Writer Michael Barnard argues that AI has become the latest marketing layer added to climate technology projects, similar to what previously happened with blockchain, hydrogen, and modular nuclear power. The point is not that AI is useless, but that introducing it into an existing commercial proposition may give a project whose operating economics have not yet been settled a new story that is more appealing to investors and customers.
The analysis cites Panthalassa, which raised $140 million during 2026 to build autonomous platforms in the ocean that generate electricity from waves and use it directly to operate AI inference computing at sea. This design partially removes a real problem: the need to transmit electricity to shore through undersea cables and connect it to the grid. But it does not settle the rest of the equation, since the platform must still withstand salt water, stress, storms, corrosion, and marine maintenance while generating electricity at a cost that justifies the equipment.
What Changes in Practice?
Placing computing loads next to the energy source may reduce dependence on cables, but it moves high-value electronics, power-conditioning equipment, and communications and computing systems requiring high operational availability into the marine environment. Therefore, the project should not be compared with a hypothetical onshore data center that faces none of the costs of electricity, cooling, or grid connection. The more relevant comparison is with terrestrial data centers that obtain power from grids or renewable sources and batteries.
Data centers’ pressure on electricity, land, cooling, and grid connections gives these ideas initial significance. But the existence of a real infrastructure problem does not automatically mean that any new energy source or new computing location will be more competitive. The same applies to orbital computing concepts, which may avoid some constraints of terrestrial grids but add launch costs, space-qualified electronics, radiation exposure, communications, replacement, and heat disposal through radiators.
Examples of Changing the Market Rather Than Solving the Original Problem
Barnard offers Destinus as an example. In its initial presentation, the company was associated with supersonic passenger aircraft powered by liquid hydrogen. That required combining a difficult commercial aircraft, cryogenic fuel storage, a new propulsion system, fueling infrastructure, and certification work. According to the article, the company’s commercial focus later shifted toward defense, autonomy, and missile systems, and in April 2026 it agreed to establish a joint missile venture with Rheinmetall.
This shift, according to the analysis, does not prove the viability of hydrogen-powered supersonic passenger aircraft, but it shows that some core capabilities, such as propulsion, autonomy, high-speed flight, manufacturing, and systems integration, may find a market willing to pay for them when they are separated from the original chain of dependencies.
Aptera presents a different case. Solar cells integrated into the vehicle’s body may produce useful energy in favorable conditions, but the larger commercial challenge remains manufacturing an unconventional vehicle and selling it at scale. The company’s filings, as the article reports, say that it needs approximately $40–45 million to complete production tooling and validate limited production, and an additional $140–160 million to increase the capacity of its current facility to approximately 20,000 vehicles annually. The company has also collected approximately 50,000 reservations, generally with a refundable $100 deposit, while the balance of unearned reservation fees was approximately $4.1 million.
The writer uses these figures to distinguish between a genuine technical advantage and its ability to change the economics of the entire product. Solar power may be a useful addition, but it does not by itself settle questions of manufacturing, financing, user experience, and scalability.
The Test: Did It Remove a Constraint or Add Dependencies?
The analysis also cites Gravitricity, which developed an energy-storage concept involving lifting weights inside mine shafts and later added the H₂FlexiStore project for underground hydrogen storage. The writer argues that the two concepts required different equipment, economics, customers, and commercial pathways, and that the company entered voluntary liquidation from creditors in October 2025. The article does not conclude that adding hydrogen caused the failure, but rather that the addition did not produce a repeatable commercial operation before the capital ran out.
The editorial reading from certi.news is that the practical value of this analysis does not lie in rejecting technology integration, but in proposing a more rigorous test for it. The question should be which specific constraint the new technology removed, how much cost or operational burden it reduced, and whether the full system became simpler, cheaper, or easier to operate than the conventional alternative.
The addition makes sense when different components perform complementary functions that can be measured, as in combining heat pumps with thermal storage or electric vehicles with managed charging. But when the original product remains difficult and another immature or expensive technology is then added to it, failure points may multiply rather than risk being distributed as they would be in an independent investment portfolio.
This remains the writer’s analytical conclusion, not a general judgment on every project linking AI with energy or climate. The open question established by the source is whether these projects will provide sufficient operational and economic data to demonstrate superiority over simpler terrestrial alternatives, rather than merely offering a more attractive financing story.