence, robots, moves in a manner in which it is moving, but even faster. Very soon, you’re going to have robots producing everything. Now, the robots do not want to consume that which they produce. And the rest of humanity is not going to have the money to buy it. So, capitalism is going to have a massive crisis, simply because it will have a humongous capacity to produce stuff, and no capacity to consume it, which is already what we are observing”(Democracy Now, 14 May 2018, no page number)

The Future of Capitalism

And if AI indeed will be the cause of a crisis in capitalism, what will happen? Will this be the driver to a more collectivist society and will companies distribute the value surplus over the broader society?

“If machinery was to be introduced in a truly collectivist society in which wealth was distributed fairly between citizens, it would have benefits both in reducing the amount of work that is necessary for production, while simultaneously increasing the wealth of both society and each individual member” (Hughes et all, 2019)

Maybe, a more fairly distributed wealth would not be a danger to capitalism, but the only option capitalism has to survive.

The Plot

Dahl’s story ends with capital providing lifetime payments for established writers to stop writing and let the machine do the work for them. Similarly, Elon Musk predicted at VivaTech 2024 in Paris that in the future, when AI and robots produce all we need, jobs will become optional—we will work only if we want to, not because we need to. For this scenario to work, Musk suggested the necessity of a ‘universal high income,’ although he did not elaborate on what that would entail.

If such a scenario unfolds, who will fund this universal income? Will it be our governments, or will capital itself bear the cost, as it might be the only way for capitalism to survive? This raises significant questions about the future distribution of wealth and the role of capital in an AI-driven economy. I will leave you with these questions, and the final words of Roald Dahl’s story:

“This very moment, as I sit here listening to the howling of my nine starving children in the other room, I can feel my own hand creeping closer and closer to that golden contract that lies over on the other side of the desk. Give us strength, Oh Lord, to let our children starve”

Let’s talk

Business leaders face the challenge to create future-proof and sustainable people strategies, ensuring both productivity and employee well-being while navigating complex AI integrations. I have been discussing AI’s impact on companies Employee Value Proposition with tens of both HR and Business leaders over the last months.

Interested to have a chat and find out what others are doing?

Let's Talk