As Elon Musk once said, “If you can think of it, you can have it,” and he imagines a future where the grind is optional and even wealth itself might fade into irrelevance. In this conversation with Nikhil Kamath, his forecast hangs on the accelerating convergence of AI and robotics-a technological wave he believes will crest within the next 10 to 20 years.
Musk puts the looming transformation in perspective: Some people will do it out of pleasure, like people who like to tend to their gardens, but there will basically be no need. This is a vision based on the development of autonomous systems-both cognitive AI “agents” and physical robots-that would be capable of doing most of the tasks that humans perform today. Available studies indicate that already-demonstrated technologies could theoretically automate about 57% of U.S. work hours: agents doing complex reasoning and information processing, while robots would do increasingly dexterous physical work.
The implications go far beyond labour. Musk believes that AI and robotics would eventually “run out of things to do to make humans happy”, satiating every conceivable human desire. Beyond that, machines might work for other machines – an economy of algorithms and actuators, operating independently of humans. In such a post-scarcity model – think of the societies imagined in Iain M. Banks’ Culture novels – money itself might cease to matter. Goods and services would be plentiful, but physical constraints like energy and mass would remain, binding even the most utopian visions to the laws of physics.
Those constraints are already surfacing in the power needs of AI. The IEA estimates data centres worldwide will more than double their electricity consumption by 2030 to 945 terawatt-hours, more than Japan currently uses. AI-optimized facilities alone can increase their power draw fourfold in the same time frame. In the United States, data centers are forecasted to account for nearly half the overall growth in power demand, outstripping the use of energy-intensive manufacturing combined. Meeting this surge will take rapid deployment of gas, solar, and storage alongside strategic investment in nuclear power to provide the high-capacity, low-carbon baseload generation critical for hyperscale AI clusters.
Advanced robotics, by contrast, are on a much slower arc. Physical work still requires a degree of fine motor skills and situational awareness that machines struggle to reliably replicate. Economists also say that while the cost of AI falls-token processing rates have fallen from $10 to $2.50 per million tokens in one year-robotics remain expensive and specialized, limiting the scalability possible in the near term. Yet Musk’s own Optimus humanoid robot program aims to overcome these barriers, with a target of contributing 80% of Tesla’s future value, despite production delays.
The transformation Musk has in mind also includes reengineering workflows rather than merely automating tasks. For instance, McKinsey research suggests that capturing AI’s potential $2.9 trillion annual economic value in the U.S. by 2030 will depend on redesigning processes to integrate people, agents, and robots. That means scaling human activities from execution to orchestration-framing problems, guiding AI outputs, and applying judgment-while machines perform routine operations. Demand for “AI fluency” as a skill has grown sevenfold in two years, making it the fastest-rising skill in job postings.
Even in a world of optional work, energy realities will enforce the tempo. Huge AI systems and fleets of robots will require enormous and dependable power. Projects like Project Stargate’s planned 5 GW data centre in Texas, powered with solar, storage, and on-site gas generation, illustrate the scale. Nuclear restarts, such as the 835 MW Three Mile Island Unit 1, and advanced reactor designs from firms like TerraPower could anchor future “nuclear computation hubs” pairing gigawatt-scale AI clusters with dedicated generation.
Musk’s timeline – 10 to 20 years – may be ambitious given engineering, economic, and political hurdles. But the trajectory is clear: AI and robotics are expanding into every corner of productivity, from administrative work to hazardous physical tasks. As they do, the definition of “job”, “income”, and even “currency” may disappear, replaced instead by a whole new calculus of meaning in a world where the machines do the work and the humans decide whether to join in at all.
