Internet Computer

There is a particular kind of confidence required to look at the internet — the whole edifice of it, the servers and clouds and data centres humming away in industrial parks from Virginia to Inner Mongolia — and declare that you intend to replace it. Not augment it, not carve out a niche within it, but rebuild its plumbing entirely on a public blockchain. This was, roughly, the pitch behind Internet Computer, and it arrived not with the usual crypto-native modesty but with the swagger of a project that believed it was solving computing itself.

Sceptics rolled their eyes. Believers rolled up their sleeves. Both reactions, in their way, tell you something true about a token that has spent years oscillating between being dismissed as an overengineered curiosity and being defended as the most technically ambitious thing in the entire industry.

The story so far

Internet Computer’s origins trace back to Dfinity, a research-heavy foundation established in Zurich by Dominic Williams, a British computer scientist with a taste for grand theoretical claims and an evident willingness to spend years — and hundreds of millions of dollars in funding from the likes of Andreessen Horowitz and Polychain — chasing them. The idea, formally unveiled around 2016 and refined over half a decade of cryptographic research, was to build what Williams called a “world computer”: a decentralised network capable of hosting entire websites, software services and applications directly on-chain, without the intervening layer of Amazon Web Services or Google Cloud that quietly underpins most of the internet as we know it.

The technical scaffolding involved was genuinely novel — chain-key cryptography, a network of independent data centres organised into subnets, and a canister-based smart contract model designed to let code run at web speed rather than at the glacial pace typical of blockchains. When the mainnet finally launched in May 2021, after years of anticipation, it arrived with a token generation event that briefly gave ICP one of the largest market capitalisations in crypto, propelled by pent-up expectation and a headline-grabbing all-time high of $750.73.

What followed was a rather brutal correction, as the gap between the project’s stated ambitions and its early, sparsely populated ecosystem became impossible to ignore. Dfinity’s response was to keep building rather than retreat: expanding subnet capacity, courting developers with grants, and pushing use cases from decentralised social media to on-chain artificial intelligence experiments. The network’s governance, run through a system called the Network Nervous System, has continued to evolve, with token holders locking ICP into “neurons” to vote on upgrades — a structure the project treats as central to its claim of genuine decentralisation.

Williams himself has remained a near-constant public presence, alternately combative and evangelical, defending the project against critics on social media with the same intensity he once used to pitch it to venture capitalists.

The case for Internet Computer

For its advocates, Internet Computer represents one of the few blockchain projects genuinely attempting to solve a problem of civilisational scale: the concentration of the internet’s infrastructure in the hands of a small number of American technology companies. A network capable of hosting fully on-chain applications, they argue, offers a credible alternative — one where censorship resistance and data sovereignty are not marketing slogans but architectural properties. The engineering underneath is not trivial either; chain-key cryptography and reverse-gas models, where users interact with applications without needing tokens for every transaction, address real usability problems that have hobbled other smart contract platforms.

There is also the matter of persistence. Long after the 2021 hype cycle collapsed, Dfinity continued shipping upgrades and expanding developer tooling rather than quietly winding down, which believers take as evidence of a team playing a longer game than most in an industry notorious for abandoning projects the moment the token price sours.

The case against Internet Computer

The case against is, frankly, difficult to avoid, given how spectacularly the token’s launch-day valuation collapsed and how long the recovery has taken. Critics point to a familiar pattern: extraordinary technical claims, a huge pre-mine allocated to insiders and early backers, and a subsequent price trajectory that left retail buyers who arrived at the top nursing enormous losses. Whether the “world computer” framing was ever realistic, or simply an effective story for raising capital in a bull market, remains a live question among sceptics who note that adoption — actual applications with actual users — has grown far more slowly than the original pitch implied.

There are also more technical objections: concerns about the degree of centralisation in node operation, the complexity of the governance model, and whether “on-chain everything” is actually a desirable design goal when simpler, cheaper alternatives exist for most of what developers want to build. Detractors further note that with a circulating supply north of 554 million tokens against a maximum near 469 million under the network’s own accounting — a discrepancy tied to its unusual inflation and burning mechanics — the token’s supply dynamics remain harder for casual observers to parse than those of most major cryptocurrencies.

The bottom line

Internet Computer occupies an unusual position in crypto’s landscape: too technically serious to dismiss outright, too commercially unproven to vindicate fully. With a market capitalisation of roughly $1.2 billion, it remains a mid-tier project by industry standards, still searching for the kind of organic adoption that would settle the argument between its engineers and its critics. Whether it ultimately becomes essential infrastructure or a cautionary tale about the gap between whitepapers and reality is a question the market has not yet answered, and probably will not for some years. This article is journalism, not financial advice.