Emerging technology

Two pages in this section carry a year in their URL, 2018 and 2019. They sit at those paths because that is where they were first published and because the links pointing at them are real. What sits there now is not a restoration. The original text belonged to the previous owner of this domain and none of it has been reproduced. Each page instead revisits the claims that were circulating in its year and asks which of them held up, judged against what could be known at the time. The 2018 page and the 2019 page both work that way.

That is not a gimmick. It is the only quality control available to a section that writes about things which have not happened yet. A publication that forecasts and never returns to its forecasts is producing entertainment, and the cost of the entertainment is paid by whoever bought the roadmap.

The marking exercise produces a consistent finding. The technology predictions that failed rarely failed on capability. They failed on adoption economics, on regulation that arrived in a different shape than expected, or on the unglamorous fact that the incumbent process was cheaper than anyone admitted. The forecasts that held were usually the boring ones about cost curves.

Which sets the entry requirement for coverage here. A technology qualifies when at least one of three things is true. Somebody is buying it with real money and the transaction is documented. It has a bill of materials, a supply chain and a physical constraint that can be examined. Or a statute, regulator or standards body has begun to treat it as a category, which is usually the moment a curiosity turns into an obligation.

Demonstrations do not qualify. Neither do funding rounds on their own, which measure investor conviction rather than utility, nor conference keynotes, nor market size forecasts for years ending in zero, which are produced by firms whose customers are the vendors selling into the market being sized.

Dated commitments are the reportable kind

Britain's clearest current example of a checkable technology bet is quantum. The National Quantum Strategy commits £2.5 billion over ten years, and the strategy attaches five national missions with dates on them. By 2035, accessible UK-based quantum computers capable of running a trillion operations. By 2035, an advanced quantum network deployed at scale. By 2030, every NHS trust benefiting from quantum sensing solutions. By 2030, quantum navigation systems including clocks deployed on aircraft. By 2030, mobile networked quantum sensors in service across transport, telecoms, energy and defence.

Whatever one thinks of the likelihood, dating a commitment is an act of accountability, and it hands any publication a diary entry. The NHS sensing mission and the aircraft navigation mission both fall due in 2030, which is close enough that procurement for them is either happening now or is not. Where the funding has actually landed is assessed at Britain's quantum position, measured against the money.

The subjects that stopped being emerging

This section also keeps the older industrial subjects, the ones that quietly stopped being emerging without ever becoming ordinary, because they are the closest thing available to a controlled experiment. Virtual reality in business is the clearest case. The consumer story collapsed and a small number of industrial uses survived on their own economics, mostly in training where the alternative is expensive or dangerous. Which ones survived, and why those and not the others, is set out in the five business uses of virtual reality that stuck.

Vehicle electrification is a second case, and the interesting country is Norway, where the position was bought with a decade of tax policy rather than invented by a technology breakthrough. Reading it as a story about batteries misses the mechanism entirely, which is the argument at Norway's lead was bought, not invented. Shared mobility is a third: the fleets that survived the funding correction are run by a handful of operators whose economics look nothing like the pitch decks of the boom, and they are counted at who actually runs the world's bike share fleets. Ocean instrumentation is a fourth, and the least covered, because its buyers are survey firms, cable operators and defence rather than consumers.

There is a structural reason these retrospectives matter more than they should have to. The hype cycle in technology media runs in months and the procurement cycle in a large British organisation runs in years, so by the time a buyer can act on an idea the coverage that formed their impression of it is three years stale and was never revisited. A section that goes back and marks its own work is doing the one thing that closes that gap. The same evidential discipline that governs data and analytics governs this desk, for the same reason: without a route back to a source, a claim about the future is indistinguishable from a claim about the past that nobody checked.