How fusion energy, satellite networks, AI agents and humanoid robots signal a new phase in human progress


Every now and then, several breakthroughs happen almost at once — separate stories that, when viewed together, mark a decisive shift in how technology and society evolve.

This week feels like one of those moments.

Google announced a milestone in fusion energy. Apple teamed up with SpaceX to connect iPhones directly to Starlink satellites. OpenAI unveiled Atlas, a new interface designed to make intelligent agents accessible to everyone. And a Norwegian start-up introduced Neo, the first humanoid robot aimed at consumers.

Individually, each story is impressive. Taken together, they reveal something deeper: the world’s next transformation will come from the fusion — quite literally — of clean energy, autonomous intelligence and physical automation.

The dawn of practical fusion

Google’s latest move in the energy sector has sparked excitement — and a fair bit of confusion. Contrary to some headlines, the company hasn’t yet “cracked” fusion power. What it has done is partner with a leading start-up to buy electricity from one of the world’s first commercial fusion reactors, expected later this decade.

It’s an important signal. Tech giants are no longer waiting for governments to make fusion real — they’re investing directly in it. And for good reason. If fusion delivers even a fraction of its promise, the implications are staggering: abundant, clean, carbon-free energy that could power data centres, AI infrastructure and space operations without environmental compromise.

The future of digital technology may depend less on processors and more on plasma.

iPhones that talk to space

Meanwhile, Apple and SpaceX have confirmed that they’re exploring direct satellite connectivity for iPhones using Starlink’s low-orbit network. If successful, it could render mobile dead zones a thing of the past — and perhaps, in time, force telecom operators to reinvent themselves.

For users, it means global, seamless connectivity. For industries such as defence, aviation and emergency response, it offers something even more valuable: resilience. When ground networks fail, a satellite link can keep critical systems alive.

In an era where connectivity equals security, this partnership may prove as transformative as the smartphone itself.

OpenAI’s Atlas: when AI starts doing things

With Atlas, OpenAI has taken another bold step towards what it calls “agentic AI”.

Instead of waiting for human prompts, these agents can act independently: search the web, organise data, draft plans, even perform small digital tasks — all within a browser.

If ChatGPT was a conversational partner, Atlas is a capable assistant. It marks the transition from passive to proactive AI — from a tool that answers questions to one that takes action.

For businesses, governments and research institutions, the change is profound. Workflows once spread across dozens of apps could soon be handled by a single intelligent layer, continuously learning and improving.

Automation is about to become much more personal.

The age of humanoids begins

And then there’s Neo. Norwegian robotics firm 1X Technologies has unveiled what it calls the first humanoid robot designed for everyday life — a machine that can move, perceive and assist within the home.

It’s not science fiction anymore: Neo will reportedly cost around $20,000 and begin shipping next year. Like the first personal computers of the 1980s, it may start as a niche curiosity but evolve into a transformative presence.

Beyond domestic chores, humanoid robots are expected to play a role in logistics, healthcare and security — areas where automation meets empathy and precision. The physical world is finally becoming programmable.

A quiet revolution — already underway

Viewed together, these developments point towards a common horizon: autonomy.

Autonomy of energy, autonomy of connectivity, autonomy of intelligence and, ultimately, autonomy of action.

And that raises an important question for Luxembourg: how do we position ourselves in this accelerating race?

What Luxembourg must do to stay in the game

Luxembourg has already demonstrated its ability to act early in emerging fields — from satellite communications and space resources to cybersecurity and high-performance computing. But the next chapter demands even more agility.

To stay competitive, Luxembourg should:

• Invest boldly in fusion and clean-energy R&D, ensuring that tomorrow’s datacentres and AI infrastructures can operate sustainably within Europe.
• Accelerate partnerships in low-orbit and sovereign connectivity, leveraging the country’s unique legacy in satellite operations and regulatory diplomacy.
• Adopt agentic AI at institutional level, modernising public administration, defence and research ecosystems with autonomous analytical capabilities.
• Position robotics and human-machine collaboration as a national innovation pillar, linked to healthtech, logistics and dual-use applications.

These are not isolated initiatives. Together, they define what strategic autonomy truly means: the ability to innovate and operate independently while contributing to Europe’s collective strength.

If Luxembourg acts now — bridging its industrial heritage with its digital vision — it can once again punch far above its size in shaping the technologies that will define the coming decade.

Published by Direct Engineering S.à r.l.

Advisory on Strategic Autonomy, Defence & Space Innovation

(October 2025)