In partnership with
Hey everyone, it’s Thursday, the inbox is humming, and FutureProof is back.
Before we jump in, a quick note. Reader Glenn G emailed this week to say this is “the best structured, smart and positive motivating asset” they read, and that the Why This Matters and Kismet sections are their favourites.
That genuinely meant a lot.
Because the whole point of this newsletter is signal over noise. Clarity over chaos. Progress over performative outrage.
So thank you, Glenn.
Now, onto a week where the transition didn’t just inch forward. It accelerated.
The big themes:
Electricity demand is surging globally, and renewables are meeting almost all of it.
Battery innovation is compounding, from sodium-ion to million-mile packs.
EVs are quietly becoming resilience infrastructure.
Construction is discovering that 98% of its emissions sit upstream, and that resilience and sustainability are the same conversation.
And yes, Australia just cut power prices as renewables scale.
Let’s get into it.
Climate
China’s Emissions Plateau for 21 Months. Peak Already Behind Us?
China’s CO₂ emissions fell 1% in Q4 2025, likely locking in a 0.3% drop for the year - extending a “flat or falling” streak that’s now lasted 21 months. And the best bit? This isn’t recession-driven. Clean power growth is outpacing electricity demand.
Key highlights:
Solar output surged 43% year-on-year, wind 14%, nuclear 8% - together adding ~530TWh, more than covering demand growth.
Power-sector emissions fell 1.5%, transport 3%, cement and building materials 7%; chemicals was the outlier, up 12%.
Energy storage jumped 75GW in 2025 - more than peak demand growth (55GW), a potential structural inflection point.
Why This Matters:: For the first time, China’s emissions are flat or falling without a slowdown in energy demand - a sign that structural decarbonisation, not economic weakness, may finally be doing the heavy lifting.
Kismet: Clean energy drove more than a third of China’s GDP growth in 2025 - meaning the climate transition isn’t a drag on the economy; it’s increasingly the engine. And that excellent news, not just for China, but for all of us. 👉 Full story here
Romania Slashes Emissions 75% While Doubling GDP. Decoupling Just Got Real.
If you want a live case study in economic growth without carbon bloat, look at Romania. Since 1990, emissions have plunged 75% while real GDP has doubled - and emissions intensity has fallen a staggering 88%. For a country once defined by coal, oil and heavy industry, that’s not incremental. That’s unprecedented systemic change.
Key highlights:
Net greenhouse gas emissions down 75% since 1990; emissions intensity down 88%.
Power-sector carbon intensity fell 52% in the 17 years after EU accession, turbocharged by EU carbon pricing and renewables support.
Nature played a role too: carbon absorbed by land use increased 77% as forests expanded onto abandoned land.
Why This Matters:
Romania shows that absolute decoupling at scale is possible, even for a middle-income, fossil-fuel-heavy economy.
Kismet: Romania now emits just 3 tonnes of greenhouse gases per person - lower than almost every other EU country, with only Sweden polluting less. 👉 Full story here
Britain’s Clean Power Sprint Just Hit the Turbo Button
For years, the UK’s 2030 clean grid target has sounded like political theatre. This week, it started to look like engineering. In its latest auction round, Britain secured 6.2GW of new solar and onshore wind at prices roughly 50% cheaper than building new gas.
Record solar haul: 4.9GW of solar awarded at ~£65/MWh, lower than the previous round.
Onshore wind revival: 1.3GW secured as England’s long-standing de facto ban fades into history.
Batteries scaling fast: Grid-scale storage is expected to jump from ~5GW today to ~25GW by 2030 - storing the midday sun for the evening peak.
But here’s the real story. Decarbonising power isn’t just about building turbines. It’s about rewiring Britain. Roughly 1,000km of new onshore lines. 4,500km offshore. Converter stations. Planning fights. Regulatory “uncertainty mechanisms.” Gas plants running 20–30% of the time even in a mostly clean system. This is infrastructure surgery on a century-old machine. And yet, the economics now vastly favour renewables. And that’s a permanent shift. Globally.
Why This Matters: For the first time, the UK’s 2030 clean power mission feels less like ambition and more like a construction timetable.
Kismet: If Britain reaches ~25GW of batteries by 2030, it will have more grid-scale storage than the entire planet had in the early 2020s - a reminder that exponential curves look implausible right up until they’re obvious. 👉 Full story here
AI News
Beyond AlphaFold: AI Just Took a Serious Step Toward Designing Drugs From Scratch
We’ve been talking about AI writing code, drafting emails, generating images. Meanwhile, in a much quieter corner of the universe, it’s starting to design medicines. Isomorphic Labs has unveiled its Drug Design Engine, and the claim is bold: it goes beyond AlphaFold 3, more than doubling accuracy on the hardest protein–ligand generalisation tests, especially where models traditionally struggle most - novel systems that look nothing like their training data.
Hardest benchmark leap: On the “Runs N’ Poses” generalisation test, Isomorphic Labs Drug Design Engine (IsoDDE) more than doubled AlphaFold 3’s accuracy on the most dissimilar protein–ligand systems.
Binding affinity breakthrough: It outperformed existing deep-learning models and even surpassed physics-based gold standards like FEP - at a fraction of the time and computational cost.
Blind pocket detection: It can identify novel, ligandable binding pockets from amino acid sequence alone, including cryptic sites previously found only through lab experiments.
This matters because predicting structure is one thing. Predicting how strongly something binds, and spotting entirely new druggable pockets before you step into a wet lab, is where time and billions are saved. The traditional drug discovery timeline runs 10–15 years. If AI meaningfully compresses that, the downstream implications are enormous for biotech, pharma economics, and global health.
Why This Matters: We may be witnessing the shift from AI as an analytical assistant to AI as a generative engine for new molecular matter - compressing the distance between hypothesis and therapy.
Kismet:: AlphaFold’s database has already been used by over 3 million researchers across 190+ countries, meaning the computational foundation for AI-designed medicines is already more globally distributed than most clinical trial infrastructure.
👉 Full story here
AI Is Now Building Family Trees for Icebergs (Yes, Really)
Icebergs don’t just float. They fracture, fragment, disperse, melt - and until now, once they shattered into thousands of pieces, much of their climate impact effectively disappeared from our models. British Antarctic Survey scientists have developed a world-first AI system that automatically identifies icebergs at birth, names them, and tracks their entire lifecycle - even linking smaller “child” fragments back to their original “parent” iceberg to build full family trees.
Major modelling gap closed: Scientists have long struggled to track fragmented icebergs once they break apart - a significant blind spot in climate prediction.
AI-powered tracking: Using satellite imagery, the system recognises iceberg shapes as they calve and follows their journey, sometimes over decades.
Climate impact clarity: Tracking where and when freshwater melts into the ocean improves understanding of ocean currents, ecosystems, and global climate feedbacks.
This is bigger than it sounds. Icebergs release vast volumes of freshwater as they melt, influencing ocean circulation patterns that regulate global climate. If you can’t accurately track that freshwater pulse, your climate projections carry uncertainty. Now we can. And there’s a bonus: the same system could improve polar navigation safety in increasingly trafficked Arctic and Antarctic waters.
Why This Matters: Better tracking of iceberg fragmentation sharpens climate models at a moment when ice loss is accelerating, reducing uncertainty in one of the most sensitive parts of the Earth system.
Kismet: For the first time, scientists can trace an iceberg fragment drifting past a ship back to the exact glacier it calved from years earlier, effectively giving melting ice a passport and a family history. 👉 Full story here
Icebergs don’t just float. They fracture, fragment, disperse, melt — and until now, once they shattered into thousands of pieces, much of their climate impact effectively disappeared from our models.
British Antarctic Survey scientists have developed a world-first AI system that automatically identifies icebergs at birth, names them, and tracks their entire lifecycle — even linking smaller “child” fragments back to their original “parent” iceberg to build full family trees.
Major modelling gap closed: Scientists have long struggled to track fragmented icebergs once they break apart — a significant blind spot in climate prediction.
AI-powered tracking: Using satellite imagery, the system recognises iceberg shapes as they calve and follows their journey, sometimes over decades.
Climate impact clarity: Tracking where and when freshwater melts into the ocean improves understanding of ocean currents, ecosystems, and global climate feedbacks.
This is bigger than it sounds. Icebergs release vast volumes of freshwater as they melt, influencing ocean circulation patterns that regulate global climate. If you can’t accurately track that freshwater pulse, your climate projections carry uncertainty. Now we can.
And there’s a bonus: the same system could improve polar navigation safety in increasingly trafficked Arctic and Antarctic waters.
Why This Matters:
Better tracking of iceberg fragmentation sharpens climate models at a moment when ice loss is accelerating — reducing uncertainty in one of the most sensitive parts of the Earth system.
Kismet:
For the first time, scientists can trace an iceberg fragment drifting past a ship back to the exact glacier it calved from years earlier — effectively giving melting ice a passport and a family history.
Icebergs don’t just float. They fracture, fragment, disperse, melt — and until now, once they shattered into thousands of pieces, much of their climate impact effectively disappeared from our models.
British Antarctic Survey scientists have developed a world-first AI system that automatically identifies icebergs at birth, names them, and tracks their entire lifecycle — even linking smaller “child” fragments back to their original “parent” iceberg to build full family trees.
Major modelling gap closed: Scientists have long struggled to track fragmented icebergs once they break apart — a significant blind spot in climate prediction.
AI-powered tracking: Using satellite imagery, the system recognises iceberg shapes as they calve and follows their journey, sometimes over decades.
Climate impact clarity: Tracking where and when freshwater melts into the ocean improves understanding of ocean currents, ecosystems, and global climate feedbacks.
This is bigger than it sounds. Icebergs release vast volumes of freshwater as they melt, influencing ocean circulation patterns that regulate global climate. If you can’t accurately track that freshwater pulse, your climate projections carry uncertainty. Now we can.
And there’s a bonus: the same system could improve polar navigation safety in increasingly trafficked Arctic and Antarctic waters.
Why This Matters:
Better tracking of iceberg fragmentation sharpens climate models at a moment when ice loss is accelerating — reducing uncertainty in one of the most sensitive parts of the Earth system.
Kismet:
For the first time, scientists can trace an iceberg fragment drifting past a ship back to the exact glacier it calved from years earlier — effectively giving melting ice a passport and a family history.
Sovereign AI Is the New Firewall - And It’s About Trust
For years, US hyperscalers sold the world a simple deal: unmatched scale, unmatched innovation, unmatched convenience. Now the counter-question is getting louder: unmatched control over whose data?
SAP’s expanded partnership with Cohere to launch full-stack sovereign AI solutions, starting in Canada is not just a product move. It’s a signal. AI models embedded inside a nationally operated sovereign cloud. Data stays in-country. Operational control remains local. No foreign jurisdictional reach.
Key highlights
AI embedded directly inside SAP’s Canadian-operated sovereign cloud, keeping sensitive data in-country
Designed for public sector, regulated industries, and critical infrastructure customers
Generative AI integrated into ERP systems rather than exporting enterprise data to external AI endpoints
Why This Matters: AI adoption is entering a trust phase. For governments and critical industries, jurisdiction and data residency are becoming decisive competitive factors.
Kismet: In the cloud era, companies asked “How fast is it?” In the AI era, more are asking “Under whose laws does it run?” 👉 Full story here
Electromobility
Electrifying Cars Cuts Emissions. Full Stop.
A new peer-reviewed study in Communications Sustainability delivers something refreshingly rare in the EV debate: system-wide clarity. Electrifying light vehicles in the United States reduces emissions across all vehicle types and powertrains, even accounting for the current grid mix. The gains are immediate, and they grow as the grid gets cleaner.
This isn’t modelled optimism. It’s lifecycle analysis at scale.
Key highlights
Electrifying light-duty vehicles lowers total greenhouse gas emissions across sedans, SUVs, and pickup trucks
Even on today’s grid, EVs outperform internal combustion vehicles on lifecycle emissions
As the power sector decarbonises, the emissions advantage of EVs widens significantly
Why This Matters: Transport electrification delivers emissions reductions now, and locks in deeper cuts automatically as electricity generation gets cleaner.
Kismet: Because EV emissions improve every time the grid adds renewables, buying an EV today is effectively pre-ordering lower emissions for the next decade without changing the car. 👉 Full story here
Your EV Is Now a Backup Power Plant
Electric vehicles are quietly evolving from transport devices into mobile batteries. Around 630,000 EVs in the U.S. already have vehicle-to-home capability, and adoption is accelerating fast. During recent winter storms, owners kept lights, heaters and even ovens running while neighbours scrambled for diesel and petrol.
I can vouch for this personally. When the Iberian peninsula blackout hit last April, my Kia EV3 powered critical devices in our home. Fridge. Internet. The coffee machine (!). Calm restored.
Key highlights
~630,000 U.S. EVs can now power homes via bidirectional charging, with one in five new EVs recently sold offering the feature
Major automakers including GM, Hyundai, Kia and others are rolling V2H out across more models
Utilities are piloting vehicle-to-grid programmes, seeing EV fleets as distributed energy assets
Why This Matters: EVs aren’t just decarbonising transport. They’re becoming resilience infrastructure, turning millions of parked cars into decentralised backup power systems.
Kismet: An average EV battery holds enough energy to power a typical home for several days, making your driveway the largest battery most households will ever own. 👉 Full story here
EV Batteries Just Took Three Big Leaps at Once
Battery innovation isn’t slowing. It’s branching. In the space of a week, we’ve seen the first mass-produced sodium-ion EV, credible progress toward solid-state deployment, and a 12-minute fast-charging battery claiming 1.5 million miles of life. This isn’t incremental improvement. It’s chemistry diversification plus performance acceleration.
Key highlights
Sodium-ion goes mainstream: First passenger EV with sodium batteries delivers ~400 km range, strong cold-weather performance, and reduces reliance on lithium supply chains
Solid-state edging closer: BYD targets 2027 pilot production, reporting improvements in cycle life and fast charging using sulfide electrolytes
Ultra-fast, ultra-durable lithium: CATL’s 5C battery promises ~12-minute charging and up to 1.5 million miles of lifespan while retaining 80% capacity
Why This Matters: Cheaper materials, longer lifetimes, faster charging, better cold performance - the combined effect is EVs that are more resilient, more affordable, and increasingly hard to argue against on technical grounds.
Kismet: The average vehicle lasts ~350,000 km. A 2+ million km battery could outlive 7–8 car bodies. Instead of scrapping the battery when the chassis wears out, you swap the body and redeploy the pack - slashing lifecycle emissions and turning the battery into a long-term capital asset rather than a consumable component.
👉 Links in the story this time!
Clean Energy
Australia’s Renewables Boom Cuts Power Prices
For years, critics argued that rapid renewable buildouts would mean permanently higher electricity prices. Australia just complicated that narrative. In 2025, wholesale power prices fell to their lowest level in four years, even as clean electricity output more than doubled since 2019 and overtook fossil fuels for the first time. This is what happens when scale, storage, and system redesign start to align.
Key highlights
National wholesale electricity prices averaged just under $92/MWh in 2025, the lowest annual level since 2021
Clean electricity output has more than doubled since 2019, while fossil fuel generation has fallen 21%
Wholesale prices across major states fell an average of 11% year-on-year in 2025, with Queensland down 15%
Why This Matters: Australia shows that aggressive renewable deployment, combined with battery storage, can stabilise and even lower wholesale power prices once the buildout phase matures.
Kismet: Australia’s clean electricity growth since 2019 has outpaced Europe, North America, and the global average combined, suggesting the “too fast, too expensive” argument may age very badly. 👉 Full story here
Battery Tech
EV Batteries Are Becoming China’s Second Energy System
The first big wave of EV batteries in China is starting to retire. But instead of heading straight for recycling, many could get a second life on the grid. One recent analysis suggests used EV batteries could meet around two-thirds of China’s grid storage needs, charging when renewables are abundant and discharging during peak demand.
At the same time, Beijing is rapidly formalising battery recycling to prepare for what officials describe as a coming “tsunami” of spent packs.
Key highlights
Used EV batteries, typically retired at ~80% capacity, could supply up to two-thirds of China’s grid storage demand
China generated nearly 400,000 tonnes of retired batteries last year, projected to reach ~1 million tonnes annually by 2030
China already dominates global battery recycling capacity, holding roughly 78% of pre-treatment and 89% of refining capacity
New policies mandate end-of-life recycling and digital lifecycle tracking to prevent grey-market diversion
Why This Matters: Second-life deployment combined with aggressive recycling turns EV batteries from a future waste headache into a strategic energy and mineral security asset.
Kismet: If a typical battery serves ten to twelve years in a car and then another decade on the grid before recycling, a single pack could underpin nearly 20-30 years of clean energy infrastructure, without a single new tonne of mining. 👉 Full story here
Latest blog post
Capital Follows Electrons
The IEA’s Electricity 2026 report was released this week and it makes one thing unmistakable: electricity is no longer a side story in the energy transition. It’s the main plot. Global demand is growing at 3–4% annually - well above historic norms, driven by EVs, heat pumps, industry, data centres, and AI.
And almost all net new generation? Renewables.
In my latest blog post, I unpack what this means for competitiveness, industrial strategy, and why some countries are quietly pulling ahead while others stall.
Key highlights
Global electricity demand is accelerating structurally, not cyclically
Renewables account for the vast majority of new power capacity additions
Coal generation has likely peaked globally; power-sector emissions are flattening
Grids, not generation, are becoming the primary bottleneck
Countries aligning industrial policy with electrification are gaining strategic advantage
Why This Matters: Access to clean, reliable electricity is becoming the decisive factor in industrial competitiveness, and capital is already flowing accordingly.
Kismet: For most of the 20th century, economic power followed access to fuels. In the 21st, it’s following access to electrons, and that shift may reshape trade, industry, and geopolitics faster than climate targets ever could. 👉 Full story here
Climate Confident:
Designing Buildings for 2125, Not 1975
We’re still designing buildings using historical climate data. The problem? Those buildings will spend most of their lives in a future climate we’ve never experienced. In this week’s Climate Confident episode, I spoke with David Sellers of Hawaii Off Grid Architecture about why designing for yesterday’s weather is quietly locking in tomorrow’s risk.
From shifting trade winds to rising sea levels and more extreme heat days, the climate assumptions architects rely on are already outdated.
Key highlights
Maui’s prevailing trade winds have shifted ~9 degrees since 1950, reshaping rainfall patterns and water availability
30–40 more days per year now exceed 90°F (≈ 32°C), undermining traditional passive cooling assumptions
Lithium battery costs have fallen ~80% in a decade, making fully off-grid, climate-controlled homes economically viable
Buildings account for ~30–40% of global emissions, meaning design decisions today shape decades of carbon output
Why This Matters: If buildings last 60–100 years, designing for historical averages is a systemic risk, to comfort, resilience, and climate stability.
Kismet: A building completed in 2026 will never experience 1970’s climate. It will experience 2050, 2080, and beyond, potentially in conditions more than 7°F (roughly 4°C) warmer than today. Designing for the past may be one of the most expensive blind spots in modern architecture. 🎧 Listen to the full episode
Resilient Supply Chain:
98% of Construction Emissions Sit in the Supply Chain
When only 2% of your emissions are under direct control, sustainability stops being a side project. It becomes a resilience strategy. In this week’s Resilient Supply Chain, I spoke with Keith O’Flynn, Group Supply Chain Sustainability Manager at John Sisk & Son, about what it really takes to decarbonise construction when 98% of emissions sit upstream.
Concrete. Steel. Data. Standards. Margins. It’s all connected.
Key highlights
~98% of Sisk’s emissions are Scope 3, with ~80% of those tied to concrete and steel
Low-carbon concrete trials are delivering up to 70% embodied carbon reductions versus traditional baselines
Construction sites often consume most of their energy after hours, revealing major efficiency gains hiding in plain sight
Data quality, Environmental Product Declaration document availability, and spend-based accounting remain major structural barriers
Why This Matters: If resilience is about understanding where risk sits, then in construction it sits in materials, standards, and supplier capability, not just on site.
Kismet: In a sector operating on 1–2% margins, the biggest risk isn’t decarbonising too fast. It may be failing to modernise supply chains before clients, regulators, and insurers demand it. 🎧 Listen to the full episode
Coming Soon to the podcasts
In the coming episodes I will be talking to Lily Hogan, Senior Product Manager, 3E and Beatrice Clark, VP Sustainability at Turtle.
Don’t forget to follow the podcasts in your podcast app of choice to ensure you don’t miss any episodes.
Featured Chart(s)
One of these things is not like the others…
Biofuels are hugely inefficient when compared to solar panels, especially wrt land use.
Spain has had the coldest, darkest, wettest January since 1996. climate change means an increase in weather extremes.
Misc stuff
Our new easier, more efficient AI age!
Anyone who has owned an Apple mouse knows this pain!
Surely the dinosaurs weren’t THAT stupid?
This one made me chuckle!
Engage
If you made it this far, very well done! If you liked this newsletter, or learned something new, feel free to share this newsletter with family and friends. Encourage folks to sign up for it.
Finally, since being impacted by the tech layoffs, I'm currently in the market for a new role. If you know someone who could benefit from my tech savvy, sustainability, and strong social media expertise, I'd be really grateful for a referral.
If you have any comments or suggestions for how I can improve this newsletter, don’t hesitate to let me know. Thanks.
*** Be aware that any typos you find in this newsletter are tests to see who is paying attention! ***
And Finally
Speak fuller prompts. Get better answers.
Stop losing nuance when you type prompts. Wispr Flow captures your spoken reasoning, removes filler, and formats it into a clear prompt that keeps examples, constraints, and tone intact. Drop that prompt into your AI tool and get fewer follow-up prompts and cleaner results. Works across your apps on Mac, Windows, and iPhone. Try Wispr Flow for AI to upgrade your inputs and save time.