Sponsored by
Hey everyone, it’s Thursday, oil is making renewables look clever again, and FutureProof is back.
This week is about backfire.
Trump’s war on Iran was supposed to be another flex of fossil-fuel power. Instead, it’s doing a remarkably good job of reminding the world why oil, gas and all the systems built around them are volatile, fragile, and strategically self-defeating. EVs look more resilient. Renewables look more dependable. Electrification looks more strategic. Awkward.
The big threads:
Trump’s oil war is pushing more people towards EVs, batteries, and clean tech instead of back to fossil comfort blankets.
AI keeps getting sharper, from brain mapping and better voice models to materials science and atomic-scale defect detection.
Clean energy keeps scaling, with global renewables now brushing 50% of electricity capacity and solar efficiency pushing into territory that used to sound faintly ridiculous.
Electrification is spreading into harder sectors too, from long-haul trucking in Australia to agrivoltaics showing, again, that food and solar are not enemies.
Plus: carbon markets try to rebuild credibility, battery costs keep collapsing, and audio AI is finally starting to sound less like a hostage reading a script.
And on the podcast front, yesterday’s Climate Confident episode with Jen Jenkins of Rubicon Carbon is out now, while a new Climate Confident+ bonus episode drops tomorrow for all subscribers, diving into a timely story and unpacking the wider global implications.
Let’s get into it.
Climate
Net Zero Isn’t the Expensive Option. Fossil Fuel Chaos Is
The UK’s Climate Change Committee has just delivered the sort of analysis fossil fuel nostalgists will hate: the total added cost of reaching Net Zero by 2050 is likely to be no bigger than the cost of a single fossil fuel price shock on the scale of 2022. Even better, the maths gets ruder for the sceptics the deeper you look, because the CCC found that for every £1 spent on Net Zero, the benefits come back at roughly £2.2 to £4.1.
Key Highlights
A single 2022-style fossil fuel price spike could cost as much as the entire net additional cost of the UK’s Net Zero pathway through to 2050.
The CCC says the Balanced Pathway to Net Zero delivers a benefit-cost ratio of 3.3 overall, with benefits still outweighing costs across every sensitivity tested, even when offshore wind or grid costs rise.
Energy losses in a Net Zero system are projected to be cut in half, falling from about £60 billion a year today to £30 billion a year by 2050, because electric technologies waste far less energy than fossil-fuelled ones. The report also says the pathway starts delivering net benefits from 2029.
Why This Matters: This matters because it flips the entire “Net Zero is too expensive” argument on its head: the real luxury belief is pretending permanent dependence on volatile fossil fuels is somehow the cheaper, safer option.
Kismet: One of the most quietly devastating details in the supplementary report is that an electric car bought in 2045 would still be delivering savings into 2060, which means even the CCC says its 2050 cut-off probably understates the real long-term payoff of the transition. 👉 Full story here
The Fossil Fuel Grab That Backfired
I keep seeing the lazy take that the Iran war is pushing countries back to coal. That’s too glib by half. What the evidence actually shows is messier and, in several markets, far more interesting: yes, some governments are leaning on whatever fossil kit they already have for short-term resilience, but the bigger structural reaction is a sharpened sprint toward renewables, storage, electrification, and home-grown power because imported fuel suddenly looks every bit as risky as it always was.
Key Highlights
In Vietnam, Vingroup wants to scrap plans for a 4.8GW LNG plant and replace it with a renewables-plus-battery project because the war has made LNG price risk look brutally exposed, with Reuters reporting LNG prices up 85% since the strikes began.
India is accelerating clearances for wind and battery storage as gas supply tightens, while in the UK record wind output has helped cut gas generation and cushion the power system from the worst of the fossil fuel shock.
On the consumer side, Britain is seeing the same pattern in miniature: Octopus says solar sales are up 54% month-on-month since the war began, alongside a jump in heat pump and EV charger sales, which is not exactly the behaviour of a public rushing lovingly back into fossil dependency.
Why This Matters: The real story here isn’t “coal comeback”; it’s that every oil and gas shock keeps doing renewables’ marketing for them, because energy security arguments land faster when households, utilities, and governments are getting mugged by fuel prices in real time.
Kismet: The most delicious twist may be in Asia: Bloomberg’s David Fickling argues the LNG shock is more likely to trigger a rooftop solar boom than any meaningful return to coal, because the missing Gulf gas volumes are small in the context of Asia’s overall coal market, while higher electricity prices make photovoltaics suddenly look like the adult in the room. 👉 Story Links Inline
AI News
Meta Says It Can Predict Your Brain. Slightly Chilling, Definitely Fascinating
Meta has unveiled TRIBE v2, a tri-modal AI model trained on video, audio and language that can predict human brain activity from fMRI data with startling accuracy across a huge range of tasks and subjects. This isn’t mind-reading in the sci-fi sense, but it is a serious leap towards using AI as an in-silico lab for neuroscience, which is both wildly promising and just unsettling enough to keep things interesting.
Key Highlights
TRIBE v2 was trained on more than 1,000 hours of fMRI data from 720 people and uses video, audio and language inputs to predict high-resolution brain responses across naturalistic and experimental settings.
The model outperformed strong linear encoding baselines, generalised to unseen subjects and tasks, and could be fine-tuned with limited new data to improve individual predictions.
In simulated neuroscience experiments, it reproduced classic findings including face, place, body and word-form regions in vision, and language-related contrasts involving speech, syntax and meaning.
Why This Matters: This matters because it points to a future where AI doesn’t just analyse neuroscience data after the fact, but helps researchers test hypotheses, design experiments, and map cognition far faster than the old one-model-per-task grind ever allowed.
Kismet: One of the cleverest bits is that the model’s internal components spontaneously lined up with well-known brain networks like language, vision, motion and the default mode network, which is a lovely reminder that sometimes the black box coughs up something interpretable before the humans have finished arguing about the labelling. 👉 Full story here
AI Just Found the Flaws That Make Better Materials
MIT researchers have built an AI model that can spot and measure atomic-scale defects in materials without cutting them open, which is a big deal because those tiny imperfections often determine whether a semiconductor, solar cell or battery material performs brilliantly or like a very expensive disappointment. Better still, the model can identify up to six point defects at once, which is exactly the kind of fiddly, high-value problem AI is annoyingly good at.
Key Highlights
The MIT team trained an AI model on 2,000 semiconductor materials and used neutron-scattering data to classify and quantify defects non-destructively.
The system can detect up to six different point defects in a material simultaneously, with concentrations as low as 0.2%, which conventional methods struggle to do.
The researchers say this could improve how manufacturers optimise semiconductors, microelectronics, solar cells and battery materials, and they now want to adapt the approach to more industry-friendly techniques like Raman spectroscopy.
Why This Matters: This matters because materials performance often comes down to defects people can deliberately tune but still can’t properly measure, and AI may finally be turning that from educated guesswork into something closer to engineering.
Kismet: The elegant twist here is that defects are not the enemy at all. They are more like seasoning: the right amount can transform the material, but too much and you’ve ruined the dish. 👉 Full story here
AI Voices Are Finally Losing the Robot Accent
Audio AI is getting a lot less clunky, and not a moment too soon. Mistral has launched Voxtral TTS, a multilingual text-to-speech model focused on natural, expressive voice generation, while Google has rolled out Gemini 3.1 Flash Live with lower latency, stronger reasoning, and more fluid real-time dialogue. Different products, same direction: voice AI is moving from “technically impressive” to “actually usable.”
Key Highlights
Mistral’s Voxtral TTS supports 9 languages, can adapt to a new voice from as little as 3 seconds of reference audio, and is built for low-latency streaming with a reported model latency of 70ms.
In Mistral’s human evaluations, Voxtral beat ElevenLabs v2.5 Flash on naturalness and voice customisation, showing listener preference win rates of 58.3% for flagship voices and 68.4% for voice customisation.
Google says Gemini 3.1 Flash Live improves both audio quality and task execution, scoring 90.8% on ComplexFuncBench Audio and 36.1% on Audio MultiChallenge, while also powering longer, faster voice conversations across Gemini Live and Search Live in more than 200 countries and territories.
Why This Matters: This matters because voice is quietly becoming the next serious interface layer for AI, and once latency, tone, multilinguality and reliability improve together, audio stops being a gimmick and starts becoming infrastructure.
Kismet: The sneaky important detail is Google watermarking all 3.1 Flash Live audio with SynthID, which means the race to make AI voices sound more human is now happening alongside a parallel race to make sure we can still tell when the human isn’t there at all. 👉 Story Links Inline
Electromobility
This Oil War Is Backfiring Into an EV Boom
Funny how a fossil fuel shock keeps turning into an electric vehicle ad. As petrol and diesel prices jump, consumers across Europe, Asia-Pacific and Australia are taking a fresh look at EVs, not because they’ve suddenly become saints, but because “pump anxiety” is starting to beat “range anxiety” in the real world.
Key Highlights
Consumer interest is rising fast: CNBC reports a 28% jump in new EV inquiries and a 15% rise in used EV inquiries on Autotrader, while Octopus Electric Vehicles saw leasing inquiries rise 36% after the war began.
In the UK, the FT reports advert views for new BYD models rose 77% year on year, used BYD searches jumped more than 375%, Renault saw EV model enquiries rise 24%, and Kia said EV test drive requests surged 84% since February 2025.
Reuters says the shift is broadening across Asia-Pacific: Australia saw a 100% uptick in EV loans in March, New Zealand registered more than 1,000 EVs in a week, close to double the previous week, and South Korean EV registrations more than doubled in March from a year earlier.
Why This Matters: This matters because every oil shock is a live demonstration that transport electrification is not just a climate strategy, it is a resilience strategy, and one that gets more persuasive every time fossil fuel markets throw another tantrum.
Kismet: The most quietly radical version of this story may be buses, not cars: as diesel tops A$3 a litre, Australia is being pushed to electrify bus fleets that currently consume about 530 million litres of diesel a year, which is a reminder that the biggest wins often come not from the shiny consumer product, but from the boring public infrastructure everyone forgot to fix. 👉 Story Links Inline
Australia’s First Long-Haul Electric Truck Just Embarrassed Diesel
Australia has just seen its first all-electric inter-city freight delivery, and the really awkward bit for diesel diehards is this: the electric truck was not only dramatically cheaper to run, it was faster too. A Windrose electric prime mover hauled a bulk load from Sydney to Canberra in a single charge, with the full warehouse-to-customer run delivered using electric vehicles end-to-end, which is precisely the sort of proof point that turns “someday” into “that excuse is dead now.”
Key Highlights
The Sydney-to-Canberra delivery covered about 460km using a 700kWh Windrose electric prime mover plus electric last-mile vans, marking what New Energy Transport says is Australia’s first all-electric long-haul delivery.
Energy costs for the trip were reportedly 85% lower than for a comparable diesel truck, and the electric truck completed the route 25 minutes faster because it could maintain speed on steep climbs.
The economics are starting to look brutally clear: the article says a diesel prime mover might cost about A$250,000 upfront but rack up around A$2 million in fuel over 10 years, while the electric version costs more upfront, around A$450,000-A$500,000, but slashes running costs.
Why This Matters: This matters because freight is one of the hardest sectors to clean up, and once electric trucks start proving they are cheaper, faster, quieter, and less exposed to diesel price chaos, the debate shifts from “can this work?” to “why are we still pretending diesel is the safe option?”
Kismet: The sneaky twist is that long-haul may actually suit electric trucking better than many last-mile routes, because when you’re dragging 60 tonnes over huge distances, diesel’s so-called advantage starts looking less like strength and more like an expensive habit. 👉 Full story here
Clean Energy
Clean Energy Nearly Hit Half the Grid. And That Was Before 2026 Got Messy
Renewables reached 49.4% of global electricity capacity in 2025, driven by another monster year for solar, which added 511GW on its own. If that happened before this year’s fossil fuel chaos really bit, 2026 could get very interesting indeed, because nothing sharpens the appeal of domestic clean power quite like watching oil and gas markets lose the run of themselves.
Key Highlights
Global renewable power capacity hit 5,149GW in 2025, up 692GW from 2024.
The annual growth rate for renewables rose to 15.5% in 2025, up from about 15.1% in 2024.
Solar was the standout again, growing by 511GW to 2,392GW, while wind added 159GW to reach 1,291GW.
Why This Matters: This matters because renewables are no longer creeping forward politely. They are now close to half of global electricity capacity, and every fossil fuel shock makes the economic and strategic case for building more of them even harder to ignore.
Kismet: The quietly huge twist is that the growth rate needed from 2025 to 2030 to hit the COP28 tripling goal is 16.6% a year, so at 15.5% in 2025 the world is no longer wildly off course, just infuriatingly close enough that policy failure, not technology, looks like the main risk. 👉 Full story here
Solar’s Efficiency Ceiling Just Got a Hole Blown In It
Researchers in Japan say they’ve developed a new material that could let solar cells squeeze far more energy from sunlight than conventional designs, using a clever “spin-flip” emitter to capture energy that would normally disappear as waste heat. If it scales, this is the kind of advance that doesn’t just make solar cheaper. It makes the whole technology stack more formidable, which fossil incumbents must find deeply annoying.
Key Highlights
A team at Kyushu University developed a material using a “spin-flip” emitter to harvest solar energy that conventional cells usually lose as heat.
The approach uses singlet fission, splitting excitons from higher-energy photons into two lower-energy excitons, theoretically doubling the usable energy from part of the solar spectrum.
The article says the breakthrough could push solar conversion efficiency to 130%, at least in theory, smashing through the longstanding limits of conventional solar cells.
Why This Matters: This matters because solar is already winning on cost and scale, so every serious efficiency improvement now has a multiplier effect: fewer panels, more output, better economics, and an even stronger case for electrifying absolutely everything that can be electrified.
Kismet: The fun part is that solar’s reputation is still often shaped by yesterday’s panels, while the actual technology keeps evolving like consumer electronics, with chemistry and physics quietly turning rooftop kit into something that would have looked absurdly futuristic not that long ago. 👉 Full story here
Agrivoltaics Keeps Winning the Dumbest Culture War in Energy
I keep seeing solar and farming framed as if they’re locked in some zero-sum cage match. They’re not. This four-year Italian field study on potatoes found that agrivoltaics can preserve healthy yields, and in some cases even match or beat full-light plots, provided the shading is managed intelligently. In other words, the panels and the plants can get along just fine, which must be terribly inconvenient for people who prefer shouting to data.
Key Highlights
A four-year field trial in northern Italy found that standard agrivoltaic shading caused only limited yield penalties on potatoes, with average losses of about 12%, while heavier shading pushed yield reductions above 30%.
The real breakthrough was dynamic light management: in 2024, an anti-tracking setup reduced shading during the critical tuber initiation phase, and yields under this setup reached 32.7 t/ha, compared with 30.3 t/ha under full light and 29.9 t/ha under standard tracking.
The study found that timing of shading mattered more than seasonal average shading, showing that crop physiology and panel operation can be coordinated rather than treated as opposing forces.
Why This Matters: This matters because agrivoltaics is shaping up not as a compromise, but as a design problem, and once you solve the design properly, you stop asking whether land should produce food or power and start asking why it shouldn’t do both.
Kismet: The quietly brilliant twist is that moderate early-season shading can delay soil moisture depletion and improve water-use efficiency, which means in a hotter, drier world, a bit of well-timed shade may turn out to be a feature, not a bug. 👉 Full story here
Energy Storage
Batteries Didn’t Just Get Better. They Got 99% Cheaper
One reason clean energy is suddenly looking far less “alternative” and far more inevitable is brutally simple: batteries have become astonishingly cheap. According to Our World in Data, lithium-ion battery cell prices have fallen 99% since 1991, turning storage from a niche luxury into the missing piece that makes EVs, home batteries and grid-scale battery energy storage genuinely scalable.
Key Highlights
Lithium-ion battery cell prices fell from about $9,200 per kWh in 1991 to just $78 per kWh in 2024, a decline of more than 99%.
Battery prices have followed a strong learning curve, falling by roughly 19% every time cumulative global production doubled.
Energy density has more than tripled since the 1990s, meaning batteries now store far more energy in less space and weight, opening the door not just for cars, but for trucks, ships and grid storage too.
Why This Matters: This matters because cheap batteries are what turn renewables from low-cost generation into reliable, dispatchable energy systems, and once storage gets affordable enough, the old argument that solar and wind are too intermittent starts looking increasingly like a fossil-era coping mechanism.
Kismet: The really wild part is that this revolution was not initially driven by climate policy at all, but by consumer electronics, meaning your old mobile phone and laptop helped drag battery technology down the cost curve until grid-scale storage and electric transport became economically viable. 👉 Full story here
Latest Podcasts
The
. 👉 Full story here
Climate Confident:
Can Carbon Credits Actually Earn Trust Again?
Voluntary carbon markets have become one of climate’s messiest battlegrounds, so I sat down with Dr Jennifer Jenkins, Chief Science Officer at Rubicon Carbon, to separate the serious from the sketchy. We got into what makes a credit high quality, why the market’s reputation took such a hammering, and why long-term offtake deals may be a sign that this market is finally growing up, or at least trying to.
Key Highlights
Jen argues that carbon markets only make sense after companies have pushed internal decarbonisation as far as is practical, using credits to close the gap where the technology or cost curve still won’t cooperate.
We unpacked what quality actually means in carbon markets, from additionality and leakage to why methodology reform, ratings agencies, and the ICVCM’s Core Carbon Principles matter.
We also explored why buyers like Microsoft are locking in long-term supply now, and how voluntary and compliance markets may start converging much faster over the next few years.
Why This Matters: Carbon markets are either a useful flexibility mechanism or a dangerous distraction, and the difference comes down to integrity, transparency, and whether buyers are using them to accelerate decarbonisation rather than dodge it.
Kismet: One of the more interesting takeaways was Jen’s point that the real sign of success in five years would not just be better projects, but far more companies actually setting serious climate targets in the first place because markets give them a credible path to hit them.
And a quick heads-up: I’ll be dropping a bonus episode tomorrow for Climate Confident+ subscribers only, where I dig into a topical story and unpack the wider global implications. 🎧 Listen to the full episode
Resilient Supply Chain:
This War Isn’t Just a Crisis. It’s a Supply Chain X-Ray
In my debut Resilient Supply Chain+ bonus episode, I dug into the war on Iran not as a news event, but as a business stress test, for energy dependence, maritime choke points, fertiliser supply, inflation, and the brittle assumption that fossil fuels will always stay cheap and available. It’s a subscribers-only episode, and this is exactly what these bonus editions are for: timely analysis on fast-moving stories and the global implications supply chain leaders need to understand now, not months from now when the damage is already baked in.
Key Highlights
I unpack how the war is disrupting far more than oil and gas, including jet fuel, marine fuels, fertiliser inputs, sulphur, petrochemicals, helium and wider industrial supply chains.
The episode argues that reshoring manufacturing is not enough if firms remain dependent on imported fossil fuels and fossil-linked inputs moving through militarised choke points.
I make the case for “reshoring energy” too: more domestic renewables, storage, electrification, grid investment and flexibility, because resilience and energy strategy are the same conversation now.
Why This Matters: This matters because the companies that treat energy as a strategic dependency rather than a utilities line item will be far better placed for the next shock, and there will be a next shock, because apparently we remain committed to learning everything the hard way.
Kismet: One of the sharpest points in the episode is that oil shocks are loud but fertiliser shocks are quieter, slower and in some ways more destabilising, because they creep into food systems with a delay and show up later as inflation, weaker harvests and political stress. 🎧 Listen to the full episode
Featured Chart(s)
Cherry blossoms are blooming earlier due to climate change
Oil prices are spiking since the US and Israel embarked on their half-baked venture to attack Iran
This is interesting.
I wouldn’t have bet on Hungary!
China is becoming the world’s first electrostate.
Meanwhile Red Electrica España (the Spanish grid operator) reports that today the Spanish grid is operating on 85.7% clean energy sources (wind, solar, nuclear, and hydro)
Misc stuff
We’ve all seen the movie the eggsorcist, right?
So, KitKat got robbed and the Nestle social media team posted the following. Lots of other brands piled on to have a laugh at this event:
And finally:
Nestle Australis posted this on April 1st!
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
In honour of yesterday
How Jennifer Aniston’s LolaVie brand grew sales 40% with CTV ads
The DTC beauty category is crowded. To break through, Jennifer Aniston’s brand LolaVie, worked with Roku Ads Manager to easily set up, test, and optimize CTV ad creatives. The campaign helped drive a big lift in sales and customer growth, helping LolaVie break through in the crowded beauty category.