Synthetic Minds | The Search for Climate Fixes Becomes a Design Brief
Synthetic Minds | The Search for Climate Fixes Becomes a Design Brief
The Synthetic Minds newsletter offers short daily insights to get you thinking. If you enjoy it, please forward. All signals are powered by Futurwise. If you need more insights, subscribe to Futurwise and get 25% off for the first three months!
I have just launched the Intelligence Age Scorecard! It will help you understand how ready your organization is for the Intelligence Age.
Today’s topic: Climate & Energy
Designing the Microbes That Repair the Climate
An AI trained on the language of DNA has written 16 complete genomes from scratch, and every one came alive. For the parts of climate work that run on biology, the slowest step has fallen away.
Set beside what the same models are being aimed at, one shift comes into focus: climate biology is turning from a slow search through nature into a design problem.
Researchers at the Arc Institute and Stanford designed working viral genomes with AI that reads DNA the way a chatbot reads text. The model proposed roughly 700,000 candidates; the team built 285, and 16 came alive.
The proof matters more than the viruses. It shows AI can design whole living systems, not only tweak single genes.
And the same model family has read the genetic code across the tree of life, from bacteria to plants, the organisms climate work actually leans on.
AI-designed enzymes already break down PET plastic in hours instead of centuries, and even create self-destructing plastic that leaves no microplastics.
Other models comb millions of compounds to find better materials for pulling carbon from the air.
Each result has landed on its own. Together they point one way: the biology of climate repair is becoming something we design.
That's the biotech meets climate change story. Here is the signal.
For a decade, the hardest climate problems have been biology problems in disguise: capturing carbon and holding it, breaking down plastic that outlives us, growing food without flooding the air with methane.
Nature can do all three. Only far too slowly, and never on command.
The bottleneck has always been the search. Finding the right microbe or enzyme meant combing through whatever evolution happened to leave behind, then nudging it by trial and error across years of lab work.
That bottleneck is lifting. When a model can write a genome that works on the first serious attempt, the search becomes a design brief. The question moves from what nature has already made to what we actually need built.
For climate research, this compresses time in a way funding alone never could. A carbon-fixing organism, a plastic-eating enzyme, a crop that needs less fertilizer, each has been a decade-long hunt. Design turns decades into iterations.
The energy-layer contest over who can keep clean power flowing named one part of the stack. This is the layer beneath it: the living machinery that turns sunlight, water and carbon into something useful.
The responsibility grows with the capability, and it deserves care. But the headline is opportunity. The tools that repair the climate are starting to be things we design, not things we wait to stumble upon.
The question we should be asking is not whether to capture more carbon. It is which of these living solutions to help build first.
For a century we mined the planet for what it could give us. The next chapter is learning to design the living things that give it back.
The Intelligence Age Scorecard

The biology of climate repair is turning from a slow search through nature into something researchers can design on demand. Carbon-fixing microbes, plastic-eating enzymes, lower-methane crops. The WAVE Framework, Watch, Adapt, Verify, Empower, asks which move this demands of you: are you still watching biology from a distance, or should you already be adapting your research and partnership strategy to a design-led decade?
Benchmark your readiness for the next two quarters with the Intelligence Age Scorecard. Or read the public Intelligence Age Scorecard of Verizon, Accenture, IBM, Visa, Qantas, Woolworths, Telstra or Commonwealth Bank first.
If this newsletter was forwarded to you, you can sign up here.
Thank you.
Mark
Frequently asked questions
What did researchers at the Arc Institute and Stanford achieve?
They used an AI that reads DNA the way a chatbot reads text to design working viral genomes from scratch. The model proposed roughly 700,000 candidates, the team built 285 of them, and 16 came alive as complete, functioning genomes, showing AI can design whole living systems rather than just tweak single genes.
Link to this questionWhy does AI-designed biology matter for climate change?
For a decade the hardest climate problems, capturing and holding carbon, breaking down long-lasting plastic, and growing food without heavy methane emissions, have really been biology problems. Nature can handle all three but very slowly and never on command. AI models that can design genomes, enzymes and compounds turn the old slow search through nature into a design brief, compressing decade-long hunts into fast iterations.
Link to this questionWhat other AI-driven climate biology breakthroughs are mentioned?
AI-designed enzymes already break down PET plastic in hours instead of centuries, and researchers have even created self-destructing plastic that leaves no microplastics. Separately, other AI models search through millions of compounds to identify better materials for pulling carbon out of the air, extending the same design-based approach beyond genomes to broader climate materials.
Link to this questionHow does AI change the search for climate solutions?
Previously, finding the right microbe or enzyme meant sifting through whatever evolution happened to produce, then refining it through years of trial-and-error lab work. Now that AI models can write a working genome on the first serious attempt, that search becomes a design process. The focus shifts from what nature already made to what we actually need to build, compressing decades of research into faster iterations.
Link to this question