Sophie Hackford

A lot of money will be spent on physical intelligence infrastructure in the next decade –  upwards of$7 trillion. This has already started re-orienting supply chains and geopolitics, and making a very real imprint on our planet: Meta’s Prometheus data-centre cluster will be almost the size of Manhattan.

What are the consequences will this massive injection of capital and enthusiam, beyond AI? Can we expect cheap, tested compute components to flood the market? And an acceleration of once-unfashionable energy technologies, like fusion or geothermal? The 2nd and 3rd order consequences of this capex explosion will lead to tremendous innovations in unexpected places. Perhaps a return to supersonic flight? 

These new compute and energy facilities will be robot-built and robot-maintained. We might finally see gainfully employed robots out in the real world. How shall we design for, and inhabit, these landscape-scale robotics?

If AI to evolve, we don’t just need more data, but a new class of data collected by a new class of sensors – ones co-evolved between compute and biology.

If we feed AI with human-centric data – as we do today – it can only access a limited portion of the vast hypersensory space available to many other species: if we can’t follow chemical trails in the oceans like a mysid, or navigate like a bat, perhaps our computers can? 

Great strides have been made in the past 2 decades in computer vision – but our sight is tangled up with our cognition. AI might free us from what language does to us – and allow us to see, hear, smell, taste, and sense the world in glorious technicolour. 

Sophie will explore how navigation, disease diagnostics, weather forecasting… will be improved by novel sensing technologies. 

Computation is not a human construct. After 3bn years of evolution, there are an incredible number of models of computation out there – most of which are new to science, and process information in an entirely different way. 

Language models are amazing. But if we want to understand our planet, what humans say about it is tiny and imperfect. How we compute about our planet is tiny and imperfect. 

Experimental drones already run on biological computers built from cultured mouse brain cells. Singapore is building a data-centre with racks of cultured human neurons where you’d expect silicon to be. We might be able to grow leafy supercomputers everywhere. These are not aiming to replace silicon, but be part of the orchestra of available compute.

AI is revealing physics, chemistry, and biology to us in thrilling ways and might enable us to build computers as Feynman predicted. Computers that operate on the same principles as – for example -penicillin operates itself. 

The space economy is a puzzle and few can see the full picture. 2026 is seen as an inflection point in the story of the space economy. 

Space is at the centre of global resilience. The more skyscraper-sized rockets launched into space, the more Earth’s economy relies on orbiting robotic infrastructure. It is getting very busy up there… 

The latest buzz is around million-satellite data centres in space. Space is becoming a computing domain, but also an industrial one. Orbiting autonomous warehousing, zero-G manufacturing, pharmaceutical R&D, and energy generation will accelerate as launch continues to fall in price.  

Some see space as a transport logistics layer for Earth – ‘the last 62-mile delivery’ – keeping military essentials like batteries in pods, for delivery anywhere on Earth within an hour. Musk has long promised to transport humans from Sydney to NY outside of the Earth’s atmosphere. 

We may have to share strategies with Lunar / Martian architects in designing habitats for extreme environments – housing in Texas with 45 degree summers. Will we need Earthsuits in parts of the world? 

The off-world economy may dwarf that of Earth’s in our lifetimes. Sophie will explore the competing interests, from those who see it as a theatre of war, trillion-dollar economy, mysterious scientific frontier, or sci-fi dream. 

We do not have mastery of the quantum world yet. But the cutting edge of quantum R&D is full of promise for AI, biotech, compute, security + codebreaking, materials science, sensing, military technologies, and much more. 

Today compute is (relatively) easy, it’s the framing of the problem that is where competitive advantage lies. Then it’s just an optimisation problem which classical and quantum are both seeking to solve.

Quantum optimisation R&D has led to a renaissance in classical compute optimisation, but we still act as if we don’t believe in the story of the last half-century of compute. Take-off in compute is always happening (quantum or not), and we have only just begun on this journey.

Sophie will tackle some of these questions:

– What are quantum computers? A little on the science, the weirdness, the promise.
– What other quantum hardware is being developed? (Sensors (for GPS alternatives), satellites, quantum radar, a quantum internet…)
– What to do as a company now?