Category: Insights

Wednesday June 3, 2026. Steve Walker

From AI Digital to AI Carbon

Investing in the Real World

Artificial intelligence is usually described as a digital revolution. The public story is about larger models, faster chips, bigger data centres and the race between technology companies. That story is true, but it is incomplete.

The next phase of AI will not stay inside screens, clouds or language models. It will move into the physical world. It will manage energy, optimise carbon, route feedstocks, predict supply chains, control fermentation, balance power demand and help convert waste into useful products.

That is the shift from AI Digital to AI Carbon.

AI Digital changes what we know, write, model and decide. AI Carbon changes what we make.

This distinction matters because civilisation is not built from data alone. It is built from fuels, chemicals, materials, nutrients, infrastructure and energy systems. The digital world may guide decisions, but the physical world still determines whether aircraft fly, homes are heated, food is produced, factories operate and economies remain sovereign.

The question is no longer only who has the largest model. The question is who will build the industrial platforms that AI can operate.

Carbon is not the enemy

For decades, carbon has been treated mainly as a pollution problem. That view is understandable, but it is too narrow. Carbon is also the foundation of industrial civilisation.

Carbon appears in fuels, plastics, solvents, textiles, fertilisers, packaging, proteins, pharmaceuticals, building materials and thousands of chemical intermediates. Modern economies do not need a world without carbon. They need a world where carbon is recovered, routed, converted and reused intelligently.

The problem is not carbon itself. The problem is linear carbon.

Extract it. Burn it. Emit it. Waste it.

Warsaw 05:06:2006 Steve Walker

For almost 150 years, industrial civilisation has been built around one dominant idea:

If carbon contains energy, burn it.

Coal powered the industrial revolution. Oil powered transport. Gas powered modern electricity systems. Entire cities, industries and economies were designed around combustion because combustion released energy quickly and at enormous scale.

But combustion also destroys something.

The moment carbon burns, its molecular structure disappears. Heat is released once, then the carbon leaves the system as exhaust, ash or emissions. For generations this looked perfectly rational because energy itself was the prize. Nobody seriously asked whether carbon might hold more value before it entered the furnace.

That assumption is beginning to break.

The modern industrial economy increasingly depends not only on energy, but on molecules. Fuels matter, but so do solvents, gases, polymers, chemicals, materials and biological feedstocks. A tonne of carbon-rich material may now hold greater value as industrial feedstock than as immediate heat.

This changes the industrial question completely.

The old economy asked:

How much energy can we release from carbon?

The emerging economy asks:

What is the highest possible value this carbon can become before we burn it?

That is where TITAN diverges from the traditional furnace economy.

TITAN does not simply burn carbon. It first converts carbon into hydrogen producer gas. Once carbon becomes gas, something important happens. The carbon is no longer trapped inside solid biomass. It becomes routeable.

That routing changes everything.

The same carbon stream may become electricity today, biomethane tomorrow, ethanol next week, or eventually SAF intermediates, solvents, ketones, acids or other industrial molecules. Instead of destroying molecular value immediately through combustion, TITAN attempts to preserve optionality for as long as possible.

This is the beginning of a very different industrial philosophy.

The furnace economy treats carbon as something to consume.

AI Carbon treats carbon as something to optimise.

That distinction may define the next industrial era.

Historically, combustion won because it was simple and cheap. Oil and gas became abundant. Giant centralised refineries and power systems dominated the world. Biological pathways, fermentation routes and alternative carbon systems were pushed aside because the economics of hydrocarbons overwhelmed almost everything else.

AI Digital Needs AI Carbon: Warsaw 02:06:2026 Steve Walker

Artificial intelligence has become the defining investment story of our age.

Every week brings larger models, faster chips, bigger data centres and new claims about how intelligence will transform civilisation. Governments discuss AI sovereignty. Markets compete to identify the next trillion-dollar company. Technology firms race to secure compute power, electricity and infrastructure at extraordinary scale.

The excitement is real.

Artificial intelligence is already changing engineering, medicine, logistics, research, finance and industrial planning. It compresses time, accelerates discovery and expands access to knowledge at speeds never before possible.

But beneath the excitement sits a deeper question.

How does AI ultimately pay back?

Not in theory.

In the real economy.

For those old enough to remember the dot-com era, the feeling is familiar. The internet genuinely changed the world. Entire industries were rebuilt. Commerce, communication and media transformed permanently.

But when the excitement faded, investors eventually returned to a harder question.

What exactly did we invest in?

Many companies had users, traffic and attention, but no durable economic foundation beneath them. The promise was real. The monetisation was weak.

Artificial intelligence now risks facing a similar moment.

Not because AI lacks importance.

But because the scale of investment now flowing into artificial intelligence requires a route into the physical economy large enough to justify it.

This is the contradiction emerging inside the AI race.

Artificial intelligence appears weightless from the user’s perspective, but the infrastructure behind it is anything but weightless. Data centres require enormous quantities of electricity. Semiconductor manufacturing depends upon highly specialised industrial supply chains. Cooling systems require water, energy and materials. Compute clusters require construction, maintenance, logistics and grid infrastructure.

The more powerful AI becomes, the more physical the story becomes.

AI Digital potrzebuje AI Carbon Warsaw 02:06:2026 Steve Walker

Sztuczna inteligencja stała się najważniejszą historią inwestycyjną naszych czasów.

Każdy tydzień przynosi większe modele, szybsze układy scalone, większe centra danych oraz nowe deklaracje o tym, jak inteligencja zmieni cywilizację. Rządy dyskutują o suwerenności AI. Rynki próbują wskazać kolejną firmę wartą bilion dolarów. Firmy technologiczne rywalizują o moc obliczeniową, energię elektryczną oraz infrastrukturę w skali niespotykanej wcześniej.

Ekscytacja jest prawdziwa.

Sztuczna inteligencja już dziś zmienia inżynierię, medycynę, logistykę, badania naukowe, finanse i planowanie przemysłowe. Skraca czas, przyspiesza odkrycia i rozszerza dostęp do wiedzy w tempie, którego wcześniej nie znaliśmy.

Jednak pod tą ekscytacją kryje się głębsze pytanie.

W jaki sposób AI ostatecznie się zwróci?

Nie w teorii.

W realnej gospodarce.

Dla tych, którzy pamiętają erę dot-comów, to uczucie jest znajome. Internet rzeczywiście zmienił świat. Całe branże zostały przebudowane. Handel, komunikacja i media zmieniły się na zawsze.

Ale gdy pierwsza fala entuzjazmu opadła, inwestorzy wrócili do trudniejszego pytania.

W co dokładnie zainwestowaliśmy?

Wiele firm miało użytkowników, ruch i uwagę rynku, ale nie posiadało trwałych fundamentów ekonomicznych. Obietnica była prawdziwa. Monetyzacja okazała się słaba.

Sztuczna inteligencja może dziś stanąć przed podobnym momentem.

Nie dlatego, że AI jest nieważna.

Lecz dlatego, że skala inwestycji kierowanych obecnie do AI wymaga wejścia do fizycznej gospodarki na tyle dużego, aby uzasadnić ten poziom kapitału.

To właśnie jest sprzeczność pojawiająca się dziś w wyścigu AI.

Z perspektywy użytkownika sztuczna inteligencja wydaje się niematerialna, ale infrastruktura stojąca za nią jest całkowicie fizyczna. Centra danych wymagają ogromnych ilości energii elektrycznej. Produkcja półprzewodników zależy od wysoce wyspecjalizowanych łańcuchów przemysłowych. Systemy chłodzenia potrzebują wody, energii i materiałów. Klastry obliczeniowe wymagają budowy, utrzymania, logistyki i infrastruktury sieciowej.

Im potężniejsza staje się AI, tym bardziej fizyczna staje się cała historia.