AI Is Local weather Tech’s Latest Hype Layer — The Industrial Issues Are Older

Help CleanTechnica’s work via a Substack subscription, on Patreon, or on Stripe. Assist us produce the entire high-quality, unique content material we publish week after week regardless of the challenges of content-scraping AI, delinquent media, inflation, and different hurdles.

Panthalassa raised $140 million this yr to construct autonomous ocean platforms that might generate electrical energy from waves and use it instantly for AI inference computing at sea. The proposition removes one real downside for wave power: as an alternative of transmitting electrical energy again to shore via subsea cables and grid connections, put a helpful electrical load beside the generator. The corporate now has sufficient capital to finish a pilot manufacturing facility and begin deploying its Ocean-3 methods.

That doesn’t take away the older issues of wave power. Gear nonetheless has to outlive salt water, fatigue, storms, corrosion and offshore upkeep whereas producing electrical energy cheaply sufficient to justify the equipment. Including AI compute strikes high-value electronics, energy conditioning and communications gear into the identical troublesome surroundings. Avoiding a cable might assist, however the related comparability remains to be in opposition to terrestrial compute equipped by grids, renewables and batteries fairly than in opposition to a fictional land knowledge middle with no electrical energy or interconnection prices.

AI is especially good at giving previous engineering propositions a brand new narrative as a result of AI itself is actual, helpful and attracting huge capital. Terrestrial knowledge facilities have real constraints round electrical energy, land, cooling and grid connections. That makes nearly any proposition that guarantees one other supply of energy or one other place to place compute sound newly related, together with wave-powered ocean knowledge facilities and computing infrastructure in orbit.

I’ve seen the identical commercialization transfer below completely different names earlier than. In 2023 I wrote about hydrogen supersonic passenger plane, solar-covered three-wheel vehicles, mine-shaft gravity storage buying a hydrogen-storage proposition, hydrogen eVTOLs and blockchain electrical energy markets. On the time, I handled a lot of the combination-making as comedy. Three and a half years of outcomes make it extra helpful as a diligence sample.

The attention-grabbing query isn’t whether or not AI, hydrogen, photo voltaic or nuclear are “actual applied sciences.” They plainly are. The complete TFIE Technique Briefing evaluation follows the unique 2023 firms ahead and attracts a extra helpful distinction: when does including one other expertise take away a cloth constraint, and when does it merely add one other dependency earlier than buyer worth can seem?

Destinus supplies one of many extra revealing outcomes. The corporate’s unique public narrative included liquid-hydrogen supersonic passenger aviation, combining an already troublesome industrial plane proposition with cryogenic gas storage, new propulsion, new gas infrastructure and certification work. Destinus didn’t disappear. As an alternative, its industrial centre moved decisively towards defence, autonomy and missile methods, and in April 2026 it agreed to kind a missile three way partnership with Rheinmetall.

That consequence doesn’t validate hydrogen supersonic passenger aviation. It demonstrates that a number of the underlying capabilities—propulsion, autonomy, high-speed flight, manufacturing and methods integration—can have paying markets when indifferent from the unique stack of dependencies. Destinus discovered clients ready to pay for vary, payload, velocity and scalable manufacturing in a distinct context.

Aptera illustrates a distinct model of the sample. Its body-integrated photo voltaic cells work, and below beneficial situations—assume Arizona ranch home with huge, unsheltered driveway the place the one locations are floor parking tons in malls, aka soul-deadening city sprawl with a aspect of local weather change-exacerbated warmth prostration—they will contribute helpful power. The tougher industrial denominator stays constructing and promoting an unconventional vehicle at scale. Aptera’s current filings say the corporate nonetheless wants roughly $40–45 million to finish tooling and validation for low-volume manufacturing and an additional $140–160 million to ramp its present facility towards about 20,000 autos per yr. That is, in fact, after 20 many years of promising buyers and early

The photo voltaic roof doesn’t must be fraudulent or ineffective for the advertising and marketing story to outrun its industrial contribution. Automobiles spend substantial time in garages, parkades, below bushes and beside buildings, whereas standard electrical autos preserve bettering their aerodynamics with out giving up unusual seating, cargo house and car structure. Photo voltaic contribution could be a function whereas car manufacturing, financing and broad shopper utility stay the figuring out questions.

Aptera didn’t chase each vogue. It stayed centered on the solar-EV proposition, whereas including a sequence of purported expertise differentiators round it: AI-assisted generative design, additive manufacturing, in-wheel motors and superior driver help. The since deserted in-wheel motors ought to have been a clue to how unserious the agency is, as the problems with unsprung weight at freeway speeds are extremely extensively understood, however they served the aim, getting extra press cycles for Aptera in order that they might get extra buyers to provide them extra money which they burned with out delivering something.

Prospects had been definitely not proof against the photo voltaic wrapping and different hype, nonetheless little it stands as much as actuality. Aptera gathered roughly 50,000 car reservations, with clients usually committing a refundable $100. The reservation programme has left about $4.1 million on the steadiness sheet as unearned reservation charges, not automotive income. A lot bigger sums got here from fairness buyers, together with programmes that provided precedence supply positions to folks investing tens of hundreds of {dollars}. Aptera’s lengthy, lengthy historical past of over promising, delivering nothing and never having any of the situations for industrial success was one of many focal components of a bit I revealed in 2022 trying on the junkyard of three-wheeled freeway autos. As I have a look at Aptera 4 years later, I see that it’s in precisely the identical place, over-promising, delivering nothing and having not one of the situations for industrial success.

Regardless of this lengthy and simply accessible historical past, Aptera has continued to generate remarkably favorable protection for a corporation that has but to promote a manufacturing car after 20 years of claiming it was on the verge of producing them. Over the previous yr it has been featured positively by credulous writers and editors on the Wall Avenue Journal, Washington Submit, Electrek, PV Journal and quite a few EV and expertise retailers, whereas a gradual stream of firm bulletins has been amplified via financial-news wires. It’s all hype and investor bait, not supply. (And sure, for Aptera followers, it’s performed this below three completely different industrial fashions, with the unique failing commercially, its mental capital being purchased for pennies on the greenback from the three founders by a Chinese language agency which additionally failed, then that mental capital being purchased again for pennies on the greenback by the three founders who’re on the third iteration of failing, with a enjoyable lawsuit about mental capital thrown into the combination.)

Gravitricity supplies a cleaner ending. Its unique mine-shaft gravity-storage concept not less than had a simple bodily structure: droop giant lots in present deep shafts and transfer them vertically. It later added H₂FlexiStore, an underground hydrogen-storage proposition drawing on a number of the similar subsurface experience. However gravity storage and pressurized hydrogen storage nonetheless required completely different gear, economics, clients and commercialization pathways, and the corporate entered collectors’ voluntary liquidation in October 2025.

It could be incorrect to say including hydrogen brought on the failure. The narrower commentary is sufficient: including hydrogen throughout its hype cycle didn’t produce one repeatable industrial enterprise earlier than the capital ran out. What seemed like diversification additionally meant pursuing two troublesome growth programmes with the identical finite administration consideration and financing.

The modern modifier modifications whereas the motivation stays. Blockchain as soon as gave electrical energy buying and selling, renewable certificates and peer-to-peer markets a brand new expertise layer despite the fact that meters, utilities, regulation and bodily community calculations continued doing the important work. Hydrogen later provided plane, fuel networks, engines and industrial areas a route right into a a lot bigger imagined future. Small modular reactors have been hooked up to hydrogen manufacturing, extending the chain once more from reactor commercialization to low-cost electrical energy to hydrogen manufacturing to no matter remaining market is meant to eat it.

AI now performs the identical narrative perform unusually effectively. Placing compute beside offshore technology does deal with transmission in a literal sense, however wave power’s historic downside has by no means been solely transmission. Placing compute in orbit avoids terrestrial grid connections whereas including launch prices, space-qualified electronics, radiation publicity, communications, alternative logistics and thermal rejection via radiators. Photo voltaic power in orbit is actual and launch costs can fall dramatically with out making general-purpose computing for terrestrial clients cheaper after a whole house programme is inserted between the processor and the person.

The important thing diligence distinction is due to this fact not between easy and complex methods. Electrical energy grids are difficult. Renewable technology, transmission, batteries, energy electronics and controls work collectively as a result of the elements carry out identifiable complementary jobs. Warmth pumps paired with thermal storage or electrical autos paired with managed charging are additionally integrations, however every addition addresses a measurable constraint.

Hype layering works in another way. The unique proposition stays troublesome, then a second immature or costly proposition is hooked up to it. The imagined market will get bigger earlier than the entire system turns into cheaper, less complicated or simpler to function. Each new dependency creates one other situation that should be glad on the similar time: hydrogen plane want plane, gas, airports and certification; offshore AI wants each the marine-energy system and high-availability computing to work; vehicle-integrated photo voltaic solely contributes materially when daylight and parking publicity cooperate.

That distinction issues as a result of startup danger isn’t the identical as portfolio diversification. Buyers can scale back danger by holding a number of impartial belongings whose outcomes are imperfectly correlated. An engineered product that requires a number of unsure subsystems to succeed collectively does the alternative. The chances multiply, interfaces add failure modes, and clients solely obtain worth after your entire chain features.

None of this requires cynical founders or fraudulent applied sciences, though definitely the previous helps and the latter is way from precluded. Choice strain is sufficient. Founders naturally emphasize variations of an organization that entice conferences and capital, buyers need publicity to giant present themes, governments desire initiatives hooked up to strategic industries, and journalists are extra curious about an uncommon mixture than the fifteenth iteration of an previous engineering downside. The additional layer may also postpone the uncomfortable comparability with an earthly various by including one other future situation that has not arrived but.

The sample has survived blockchain, hydrogen and now AI as a result of the narrative incentive is sturdy. The suitable response is to not reject each technological mixture. It’s to return repeatedly to the unique denominator: what constraint did the added expertise take away, how a lot price or working burden disappeared, and is the entire system now extra aggressive than the less complicated various?

When these solutions are measurable, the combination could also be genuinely productive. When they’re weak, the brand new layer could also be doing way more work within the financing narrative than within the product.

Learn the complete longitudinal evaluation and climate-tech diligence framework in TFIE Technique Briefing.