The FT’s John Gapper recently wrote a perceptive piece calling time on the exceptionalism of the internet. In its early days, online was treated as a kind of la-la-land where information wanted to be free, new business models would proliferate and the laws of economic gravity did not apply. This magical thinking has served the big internet firms well, and they have done their best to keep it alive – not least in consumer naivety about the extent to which their online lives are tracked and sold on to others.

But as those companies grow bigger and fatter it becomes harder to ignore their real-world effects, and, to their indignation, regulators are increasingly taking a hand. As they should, if you think of Uber as a minicab firm, Airbnb as a hotelier and Deliveroo as a courier. Through the same lens Google and Facebook are publishers (with an advertising business model at least as old as newsprint). Amazon is a logistics company, albeit an incredibly efficient one.

Viewed from this angle, a number of things come sharply into focus. Increasingly, ‘disruptive’ looks like la-la lingo for regulatory, legal, tax or just semantic arbitrage. So part of the reason Uber is cheaper is that it books its UK orders through a Dutch subsidiary, which enables it to avoid charging 20 per cent VAT on fares. Uber London is really is just a minicab firm. Another is that, like Deliveroo, Airbnb, and Facebook, sheltering under the claim that it is an enabling not a transport (or food, hospitality or publishing) company, it can shrug off the responsibilities and costs of materials, insurance and above all employing people to drive passengers (or deliver pizza, wash sheets or write news).

But slowly, as Gapper notes, the real world is catching up. Uber and Deliveroo drivers are demanding employment rights through the courts – they may get them. Transport for London (TfL) wants Uber to observe the same standards for crime reporting, driver background checking and competition that it requires from other minicab firms. It is, it explains, nothing to do with the app or preserving in aspic the cab trade, which it and its predecessors have been regulating with some success since the 17th century. TfL has to bear in mind the added (and self-defeating) congestion that Uber generates, just as planners justifiably have something to say about ‘Rooboxes’ (Deliveroo’s ‘dark kitchens’ for cooking takeaways for posh restaurants) causing real-world noise and nuisance in residential areas – or for that matter whole swathes of cities turning into tourist dormitories through Airbnb (which is now getting into the business of developing apartment blocks for let).

If it complies with the rules, will Uber get its London licence? It would be interesting to be a fly on the wall for that conversation. It could of course could be argued that in a sane world Uber’s reckless management would be disqualified from being in charge of any sort of company. But leaving that aside, the regulator will have legitimate questions to ask in that most real of real-world domains: finance.

So far Uber has spent seven years and $13bn in its quest for global domination. Growth has been spectacular – but so has the cost. In the words of one academic investigation, ‘The growth of Uber is entirely explained by massive predatory subsidies that have totally undermined the normal workings of both capital and labor markets. Capital has shifted from more productive to less productive uses, the price signals that allow drivers and customers to make welfare maximizing decisions have been deliberately distorted, and the laws and regulations that protect the public’s interest in competition and efficient urban transport have been seriously undermined.’ In purely financial terms, the losses make the eyes water. In Q1 2017 Uber was applauded for ‘narrowing’ its losses to a mere $708m compared to nearly $1bn in the previous quarter. The total loss for 2016 amounted to around $3bn, after more than $2bn in 2015.

No one has satisfactorily explained how Uber could claw back deficits of that magnitude through normal means in an industry with razor-thin margins and a commodity product. (There are similar doubts about losses piling up at Deliveroo, albeit on a less massive scale.) So how come investors haven’t pulled the plug? One answer might be that venture capitalists in Silicon Valley have all succumbed to magical thinking. OK, me neither. The other explanation is that they really believe Uber’s fair-means-or-foul methods will take it to the promised land where network effects and increasing returns turn the ugly duckling into a winner that takes all, or at least most, of the money in its industry, before the cash runs out.

What’s riding on Uber therefore is much more than today’s riders and drivers. In most analyses, the only way Uber can succeed is by establishing the critical mass (read, quasi-monopoly) that would allow it to jack prices up to cover the full cost of rides – a jump of a beefy 40 per cent on present levels, according to some reckonings. At the same time it also needs to slash costs – which it why it is pouring money into self-driving (SD) technology, which will permit it to dispense with human drivers altogether. At that stage, of course, it won’t give a toss what anyone, including the courts, think of its employment practices – although it may come as a bit of a surprise to the thousands of London drivers whose support for its licence it is happy to claim today.

Uber, in short, is in a race against time, and, some would say, to the bottom. If it succeeds (or is allowed to) it will have legitimised an innovation model that destroys more value than it creates and turns what it does create into rents for the very few winners at the top. What riders gain as consumers they lose as producers – with SD they won’t even get gigs. To say that Uber should be subject to regulation is not anti-innovation – rather, it’s a call for regulation to get to grips with the new tech-enabled monopolies that are in reality old-style robber-baron rentier capitalism brought up to date. London is the theatre for this test case, in which we are both spectators and participants.