Midway through the 19th century, when the government of France set out to build Europe's most impressive railway system, there wasn't much doubt about how to do it: start in Paris, and build tracks in straight lines radiating outwards. "What a fine role for the state," the system's designer, Victor Legrand, rhapsodised, "if it can take charge and plan the main lines  and by this means of rapid, long-distance communications bring about the full integration of our fine country." But in crucial ways the "Legrand Star", as it came to be known, was a disaster. When war broke out with Prussia in 1870, Bismarck was able to transport nearly twice as many troops to the front as France, because Prussia's railways were a decentralised, unplanned hodgepodge of country lines; every French soldier, by contrast, had to travel via Paris. In the parlance of network theory, the Prussian system had "redundancy": multiple pathways to the same destination. The Prussians, you may remember, won that war.

The internet is the modern decentralised network par excellence, which helps to explain why the Silicon Valley air carries the distinct aroma of libertarianism. Many web entrepreneurs seem to view themselves as individualist frontiersmen on the unregulated plains of cyberspace, albeit with high-end mountain bikes. An extreme example is Peter Thiel, co-founder of PayPal, who dreams of escaping the burdens of government entirely by establishing a floating autonomous ocean colony, or "seastead".

But this identification of progress with free-market libertarianism, the technology writer Steven Johnson contends, is as mistaken as Legrand-style central planning. Real progress, he argues in Future Perfect, emerges from "peer networks" such as the internet because they're open and collaborative, not because of private, profit-motivated competition. His book is a call to support what he presents as a new kind of politics, based neither on traditional left-wing ideas of big government nor traditional right-wing ideas of big markets. "We believe in social progress, and we believe the most powerful tool to advance the cause of progress is the peer network," he writes, describing himself and his fellow-thinkers. "We are peer progressives."

This kind of declaration, coupled with Johnson's gung-ho title, is liable to induce a sinking feeling: is this another tract by a cyber-utopian, convinced that the web will solve all human ills? In the wake of the Arab revolutions of 2010 and 2011, for example, we've been subjected to a remarkably tedious argument about the role that social networks such as Twitter played in the unseating of dictators. Cyber-utopians argue that they made all the difference; mirror-image cyber-curmudgeons argue that they were irrelevant. But what normal person seriously doubts that the answer is "somewhere in between"? Johnson, refreshingly, is uninterested in such debates. The point about peer progressivism, he makes clear, is that it takes the internet not as a cure-all, but as a model. (And not the only model: the trading towns of the early renaissance were peer networks, too.) "A growing number of us," he writes, "have started to think that the core principles that governed the design of the net could be applied to solve different kinds of problems  the problems that confront neighbourhoods, artists, drug companies, parents, schools." These solutions are as likely to emerge from outside the market as within it; the internet itself, after all, was created not by some Thiel-esque superman, but by government-funded collaborations between academics and the military. Besides, private corporations can become Legrand Stars as easily as states. Even Facebook, on the inside, "prefers top-down control centralised in a charismatic leader".

What might a "peer progressive" world look like? Consider the ordinary challenges of an urban neighbourhood: potholes, noise pollution, broken street lights, empty wasteland. There are already numerous "crowdsourced" systems for alerting local councils to such problems, including New York's pioneering 311 phone line, or the splendidly named Australian website It's Buggered Mate. But what if residents were also empowered to spend a portion of their taxes on solving those problems, through a "crowdfunding" network modelled on the website Kickstarter? Members of the community  residents' groups, local builders, landscape gardeners  could propose improvement projects, and members of the community could collaborate to decide which ones to fund. This system of "participatory budgeting", similar to an approach already used in some parts of Brazil, doesn't count as big government: it shuns central planning, drawing wisdom instead from the edges of the network. But it isn't small government, either: it doesn't involve slashing taxes, nor a "Big Society"-style reliance on volunteers. It's a networked form of direct democracy  not the herd mentality of government-by-referendum, but a fine-grained, ongoing response to daily problems.

The principle could be extended, Johnson suggests, to election campaign financing. Perhaps we don't have to choose between centralised state funding and a free-market free-for-all. Why not give citizens the right to spend, say, £50 of what they've already paid in tax, in an election year, to the candidate or party of their choice?

Even in the world of commerce, the book shows, collaborative peer networks outperform free-market arrangements all the time. Johnson celebrates a resurgence of interest in "prize-backed challenges", funded by governments or private individuals, similar to the one that famously led to the solving of the problem of longitude. The best hopes for everything from a cure for Aids to viable private spaceflight may lie in such schemes. It is a central premise of such challenges that the resulting knowledge be shared, so that the network's "hive mind" can implement and improve it. The pure market approach would be to protect it with a patent. This is a key distinction between libertarians and peer progressives: many libertarians love patents. "All of that emphasis on freedom," Johnson writes, "vanishes when intellectual property is on the table."

There are still traces, in Future Perfect, of the cyber-utopianism that its title might imply; for example, Johnson never fully explains how his vision of networked democracy would avoid the perils of what Alexis de Tocqueville called "the tyranny of the majority". His brief mention of climate change, on the other hand, suggests that peer progressivism might not be able to solve the most pressing challenge of all. But the book is a wide-ranging sketch of possibilities, not a detailed policy prescription, and read as such, it's frequently inspiring. Above all, it's exciting to reflect on the possibility that the many achievements of the Silicon Valley revolution might be compatible, rather than in tension, with a progressive focus on social justice and participatory democracy. But then I always suspected that Peter Thiel's confidence in the power of the free market to devise the most elegant solutions to everything was misplaced. Have you tried using PayPal recently?

 Oliver Burkeman's The Antidote is published by Canongate.