In partnership with

Hey everyone, it’s Thursday, petrol is losing the pricing argument, and FutureProof is back.

This week is about thresholds.

Electric cars are now cheaper than petrol cars in the UK. Renewables have overtaken coal globally. And in AI, the week managed to include design tools, drug discovery, cyber defence, and a leaked model, because apparently nobody in tech can just have a normal product launch anymore.

The big threads:

  • The economics are turning. EVs, renewables, and storage keep getting cheaper, while fossil fuels keep arriving with price shocks, volatility, and geopolitical drama attached.

  • China and India are scaling clean energy fast enough to stop looking like “future potential” and start looking like the centre of gravity.

  • AI is moving deeper into actual work, from coding and design to science and security, while the risks are becoming harder to keep in the lab.

Plus: my latest blog post on why oil shocks make electrification a business imperative, a new Climate Confident episode on why the real climate problem in buildings is now materials and resilience, and a new Resilient Supply Chain episode on why most logistics teams are still making decisions half-blind.

Oh, and look out for the latest Resilient Supply Chain bonus episode which drops tomorrow for all Resilient Supply Chain+ subscribers.

Let’s get into it.

Climate

Solar Slows the Damage, Even as America Burns More Coal

This is the sort of story that deserves more attention than it’ll get. Global energy-related CO2 emissions still rose in 2025, but only by 0.4%, as a huge build-out of solar, especially in developing economies, helped put a serious brake on what could have been a much uglier year.

Key Highlights

  • The IEA says global emissions growth slowed sharply in 2025, with solar doing much of the heavy lifting as it reshaped the power mix.

  • China’s emissions fell as it led solar capacity additions, while India’s emissions dropped for the first time on record in normal economic conditions, helped by a strong monsoon and more renewable generation.

  • Advanced economies moved the wrong way, though, with the US driving the first annual rise in emissions from that group since 2018, thanks to higher coal use, strong electricity demand from data centres, industrial growth, and colder weather.

Why This Matters: This matters because it shows the clean energy transition is no longer some future promise. It is already bending the emissions curve, even while parts of the old system keep fighting like a pub regular being asked to leave at closing time.

Kismet: India’s emissions had previously fallen only during the pandemic and the oil shocks of the 1970s, so the fact they dropped in an otherwise normal economic year is a really startling sign that renewables are starting to alter the structure of growth itself. 👉 Full story here

Colombia Tries What COP Won’t: Build a Fossil Fuel Exit Club

I’m genuinely encouraged by this one. Colombia, hardly a country with no skin in the fossil game, is convening a “coalition of the willing” to push the transition away from oil, gas, and coal because the usual COP process has become a masterclass in delay, dilution, and diplomatic theatre. And it was good to see Tzeporah Berman of the Fossil Fuel Non-Proliferation Treaty quoted in the piece, especially since I had her on the Climate Confident podcast last year to talk about why climate policy has to tackle fossil fuel supply, not just demand.

Key Highlights

  • Colombia and the Netherlands are co-hosting a new Santa Marta conference, backed by 54 countries, to move fossil fuel transition talks forward outside the usual deadlocked UN process.

  • The timing is painfully apt: war-linked oil shocks are pushing up prices and making renewables, EVs, heat pumps, and energy independence look less like climate virtue and more like economic self-defence.

  • The conference aims to focus on tangible outcomes, especially finance for developing countries, technology transfer, and ensuring Indigenous and marginalised communities aren’t sacrificed in the name of a messy transition.

Why This Matters: This matters because it suggests climate progress may now depend less on waiting for universal agreement and more on willing countries moving ahead without asking petrostate blockers for permission.

Kismet: The next conference is already planned for Tuvalu, which is a striking twist, one of the countries least responsible for the climate crisis helping shape the diplomatic architecture that will finally weaken fossil fuels’ grip on global politics. 👉 Full story here

Hungary Just Fired Europe’s Biggest Climate Blocker

This is one of those political stories that looks national on the surface but lands squarely in Brussels. Orbán’s defeat does not mean Hungary suddenly turns into a climate hero, but it does mean one of the EU’s most stubborn blockers of climate action and Russian energy phaseout is finally out of the way.

Key Highlights

  • Orbán’s government had repeatedly slowed or obstructed EU climate and energy policy, including action on net-zero, Fit for 55, the 2035 combustion-engine phaseout, and efforts to cut Russian fossil fuel dependence.

  • Péter Magyar’s Tisza party is expected to be far more cooperative with the EU, and has quickly moved to unlock billions in frozen EU funds, much of it earmarked for green measures such as energy efficiency, home insulation, rail electrification, and drought resilience.

  • This is not some radical green uprising, though. Tisza looks more like a pragmatic centre-right government that may remove stupid restrictions on wind, back geothermal and renewables, and be less disruptive, without becoming a climate vanguard.

Why This Matters: This matters because in EU climate politics, removing a serial blocker can be almost as important as electing a champion, especially when energy security, industrial policy, and Russian fossil fuel dependence are all colliding at once.

Kismet: Hungary already gets roughly three-quarters of its electricity from clean sources, and nearly a third from solar alone, so the strange twist here is that one of Europe’s most politically obstructive governments was sitting atop a power mix that was cleaner than many of its louder climate peers. 👉 Full story here

AI News

OpenAI Goes Full Stack: Code, Images, and Drug Discovery

I think this was the week OpenAI stopped looking like a chatbot company and started looking more like a layer of operating infrastructure. In the space of a few days it pushed Codex towards full agentic software work, launched ChatGPT Images 2.0 with far better control over text and layout, and unveiled GPT-Rosalind for life sciences, while the image I’m using for this story was itself made with Images 2.0 after I fed it a couple of photos of me and asked for a 1950s-style newsroom scene, because apparently that’s a normal sentence now!

Key Highlights

  • Codex is no longer just helping write code. OpenAI says it can now operate a Mac, use apps and the web, remember preferences, schedule future work, connect to remote devboxes over SSH, and work across much more of the software development lifecycle for the more than 3 million developers already using it each week.

  • ChatGPT Images 2.0 looks like a serious leap from “fun AI picture generator” to practical creative tool, with OpenAI highlighting greater precision and control, stronger multilingual text rendering, and better fidelity across complex layouts and visual styles.

  • GPT-Rosalind is OpenAI’s new life sciences model for biology, drug discovery, and translational medicine, built to help with evidence synthesis, hypothesis generation, experimental planning, and tool-heavy research workflows, with access gated through a trusted programme for qualified customers.

Why This Matters: This matters because OpenAI is no longer just competing to build a smarter assistant. It is moving, fast, into three much bigger markets at once: software production, creative production, and scientific discovery. That is a far more consequential shift than another week of model benchmark chest-beating.

Kismet: The strangest and most telling detail in all this is the spread: one OpenAI system is now aimed at helping scientists shorten research paths that can take 10 to 15 years in drug development, while another is busy turning people like me into faux mid-century newsroom editors. Same company, same week, wildly different parts of the economy suddenly up for grabs. 👉 Links Inline

Anthropic Wants AI to Design Your Next Deck, Prototype, and Pitch

Anthropic has launched Claude Design, a new product that lets users create prototypes, slides, one-pagers, mockups, and marketing assets by collaborating with Claude, which is a pretty clear sign that AI design tools are moving from gimmicky image generation into actual workplace output. In other words, the AI race is no longer just about who writes best, it’s about who gets embedded deepest into how teams build, present, and ship ideas.

Key Highlights

  • Claude Design is powered by Claude Opus 4.7 and is being rolled out in research preview for Pro, Max, Team, and Enterprise subscribers.

  • Anthropic says users can start from prompts, uploaded files, websites, or codebases, then refine designs through conversation, inline comments, direct edits, and Claude-generated sliders.

  • The product is built for practical workflow use, with support for team design systems, collaboration, exports to PDF, PPTX, Canva, and HTML, plus handoff bundles for Claude Code.

Why This Matters: This matters because design is becoming another frontline in the AI platform war, and whichever model gets trusted with presentations, prototypes, and branded assets could end up owning far more of the modern knowledge workflow than people currently realise.

Kismet: One of the most revealing details is that Anthropic built export paths to both PPTX and Canva, which tells you this is not aimed at replacing designers outright, but at inserting AI directly into the gloriously messy corporate machinery where ideas become slides, and slides unfortunately become decisions. 👉 Full story here

Anthropic’s Mythos Is Already Fixing Bugs and Escaping Containment

We talked about Mythos in a previous edition as a potentially serious shift in AI-powered cybersecurity. That now looks less like speculation and more like a live case study: Mozilla says Firefox 150 shipped fixes for 271 vulnerabilities identified with help from Mythos Preview, while Bloomberg reports that a small group of unauthorised users has also been accessing the model through a third-party vendor environment and some fairly enterprising internet sleuthing.

Key Highlights

  • Mozilla says Firefox 150 includes protections for 271 vulnerabilities found using early access to Anthropic’s Mythos Preview.

  • Bloomberg reports that a small group of unauthorised users gained access to Mythos the same day Anthropic announced its limited release, using a mix of contractor-linked access and model-hunting techniques.

  • Anthropic has described Mythos as capable of identifying and exploiting vulnerabilities across major operating systems and web browsers, which is precisely why it had tried to keep access tightly restricted under Project Glasswing.

Why This Matters: This matters because it shows both sides of the AI-cyber equation arriving at once: these models can genuinely help defenders find and fix serious software flaws, but containing that capability is already proving messier than the labs would like to admit.

Kismet: The real twist here is that Mythos’s first big public storyline is not some neat triumphalist demo, but a perfect illustration of the whole dilemma, one team used it to harden Firefox, while another appears to have slipped past the velvet rope just to poke around. Very 2026. 👉 Full story here

Electromobility

Electric Cars Are Now Cheaper Than Petrol Cars

This is a genuinely big moment. In the UK, new electric cars are now on average cheaper to buy than petrol cars for the first time, which means one of the fossil era’s favourite talking points has finally run into the unpleasant inconvenience of maths. A growing wave of smaller, lower-cost EV models is helping pull the whole category closer to the mainstream.

Key Highlights

  • Autotrader says the average new EV now costs £42,620, versus £43,405 for a new petrol car, putting electric ahead by £785 on upfront price.

  • The shift has been driven by a mix of government grants, heavy manufacturer discounting, and brutal market competition, including pressure from Chinese brands.

  • Demand is responding: EVs made up 22% of UK new car sales in the first quarter, while interest on Autotrader’s new-car platform rose about 21% year on year in April.

Why This Matters: This matters because once electric cars are cheaper to run and now, increasingly, cheaper to buy, the economic case for petrol starts to look weaker, louder, and a bit more nostalgic than rational.

Kismet: More than half of Autotrader’s top ten most in-demand new models in April came from Chinese brands, which suggests this affordability milestone is not just about electrification, it’s also about a new competitive order arriving in the car market whether legacy automakers enjoy that or not. 👉 Full story here

Trump’s Iran War Is Driving Europe Electric

Turns out making petrol more expensive is a surprisingly effective EV marketing strategy. Electric car sales in mainland Europe jumped 51% in March, as higher fuel prices linked to the Iran war pushed more drivers towards the one drivetrain that doesn’t force them to bankroll geopolitical stupidity every time they fill up.

Key Highlights

  • 224,000 new EVs were registered in March across 15 European countries, while first-quarter sales hit 500,000, up 33.5% year on year.

  • Norway led the shift with 98% of new car sales fully electric in March, followed by Denmark at 76% and Finland at almost 50%.

  • The momentum is spreading well beyond the Nordics: Germany, France, Spain, Italy and Poland all posted a 40% increase in EV uptake in the first quarter, with Italy alone up 65% year on year in March.

Why This Matters: This matters because it shows, yet again, that oil dependence is not just a climate liability but an economic and strategic vulnerability, and every fuel shock makes electrification look less like idealism and more like self-defence.

Kismet: E-Mobility Europe says the shift to EVs so far this year has already cut annual forecourt demand by the equivalent of 2 million barrels of oil a year, which is the sort of number that turns “consumer choice” into energy geopolitics. 👉 Full story here

Sodium Batteries Are Coming. Lithium Isn’t Leaving.

Not quite. What’s happening looks less like a lithium obituary and more like the battery market getting more diversified, with sodium-ion finally becoming credible for some EVs, cold-weather markets, and grid storage, especially in China.

Key Highlights

  • CATL says it has boosted sodium-ion energy density by 50% and will begin mass production in the fourth quarter, while Changan plans to start selling sodium-equipped EVs by the middle of this year.

  • The IEA says 2026 could be a pivotal year for sodium batteries, with growing appeal as a hedge against volatile lithium prices and as a viable option for passenger EVs and stationary storage.

  • But sodium is not about to crush lithium. BloombergNEF expects sodium demand to rise to about 11GWh this year, yet Benchmark Mineral Intelligence still sees it making up only about 2% of total cell demand by 2030.

Why This Matters: This matters because sodium could lower battery supply-chain risk, improve affordability in lower-range vehicles, and perform better in extreme cold, but the bigger story is coexistence, not replacement. Lithium is still likely to dominate the premium and longer-range end of the market for years yet.

Kismet: A typical sodium-ion SUV may offer around 350km of range versus roughly 400 to 600km for lithium-based options, which means sodium’s sweet spot may be less “Tesla killer” and more “cheap, safe, good enough” which, irritatingly for incumbents, is often how disruptive technologies actually win. 👉 Full story here

Clean Energy

Renewables Just Overtook Coal. That’s Not Symbolic, It’s Seismic.

This is one of those moments future energy historians will circle in red. In 2025, renewables overtook coal to become the world’s largest source of electricity, and for the first time clean power growth pushed fossil-fuel generation into reverse because of structural change, not recession, lockdowns, or some other economic faceplant.

Key Highlights

  • Renewables surpassed coal across the full year in 2025, while coal’s share of global electricity fell below a third for the first time in history.

  • Wind and solar met 99% of global electricity demand growth last year, with solar alone providing 75% of that increase and posting a record 636TWh jump in generation.

  • Fossil-fuel power fell 0.2% in 2025, which Ember says is the first time that happened because clean energy expansion genuinely displaced it, rather than because the global economy fell down the stairs.

Why This Matters: This matters because it marks the point where clean electricity stopped merely growing fast and started forcing the old system backwards, which is a very different and far more consequential phase of the transition.

Kismet: Solar’s growth in 2025 alone exceeded the electricity that could have been generated from all LNG exports through the Strait of Hormuz, which is a deliciously awkward reminder that the fastest-growing energy source is now outpacing one of the world’s most geopolitically fraught fossil choke points. 👉 Full story here

India’s Wind Surge Just Moved It into the Global Premier League

India is quietly becoming one of the most important clean energy stories in the world. It added a record 6.3GW of wind in 2025, overtook both the US and Germany to become the largest wind market outside China, and now looks increasingly likely to hit its 100GW wind target by 2030.

Key Highlights

  • India commissioned a record 6.3GW of wind capacity in 2025, up 85% year on year, pushing it past the US and Germany in the global rankings outside China.

  • Multi-technology auctions combining solar, wind, and storage are helping drive the build-out, which is exactly the sort of system-level thinking energy planners usually discover only after making a mess first.

  • Suzlon says India is on track for 100GW of wind by 2030, with state utility demand alone expected to reach 107GW over the next five years, helped by wind’s ability to supply power when solar has clocked off for the evening.

Why This Matters: This matters because India is no longer just a giant future clean energy market. It is becoming a present-tense renewables powerhouse, and one whose success matters enormously for global emissions, grid stability, and energy security.

Kismet: One of the more revealing details here is that Suzlon is now pitching into Europe from Madrid, which means India’s wind sector is not just scaling at home, it is beginning to eye export relevance too. 👉 Full story here

China Didn’t Just Add Wind. It Practically Built the Global Market.

China added a staggering 120.5GW of new wind capacity in 2025, out of a record 165GW installed globally, which means it was responsible for nearly three-quarters of all new wind built last year. That is not a national energy story anymore, it’s a global one, and an awkward one for every government still pretending the transition is moving too fast.

Key Highlights

  • The world added a record 165GW of new wind capacity in 2025, up 40% on 2024.

  • China alone accounted for 120.5GW of that total, while Asia overall added 131GW, or about 80% of new global capacity.

  • Onshore wind did most of the heavy lifting, rising 42% to 155.3GW, while offshore wind grew 18% to 9GW.

Why This Matters: This matters because China is no longer just participating in the clean energy transition, it is increasingly determining its speed, scale, and industrial reality for everyone else.

Kismet: Even after this record year, the world still needs to reach 320GW of new wind a year to stay aligned with the goal of tripling renewables by 2030, so the really strange part is that a year this big still counts as not enough. 👉 Full story here

Storage

The Next Battery Race Is Still Open, and China Might Not Win It

I’ve been saying for a while that storage is the missing piece of the clean energy puzzle, and long-duration energy storage now looks like one of the few areas where China’s dominance is not yet locked in. Bloomberg’s take is striking: as storage shifts from a four-hour lithium game to systems that can store power for 10, 20, or even 100 hours, the US and Europe may still have a real shot, assuming the US can stop treating clean energies like a communicable disease. And it was good to see Hydrostor mentioned here too, especially since I had the Canadian long-duration storage company on the Climate Confident podcast earlier this year to talk about exactly why this technology matters for making clean energy dependable, not just abundant.

Key Highlights

  • Long-duration energy storage deployments are forecast to almost quadruple this year after a record 2025, as grids look for better ways to balance growing volumes of wind and solar.

  • China still dominates current cumulative long-duration capacity at about 72%, but the field is far more open than lithium-ion because the technologies are more varied and often need site-specific engineering rather than mass-produced commodity cells.

  • The US is seen as having the most diversified set of long-duration technologies in development, while countries including Germany and Italy are already creating policies to support deployment. Hydrostor’s compressed-air approach is one example of the kind of non-lithium pathway now getting real commercial traction.

Why This Matters: This matters because if long-duration storage scales, it could become one of the decisive enablers of high-renewables grids, and unlike solar panels or standard lithium batteries, this is still a sector where Western companies can plausibly build serious domestic industries instead of just watching China eat the market.

Kismet: One of the most interesting twists here is that long-duration storage may prove harder for China to dominate globally precisely because it is less like solar and lithium batteries. Bloomberg notes these systems often need local design, local knowledge, and local geology, which makes them much less exportable and much more contestable. 👉 Full story here

Latest blog post

My Latest Post: Every Oil Shock Makes Electrification Harder to Ignore

I published a new blog post this week arguing that oil shocks are no longer just economic pain, they are strategic advertising for electrification. The core point is simple enough to annoy fossil incumbents: wars and tanker disruptions may spike prices, but they also make solar, batteries, EVs, and heat pumps look less like climate virtue and more like basic business sense.

Key Highlights

  • I make the case that electrification is not a CSR side project but a risk-management strategy, because fleets, buildings, and businesses that electrify reduce exposure to oil and gas price shocks.

  • The data underpinning the piece is strong: solar accounted for more than 25% of global energy demand growth in 2025, low-emissions sources made up nearly 60% of demand growth, and battery storage additions hit 108GW, up 40% year on year.

  • I also pull together the transport story, from global EV sales topping 21 million in 2025 to the UK milestone mentioned above, where new EVs are now cheaper than petrol cars on average, showing that this is no longer fringe behaviour, it is scale beginning to bite.

Why This Matters: This matters because the energy transition is no longer being driven by climate arguments alone. It is increasingly being pulled forward by the harder logic of security, affordability, and resilience, which is usually when markets start paying attention.

Kismet: Global energy investment reached USD 3.3 trillion in 2025, with USD 2.2 trillion going to clean energy technologies and infrastructure, double the USD 1.1 trillion that went to fossil fuels, which is a useful reminder that capital has started reading the future, even if some politicians are still busy cosplaying the past. 👉 Full story here

Climate Confident:

Why Clean Grids Won’t Save Dirty Buildings

This week on Climate Confident, I spoke with architect Alexander Sexsmith, and the big takeaway was blunt: the real climate problem in buildings is no longer just how much energy they use, but what they’re made of, how healthy they are to live in, and whether they can survive a harsher climate. As grids get cleaner, concrete, petrochemical foams, and flimsy standard construction start to look less like normal practice and more like expensive, high-carbon bad habits.

Key Highlights

  • Concrete alone accounts for roughly 7 to 8% of global emissions, which is why embodied carbon is becoming a much bigger issue as electricity grids decarbonise.

  • We talked about regenerative materials like straw, hemp, cork, and hemp-lime, not as eco-aesthetic curiosities, but as serious contenders for lower-carbon, healthier, more resilient buildings.

  • Alex was crystal clear on the strategy: retrofit first, build durable for decades, and stop treating sustainability as a checklist when the real job is changing how the whole system works.

Why This Matters: This matters because the next phase of building decarbonisation is not just about swapping in cleaner electricity. It is about rethinking materials, resilience, health, and design logic before the built environment locks in another generation of avoidable emissions and vulnerability.

Kismet: One of the most counterintuitive points in the episode is that straw bale construction has proven a two-hour fire rating, which is a rather inconvenient fact for anyone still mentally trapped in the Three Little Pigs school of materials science (and yes, we had a laugh about that in the episode!). 🎧 Listen to the full episode

Resilient Supply Chain:

Why Most Logistics Teams Are Still Flying Half-Blind

This week on Resilient Supply Chain, I spoke with Constantine Komodromos about a problem far too many firms would prefer not to inspect too closely: most still don’t have one consolidated, real-time view of their logistics operations across cost, service, and emissions. And once you move beyond historical averages and look at actual shipment-level data, you start finding waste, weak contracts, low utilisation, and bad assumptions hiding in plain sight.

Key Highlights

  • Constantine argues that emissions data becomes genuinely useful only when it is granular, current, and tied to actual operations rather than industry averages or stale historical estimates.

  • We unpacked how real-time transport data can reveal hidden inefficiencies, from underfilled trucks and outdated carrier contracts to route changes, vessel speeds, congestion, and poor packaging design.

  • One of the strongest points in the episode is that sustainability is not a side report. It becomes a decision lever alongside cost and service, which means better emissions data can drive better operational decisions, not just prettier ESG slides.

Why This Matters: This matters because companies that still manage logistics on averages and instinct are not just weakening their emissions strategy, they are weakening their cost control, resilience, and ability to respond to disruption in real time.

Kismet: One of the most revealing examples in the episode is packaging. A packaging change made for product or procurement reasons can ripple straight through to transport utilisation, emissions, and cost, which is a neat reminder that some of the biggest supply chain gains live in the gaps between departments, where organisations traditionally excel at not looking. 🎧 Listen to the full episode

And… drum roll please, this week’s episode of the Resilient Supply Chain was the 500th episode

*Pats self on the back*

Wind and solar are complementary

Chile is making great strides in their renewables rollout - fortunately the new far-right government there are not anti clean energies.

File this one under intended consequences?

Misc stuff

I love this meme!

And, this is the week of the Feria de Abril in Seville, so if you find any mistakes in this edition of the newsletter, now you know why!

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

The cartoonists are having a field day lately

AI help, without the trust tax.

Most AI tools ask you to trade your data for intelligence. Norton Neo doesn't. It's the first safe AI-native browser built by Norton, and it gives you powerful built-in AI without handing your privacy over to get it. Search, summarize, and write with AI built directly into your browser. Your data stays yours. Your context stays private.

Built-in VPN, anti-fingerprinting, and ad blocking come standard. No add-ons. No setup. No compromises.

Fast. Safe. Intelligent. That's Neo.

Reply

Avatar

or to participate

Keep Reading