Sponsored by
Hey everyone, it’s Thursday, and FutureProof is back.
This week, fossil-fuel volatility keeps making electrification look smarter, AI gets ambitious, and clean energy keeps scaling in places it supposedly shouldn’t.
Highlights this week:
The Stakes: Record heat, a strengthening El Niño, and even Champagne feeling the climate squeeze.
Climate: Cleaner air and climate action could return $15 for every $1 invested.
AI: DeepMind maps 9 billion DNA mutations, AI tackles contrails, and GPT-6 arrives.
EVs: Russia’s fuel shortages double EV sales, while electric trucks push further into diesel territory.
Clean energy & science: UK solar breaks records, floating wind moves offshore, and fusion attracts $14bn.
Research & podcasts: New executive research on long-duration storage, plus Ray Wills on energy-transition risk and Friddy Hoegener on AI hollowing out the leadership pipeline.
A lot of transition stories this week share the same theme: the economics are moving faster than many assumptions.
Let’s get into it.
The Stakes
The Planet Is Heating Up - and Now It’s Coming for the Champagne
August tied for the hottest month ever recorded at 1.65°C above pre-industrial levels, Spain endured a record 61 heatwave days, and a rapidly strengthening El Niño threatens to pile droughts, floods and food disruption on top of an already overheated climate. And, because apparently the climate crisis has decided subtlety is overrated, France now expects Champagne production to plunge 48% after brutal heat and drought hammered the vineyards.
Spain spent roughly two out of every three summer days under heatwave conditions, with 5,234 heat-related deaths estimated between mid-May and the end of August.
The developing El Niño could become the strongest on record, with the World Food Programme estimating its effects could push roughly 49 million additional people into acute hunger.
France is heading for one of its smallest wine harvests in decades: Champagne suffered its earliest harvest on record, grape bunch weights hit a 20-year low, and production could almost halve.
Why This Matters: Climate change is no longer an abstract curve on a graph; it is increasingly showing up simultaneously in mortality statistics, food prices, agricultural yields, supply chains and economic losses, precisely why accelerating clean energy and electrification matters.
Kismet: Champagne may sound like the frivolous casualty here, but France estimates this summer’s climate-related economic damage at €10-15 billion, up to roughly 0.5% of GDP, the missing bubbly is merely the fizzy bit of a much bigger bill.
👉 Links Inline
Climate
Climate Action Pays Back $15 for Every $1 Invested
Climate action as an economic drag? That argument just took a fairly serious beating. A new UN assessment finds that tackling air pollution and climate change together could lift global GDP by 2.8% by 2035, and I find the return particularly striking: every $1 invested generates about $15 in combined economic and societal benefits.
The report identifies 25 existing measures - including renewables, EVs, cleaner cooking and cutting oil-and-gas venting, costing about 0.7% of global GDP over the coming decade.
Even counting only measurable market benefits such as lower healthcare costs and higher productivity, the return is roughly $4 for every $1 invested.
Delay is expensive: each year of inaction forfeits more than $1.5 trillion in potential benefits.
Why This Matters: Decarbonisation increasingly looks less like an environmental expense and more like an extraordinarily attractive investment in productivity, public health and economic resilience.
Kismet: China has cut harmful air pollution by almost 60% since 2013, with the share of hospital admissions attributed to that pollution falling by roughly 30%, an unusually tangible dividend from cleaner air. 👉 Full story here
Clean Energy Has a Tipping Point, and We May Already Be Crossing It
Here’s a climate tipping point I’m very happy to see: researchers say clean electricity is now the highest-leverage environmental transition, because cheaper wind and solar make EVs, heat pumps and cleaner industry more attractive, which increases electricity demand, drives further investment and pushes costs down again. In other words, decarbonisation can start behaving less like a policy programme and more like a flywheel.
Solar and wind have already crossed positive tipping points in many major markets, where deployment lowers costs and lower costs trigger yet more deployment.
Coordinated policies across electricity, transport and heating could bring forward tipping points by two to eight years.
The biggest brake may now be infrastructure and capital: Africa receives less than 2% of global clean-energy investment despite having 60% of the planet’s best solar resources.
Why This Matters: Once clean technologies become cheaper, easier and more politically attractive than the incumbent alternatives, the transition can accelerate under its own economic momentum - which changes the entire conversation about how quickly decarbonisation can happen.
Kismet: For every £1 flowing into nature protection, roughly £30 finances activities that degrade nature, suggesting the shortage isn’t capital so much as our rather spectacular choice of where to put it. 👉 Full story here
AI News
Google’s AI Just Pre-Calculated 9 Billion Possible DNA Mutations
Google DeepMind has released AlphaGenome Atlas, a one-petabyte database predicting the biological impact of every possible single-letter change across the human genome, all 9 billion of them. What grabs me is the shift from AI answering questions on demand to AI doing an enormous body of scientific computation upfront, so researchers can query the results almost instantly.
The Atlas covers both the 2% of DNA that codes for proteins and the far less understood 98% that doesn’t.
Researchers at the Broad Institute used its new Variant Impact score to help identify a mutation implicated in an unsolved rare-disease case.
Analysis of more than 54,000 UK Biobank participants uncovered 22% more associations involving non-coding genetic variants.
Why This Matters: AI is starting to turn previously impractical scientific searches into searchable infrastructure, potentially accelerating everything from rare-disease diagnosis to fundamental biology.
Kismet: The entire dataset weighs about one petabyte - roughly a million gigabytes, yet DeepMind has built a browser-based interface requiring no coding skills, putting a decidedly enormous genomic calculation behind a relatively ordinary-looking search box. Truly Amazing! 👉 Full story here
AI Is Coming for Contrails. Conspiracy Theorists, Please Update Your Files
Google and Cathay Pacific are using AI, satellite imagery and weather data to steer aircraft away from the atmospheric pockets where persistent contrails form, and early trials cut their warming impact by 40%. Given contrails may account for roughly a third of aviation’s total climate impact, that is an astonishingly large climate gain from what can amount to a modest altitude adjustment.
Google’s system predicts where contrails are likely to form and gives flight planners suggested altitude changes to avoid those zones.
The approach works with existing aircraft and existing fuels, making it potentially far faster to deploy than replacing fleets or waiting for sustainable aviation fuel to scale.
Cathay is now expanding testing across intra-Asia, transpacific and polar routes to see how well the approach works on ultra-long-haul flights.
Why This Matters: Aviation has relatively few easy decarbonisation levers, so knocking a substantial chunk off its non-CO₂ warming effect using software and small operational changes could be one of the cheapest climate wins available to the sector.
Kismet: Should the skies suddenly contain fewer white streaks, I await with interest the inevitable theory explaining who has taken all the contrails, where they’re storing them, and what Google plans to do with the enormous secret contrail stockpile. 👉 Full story here
AI Had a Busy Week: GPT-6 Arrives, Meta Gets Agentic, and Apple Starts Proving Reality
This was one of those weeks when AI announcements stopped looking like incremental model updates and started looking more like pieces of a new computing stack snapping into place. OpenAI launched GPT-6 Astra and ChatGPT Images 2.5, Meta unveiled an agent called Muse that can actually do things for you, while Apple tackled the increasingly awkward question of whether the photograph you’re looking at depicts something that ever existed.
GPT-6 Astra pushes hard into autonomous professional work: computer use, software engineering, scientific research and multi-step workflows, with OpenAI reporting 99.9% on ARC-AGI-3, 97.6% on FrontierMath Tier 4 and substantially faster computer-use performance than GPT-5.6 Sol.
Meta’s Muse gets its own persistent virtual computer and browser, can book appointments, fill forms, deal with customer service and work across connected apps, while requiring approval before consequential actions such as purchases or sending emails.
ChatGPT Images 2.5 promises up to 50% faster generation, better preservation of people and objects during repeated edits, and a new Sketch feature; meanwhile Apple’s Reference Image captures cryptographically signed sensor data on the iPhone 18 Pro, creating what amounts to a digital negative for checking whether a photograph has subsequently been altered.
Why This Matters: The AI race is rapidly shifting from “which chatbot answers best?” to something much bigger - models that see, create, operate computers, use software and take actions, accompanied by an emerging authenticity layer for distinguishing generated media from captured reality.
Kismet: There’s a lovely symmetry here: in the same week AI became dramatically better at manufacturing convincing images, Apple announced technology for proving that an image wasn’t manufactured by AI. The arms race has apparently developed its own immune system. 👉 Links Inline
Electromobility
EV Sales Keep Surging - From Britain to Australia, Portugal and Even Russia
If anyone still thinks EV adoption is a niche phenomenon waiting politely for perfect conditions, August’s numbers are becoming rather inconvenient: battery-electric sales jumped 30% in the UK, 65% in Portugal and roughly 170% in Australia, while Russia’s electric and plug-in sales more than doubled over the summer. China tells an equally revealing story, pure BEVs edged higher even as the overall car market plunged, while sales of fossil-powered vehicles collapsed 40% year-on-year.
In Australia, BEVs reached 25.6% of August sales, becoming the country’s largest individual drivetrain category; add hybrids and plug-ins and electrified vehicles passed 51%.
In the UK, BEVs captured almost 30% of August registrations, overtaking every other individual fuel type, while Portugal’s pure-EV share reached 36.1%.
China’s BEV sales rose 0.8% despite total retail car sales falling 23.6%, while Chinese new-energy-vehicle exports exploded 154.7% as manufacturers increasingly push their electric cars into global markets.
Why This Matters: EV growth is no longer being driven by one country, one policy or one manufacturer - electrification is appearing across very different markets because economics, model availability, energy security and industrial competition are increasingly pointing in the same direction.
Kismet: Perhaps the strangest datapoint comes from oil-rich Russia, where refinery attacks pushed petrol production down to roughly 70% of domestic demand, leaving motorists queueing for fuel, and suddenly making plugging in an electric car look considerably less radical. 👉 Links Inline
Electric Trucks Have Stopped Asking Permission to Replace Diesel
The old objection to electric HGVs was range; that argument is ageing badly. Trucks capable of 400–500 km are already covering the boring, predictable routes where freight actually makes money, while a new generation is pushing towards 600–800 km, and I think the more important story is that, for suitable high-utilisation routes, the operating economics can already favour electricity long before maximum range reaches diesel territory.
Mercedes, MAN, DAF, Volvo and others already offer electric heavy trucks around the 400-500 km class, while newer European and Chinese platforms are pushing towards 600 km and beyond.
The article’s operating-cost model suggests an electric semi running roughly 700 km a day could save around $200–$300 per day on energy alone, before maintenance and downtime are counted.
China is turning freight electrification into industrial strategy: electric heavy-truck sales reportedly surged 182% in 2025, while enormous charging hubs are emerging to support hundreds of trucks per day.
Why This Matters: Freight electrification doesn’t need every truck to drive 1,000 km non-stop, once electric HGVs can economically cover the repeatable routes that dominate logistics, diesel starts losing kilometres long before it loses every use case.
Kismet: One Chinese electric-truck charging hub is being built with 100 MW of capacity, including 18 chargers rated at 1.44 MW each (for context that is over 10 times faster charge than my EV3 can accept!), at which point “EV charger” starts sounding rather inadequate; that’s basically a small power station with parking spaces.
Clean Energy
Britain Just Had a Record Solar Summer. Yes, Britain.
The UK’s famously dependable sunshine generated a record 8.8TWh of solar electricity this summer, up 23% on 2025 and double the level produced just five summers ago. July alone delivered 3.3TWh, enough to meet 15% of Britain’s electricity demand, another reminder that solar stopped being a Mediterranean speciality some time ago.
Solar generation set new monthly records in June, July and August, helped by both exceptionally sunny weather and rapidly expanding capacity.
Installed UK solar capacity is now approaching 24GW, with nearly 172,000 new installations added since the start of 2026 and roughly 1.7 million homes now sporting panels.
During 26 heatwave days, households with rooftop solar are estimated to have avoided £86.7 million in electricity costs, while extra daytime generation also reduced the need for expensive gas-fired power.
Why This Matters: Solar is becoming a meaningful component of electricity systems far beyond traditionally sunny markets, and every additional rooftop or solar farm compounds the effect by cutting gas consumption, wholesale electricity costs and emissions simultaneously.
Kismet: In July, Britons were installing rooftop solar at a rate of roughly one system every two minutes, which may be the strongest evidence yet that even the British have finally decided not to waste the occasional appearance of the sun. 👉 Full story here
Floating Wind Is Heading for Deep Water - Literally
Offshore wind has spent years bolted to the seabed, but the next phase is increasingly likely to float. With much of the planet’s best offshore wind resource sitting in water too deep for fixed foundations, developers are borrowing technology from offshore oil and gas to anchor turbines on floating platforms - expensive today, potentially transformative once scale and learning curves kick in.
Floating wind still accounts for less than 1% of the world’s 91GW of offshore wind capacity, but it represents 22% of the future project pipeline, with nearly 200GW announced.
The big prize is access to deep-water locations near major coastal demand centres where fixed-bottom turbines simply cannot go.
The economics remain the headache: even in China, currently the cheapest market, floating projects can cost two to four times more than conventional offshore wind.
Why This Matters: If costs fall, floating wind could unlock vast new areas for renewable generation and give land-constrained coastal economies another large-scale source of clean power.
Kismet: There’s a pleasing irony in using engineering perfected for offshore oil and gas rigs to build the infrastructure that could help displace fossil fuels. 👉 Full story here
Science
Fusion Has Attracted $14 Billion. Now It Has to Become a Power Plant
Fusion has spent decades being the energy technology of tomorrow, but investors have now put more than $14 billion into trying to make tomorrow arrive rather sooner. The physics milestone has been crossed; what interests me now is the much harder transition from spectacular laboratory experiment to repeatable, affordable electricity.
Fusion companies raised a record $4.5 billion in the past year, with more than 50 firms worldwide now pursuing commercial power.
Pacific Fusion is building a $1 billion demonstration facility in New Mexico for 2030, while Commonwealth Fusion Systems expects to switch on its demonstration machine next year.
The economic target is getting serious: several developers ultimately hope to produce electricity for roughly $50–100/MWh, potentially competitive with other major generation technologies.
Why This Matters: Fusion doesn’t need to replace cheap solar and wind to matter enormously; commercially viable, dispatchable zero-carbon power could become a valuable complement in an electricity system facing rapidly rising demand.
Kismet: The old joke says fusion is always 20 years away, but Microsoft and Google have already signed agreements to buy electricity from fusion plants that don’t yet exist, corporate procurement has apparently decided to start arguing with the joke. 👉 Full story here
Latest blog post
I’m Turning 800+ Podcast Conversations Into Research, and This One Tackles Energy Storage
My third executive brief asks a deceptively simple question: if long-duration energy storage is increasingly technically viable, why isn’t much more of it getting built? Drawing on 50 relevant podcast interviews, the answer points less to another breakthrough chemistry and more to market design - electricity systems still struggle to value, procure and finance the ability to move energy across time.
The Missing Market for Time is the third executive brief in my growing Research library, with all reports free to download.
Its central argument is that LDES needs clearer system requirements, duration-sensitive revenues and contracts capable of turning technical capability into bankable projects.
I’m now exploring sponsorship for future briefs, including research built around the thorny strategic questions companies and organisations genuinely need answered, using evidence from hundreds of expert conversations rather than producing another corporate white paper destined for PDF purgatory.
Why This Matters: There’s an enormous amount of hard-won expertise sitting inside long-form conversations, and turning that archive into structured, independent research creates a useful bridge between thought leadership and executive decision-making.
Kismet: The storage numbers illustrate the problem beautifully: the world added 108GW of battery storage in 2025, eleven times the installed capacity of 2021, yet most projects still cluster around roughly two hours, proving that “storage is booming” and “we’ve solved long-duration storage” are very different statements. 👉 Full story here
Climate Confident:
The Biggest Energy Risk May Now Be Betting on Yesterday
On this week’s Climate Confident, Prof Ray Wills argues that the clean-energy transition is moving so quickly that the risk equation has flipped: moving too slowly can now be considerably more expensive than moving too fast. Batteries, solar and EVs are following steep adoption and cost curves, while companies investing in long-lived fossil infrastructure face a rather awkward question, who will still want to buy what they’re selling by the time those assets need to earn their keep?
In Western Australia, batteries can now supply up to 30% of electricity during some peak periods, despite contributing only around 4% across the day, that peak contribution was effectively zero in 2024.
Ray argues that China’s advantage comes from treating clean tech as an industrial ecosystem: build solar, batteries and EVs domestically at enormous scale, drive costs down, improve quality, then export the resulting technologies.
For business leaders, his message is particularly sharp: EVs, batteries and renewable energy increasingly represent competitiveness and energy-security decisions, rather than merely climate initiatives.
Why This Matters: Capital allocation depends on assumptions about the future, and if clean technologies continue scaling along nonlinear curves, investing as though today’s energy system will persist could become the more dangerous strategic bet.
Kismet: Ray has been publishing clean-tech adoption curves for roughly 15 years, and his advice for spotting a dubious forecast is wonderfully simple: if the clean-tech line doesn’t bend upwards, be suspicious of the model rather than the transition. 🎧 Listen to the full episode
Resilient Supply Chain:
AI Can Cut Headcount Today, and Create a Leadership Crisis Tomorrow
On this week’s Resilient Supply Chain, Friddy Hoegener makes a point more companies should be worrying about: if AI removes the entry-level work where people learn the business, where do your future supply chain leaders come from? The smarter play may be to use AI to strip out tactical grunt work while pushing junior employees into analysis, judgement and strategic problem-solving much earlier.
Friddy sees companies splitting into two camps: those using AI primarily to remove roles, and those using the capacity it creates to make existing jobs more strategic.
Future-proof supply chain skills increasingly look like data analysis, stakeholder management, cross-functional communication, adaptability and the ability to make a business case, rather than years of narrow category expertise.
Companies also damage themselves through poor hiring discipline: vague role definitions, generic job descriptions and endless interview rounds can lose excellent candidates before a decision is made.
Why This Matters: AI-driven productivity gains are attractive, but companies that hollow out their junior talent pipeline may discover several years from now that they optimised away the people they needed to become tomorrow’s managers and directors.
Kismet: One of the more counterintuitive effects of AI may be that entry-level jobs become harder rather than easier, less data entry and PO processing, more judgement, interpretation and strategic thinking from day one.
🎧 Listen to the full episode
Don’t forget to follow the podcasts in your podcast app of choice to ensure you don’t miss any episodes.
Featured Chart(s)
On paper plug-in hybrids sound great - the reality however is somewhat different.
EVs are really taking over the car market in Europe, and with oil once more going over $100 per barrel, that trend is not going to reverse.
This is why China is referred to as an electrostate.
Misc stuff
Harsh, but true!
This is a Spanish cartoon, so the IA is Inteligencia Artificial - Artificial Intelligence in Spanish!
I bought my car and my iPhone in Dec 2024. My iPhone’s battery is now at 93%, whereas my car’s is still at 100%. EV batteries are completely coddled and last for 100’s of thousands of kilometres.
Obligatory Trump Cartoon
Not a cartoon per se, but on point!
And I love this photo of the German far-right leader with an appropriately positioned microphone shadow. Not Trump, but Trumpian!
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
Did an ACME device finally work?
How 2M+ Professionals Stay Ahead on AI
AI is moving fast and most people are falling behind.
The Rundown AI keeps you ahead of the curve.
It's a free AI newsletter that keeps you up-to-date on the latest AI news, and teaches you how to apply it in just 5 minutes a day.
Plus, complete the quiz after signing up and they’ll recommend the best AI tools, guides, and courses — tailored to your needs.