Elon Musk and His Predictions About AI
A few weeks ago, Elon Musk, was interviewed by The Economist for The Insider. The 90-minute interview took place at one of Tesla's enormous factories in Texas.
Image created with GPT Image.
And, as expected, Elon Musk put his visionary talents on full display, making predictions about the future of AI and its impact on people's lives.
AI will make work optional and money irrelevant.
Elon Musk is one of those people who can inspire both admiration and strong opposition at the same time. Nevertheless, we have to acknowledge his vision and willingness to pursue projects that many others would never even dare to imagine — PayPal, OpenAI, Neuralink, SpaceX, Tesla…
However, it's also worth remembering that Elon Musk is never short of words, especially when invited to play the role of "humanity's great guru". Not everything he says should necessarily be taken at face value.
We've listened carefully to the entire interview, and here's our perspective on the statements and predictions Elon Musk shared.
The Transformation
We believe Musk is right about the scale of the transformation, but probably less accurate about the timelines and, above all, about his assumption that technological progress will automatically lead to greater prosperity. After all, he has a vested interest in this transformation.
According to Musk, within approximately five years, AI could surpass the combined intelligence of all human beings. Within ten years, we could have lost effective control over these systems. The combination of artificial intelligence and robotics would make it possible to produce virtually any good or service, work would become optional, and money could eventually lose much of its purpose.
Ultimately, Elon Musk paints a picture of a society of technological abundance, in which machines take over much of the physical and intellectual work, while citizens receive something resembling a generous universal basic income.
We believe the transformation and its overall direction are real and that Musk is broadly right about them. But it won't happen quite so quickly, and we certainly can't simplify the economic, political and social challenges involved in getting there. His narrative is clearly influenced by the strategy and future he envisions for his own businesses.
Elon Musk: His Five Major Predictions
Elon Musk builds his argument and vision around five major predictions:
Artificial intelligence will significantly surpass human capabilities over the coming years.
Humanity could lose effective control over AI systems within approximately a decade.
The combination of AI and robots will generate extraordinary production capacity.
Work will become optional, and money will gradually lose its relevance.
China could eventually lead the global race in artificial intelligence and robotics.
However, these five predictions are far from equivalent or comparable in terms of maturity. The rapid advancement of AI and China's growing industrial influence are clearly observable trends. By contrast, the disappearance of work, money or scarcity relies on much more debatable economic and political assumptions.
AI Surpassing All Human Intelligence
Perhaps Musk's most striking claim is that, before 2031, artificial intelligence could surpass the combined intelligence of all human beings.
The problem is that there is no accepted way to measure this supposed collective intelligence.
Intelligence isn't a homogeneous quantity that we can simply add up and compare. This is something we constantly emphasise in our training sessions, particularly because human intelligence and so-called artificial intelligence are far from directly comparable.
Human intelligence encompasses knowledge, reasoning, creativity, perception, experience, intuition, social skills, judgement and contextual understanding. Measuring all of this in machines is anything but straightforward, and even if we could, we would still need to agree on a common measurement scale.
What is clearly evident is that AI systems are rapidly expanding the number, duration and complexity of the tasks they can perform. But that doesn't necessarily mean they're becoming more intelligent. According to METR, the length of software tasks that AI models can complete autonomously has been doubling approximately every seven months.
It's entirely possible that, before 2031, we'll have systems that outperform most people in programming, analysis, diagnosis, research, design or planning.
But outperforming a person across multiple tasks doesn't necessarily mean replacing an entire profession. This is something we explore in our training sessions using the Three-Bucket Method:
What AI can do autonomously on your behalf.
What you can do more effectively with AI.
What only you can do.
And while it's true that, in a society that rewards productivity and optimisation, the tendency is to focus on Bucket 1, there is still enormous potential to explore across virtually every profession in Buckets 2 and 3.
AI can deliver extraordinary results in controlled tests while still failing to generate comparable improvements within a real organisation.
Losing Control of AI
Musk suggests that, within a decade, human beings could lose control of artificial intelligence.
That statement immediately brings to mind an autonomous superintelligence — something straight out of The Matrix — capable of taking over.
But there is probably a more immediate risk, one that is far less cinematic and much more plausible: losing practical control without any machine ever needing to rebel.
Companies automate decisions that nobody knows how to review anymore.
Professionals delegate so much to AI that they stop developing their own judgement.
Public administrations use opaque algorithms to allocate rights, resources or services.
AI-generated information feeds other AI systems, multiplying errors.
The speed of automated decisions exceeds our ability to supervise them.
We focus almost exclusively on finding solutions that belong in Bucket 1 (and this has already happened in many organisations, some of which have subsequently had to reverse course).
So yes, losing control is a real possibility, but it doesn't necessarily have to happen through a confrontation between humans and machines. It could emerge through a gradual surrender of knowledge, autonomy and decision-making ability.
Will Work Disappear?
Elon Musk tells us that AI and robots will make work optional.
Automation could dramatically reduce human involvement in many activities. It's also reasonable to expect that much smaller teams will be able to achieve levels of output that currently require dozens or even hundreds of people.
But there's a significant gap between being able to automate a task and making that automation profitable, reliable, legal, insurable, socially acceptable and scalable.
The International Labour Organization estimates that one in four jobs has some degree of exposure to generative AI, but considers job transformation more likely than the complete disappearance of employment.
The relevant question isn't whether we'll stop working tomorrow. It's how professions will evolve, which tasks will lose value, which new ones will emerge and how the productivity gains will be distributed.
The Impact of Producing More
From our perspective, this is probably the weakest part of Musk's vision.
He assumes that enormous production capacity will lead to a society of abundance. However, technology can generate wealth without guaranteeing that this wealth will be distributed.
We could produce far more and, at the same time, live in a more unequal society if AI models, robots, chips, data, energy and infrastructure are controlled by a small number of companies.
Artificial intelligence doesn't automatically eliminate scarcity in areas such as:
Housing and land.
Energy and raw materials.
Attention and time.
Security and privacy.
Exclusive goods.
Natural resources.
Influence and power.
As long as scarce goods exist, some mechanism will still be needed to allocate them. It could be money or another system, but competition for access to those resources is unlikely to disappear.
A generous universal basic income (something some politicians have proposed and which has even been tested on a small scale in Finland) wouldn't automatically result from technological innovation either. It would require political decisions, tax systems, redistribution mechanisms and new agreements on the ownership of technological capital.
The International Monetary Fund estimates that approximately 40% of global employment is exposed to AI, but stresses that the outcome will depend on public policies, workforce training and how productivity gains are distributed.
Before reaching a hypothetical society of abundance, which seems rather utopian, we could face a lengthy transition characterised by the disappearance of entry-level jobs, wage polarisation, corporate concentration, pressure on tax systems and growing social tensions.
Technology can generate productive abundance. However, turning it into shared prosperity will be a political and social decision.
China's Influence
This is perhaps one of the most convincing arguments in the interview, although it's worth pointing out that there is already plenty of solid evidence supporting it, even for those with limited knowledge of the subject.
The AI race won't be determined solely by who develops the most advanced model. It will also depend heavily on the ability to manufacture chips, generate energy, produce robots, build data centres and integrate artificial intelligence into industry.
China has an extraordinary advantage when it comes to connecting manufacturing, artificial intelligence and robotics. According to the International Federation of Robotics, it accounts for 54% of new industrial robot installations worldwide and already has more than two million robots operating in its factories.
The United States may retain its leadership in frontier AI models (the digital interface), but China appears particularly well positioned to translate AI into the physical world and deploy it at scale.
Energy is also becoming a strategic factor. The International Energy Agency estimates that global electricity consumption by data centres will almost double between 2025 and 2030. Here too, China has a considerable advantage over the rest of the world in its capacity to generate electricity.
The next phase of artificial intelligence will therefore not be determined by algorithms alone. It will also be decided in factories, power grids, supply chains and the industrial capabilities of individual countries.
In Conclusion
We must remember that Elon Musk is a visionary with vested interests, and his predictions are inevitably "coloured" by them.
His vision and predictions align almost perfectly with the areas in which his companies operate:
Artificial intelligence: xAI.
Robotics: Tesla and Optimus.
Energy and storage: Tesla Energy.
Connectivity: Starlink.
Space infrastructure: SpaceX.
Autonomous mobility: Tesla.
Elon Musk presents a narrative in which his companies occupy a central position as providers of the infrastructure needed to make that future a reality.
This doesn't invalidate everything he says, but it does mean we need to examine his arguments carefully, separating his vision of technological evolution from his business interests.
Furthermore, three recurring biases emerge throughout his argument and deserve careful consideration:
Technological determinism: If a technology can be developed, he tends to assume it will eventually be deployed on a massive scale.
Timeline compression: He turns long-term possibilities into predictions expected to materialise within five or ten years.
Underestimating institutional challenges: He treats regulation, redistribution and social transformation as secondary issues, even though they could prove more difficult to resolve than the technology itself.
From our perspective, it doesn't seem reasonable to build a strategic plan on the assumption that work or money will disappear within the next decade.
However, we should act on the assumption that profound transformations will take place between 2026 and 2031:
Automation will move from isolated tasks to entire processes.
Smaller teams will achieve much greater operational capacity.
Generic content and analysis will lose value.
Many junior positions will disappear or change radically.
Proprietary data will become a strategic asset.
Dependence on model providers, chips, platforms and energy will increase.
Pressure on the prices of professional and digital services will grow.
Trust, brand and execution capabilities will become even stronger differentiators.
Technology will increasingly become part of our everyday lives, and it will do so at considerable speed. The key challenge for professionals and brands will be how they integrate it and, above all, how they make it work alongside the human element and strengthen it.
This isn't simply about introducing new tools. Companies need to review which processes can be automated, what knowledge they must protect, which decisions require human supervision and what value proposition they will be able to defend when operational intelligence becomes significantly cheaper.
We need to understand that we're entering a world where certain intellectual and operational capabilities will become far more accessible. Activities that currently require specialised teams will be performed in less time, with fewer resources and at a much lower cost.
But artificial intelligence won't automatically eliminate responsibility, trust, judgement, relationships, purposeful creativity or the ability to turn an idea into a sustainable reality. People will continue to play — and must continue to play — a fundamental role.
Elon Musk is probably right about one thing: the scale of the transformation could be far greater than we currently imagine.
But whether that transformation creates a more prosperous, balanced and humane society will not depend on technology alone. It will depend on the decisions we make about how we design, use and govern it, and how we distribute the value it can create.
Comments