NEAR Protocol

There is a particular kind of confidence that comes from having solved a problem nobody outside a small circle of cryptographers fully understands. NEAR Protocol has always carried that confidence, the quiet certainty of engineers who believe that if the maths is right, the market will eventually notice. Whether it has noticed enough, and on what terms, is the question that hovers over any conversation about this project.

Ask a NEAR partisan what the network is for and you will hear about sharding, about usability, about the unglamorous plumbing that supposedly determines whether blockchains can ever serve more than a few million enthusiasts. Ask a sceptic and you will hear about a crowded field, a token supply that keeps expanding, and a rebrand into artificial intelligence that reads, depending on your mood, as either shrewd repositioning or an admission that the original pitch was not quite enough.

Both readings can be true at once. That tension is the subject of this essay, and it is worth taking seriously precisely because NEAR has never been a meme or an accident. It was built by people who thought carefully about scaling, argued about it in public, and then had to watch the industry’s attention drift elsewhere.

The story so far

NEAR’s origins lie not in speculative finance but in a machine-learning competition. Illia Polosukhin, a former Google engineer who had worked on natural language processing, and Alexander Skidanov, an ex-MemSQL database specialist, met while trying to get computers to write code from examples. They kept running into the same wall: the infrastructure required to coordinate distributed compute and payment at scale simply did not exist in a form ordinary developers could use. That frustration, rather than any grand ideological project, is what pushed them towards blockchain design in 2017 and 2018.

What emerged was an architecture built around sharding from first principles, a technique called Nightshade that splits the network into parallel chains of computation while presenting a single unified chain to the outside world. The pitch was speed and low fees without sacrificing the decentralisation that made blockchains worth building in the first place. NEAR launched its mainnet in 2020, timed almost perfectly with the beginning of crypto’s broader renaissance, and the project raised substantial venture backing from names including a16z and Pantera, lending it a credibility that many rival layer ones envied.

The 2021 bull run was kind to NEAR, carrying it to an all-time high above twenty dollars as capital rotated hungrily through anything promising to be an Ethereum alternative. That peak became a kind of watermark against which every subsequent cycle would be measured, a reminder of how quickly sentiment can inflate a token’s price well beyond what its underlying usage might justify. The years since have been quieter and, in some ways, more revealing: a pivot towards positioning NEAR as infrastructure for artificial intelligence and user-owned data, an attempt to ride a newer wave of enthusiasm while the sharding thesis continued its slower, less headline-friendly rollout.

Polosukhin has remained the public face throughout, an unusually technical figurehead in an industry that often rewards showmanship over substance, giving the project a certain intellectual consistency even as its marketing has shifted with the prevailing winds.

The case for NEAR Protocol

The strongest argument for NEAR is essentially an engineering one. Sharding is genuinely difficult to implement without breaking composability, the property that lets smart contracts talk to one another seamlessly, and Nightshade’s design represents a serious, peer-reviewed attempt to solve that problem rather than paper over it with marketing. Believers point to transaction costs that remain a fraction of a cent and finality measured in seconds as evidence that the architecture works as advertised, not merely in whitepapers.

There is also the developer experience argument, often underrated by outsiders. NEAR’s account model, human-readable addresses, and support for both Rust and JavaScript were deliberately designed to lower the barrier for mainstream software engineers rather than only cryptography specialists, a decision that supporters argue positions the network well if blockchain adoption ever moves beyond financial speculation into ordinary applications.

Finally, the pivot towards artificial intelligence, however opportunistic it may look, is not entirely without logic. A network built for parallel processing and cheap coordination has plausible uses in orchestrating AI agents and verifying machine-generated work, and NEAR’s leadership has spent considerable energy trying to make that case to a market currently obsessed with all things AI.

The case against NEAR Protocol

The most persistent criticism is one of narrative fatigue. NEAR has been the layer one that is about to matter for several years now, and sceptics reasonably ask how many pivots and rebrands a project can undergo before the underlying uncertainty about product-market fit becomes the real story. Competing chains with simpler, less elegant designs have often captured more developer activity and liquidity, suggesting that technical sophistication alone rarely wins a crowded market.

Tokenomics draw scrutiny too. With circulating supply having grown past 1.29 billion tokens under an inflation model designed to fund validators and grants, dilution remains an ongoing concern for holders trying to judge long-term value against a backdrop of continuous issuance. Critics also note that much of NEAR’s early prominence rested on venture capital enthusiasm and a favourable market cycle rather than organic transaction demand, leaving open the question of what happens when that initial goodwill fully expires.

There is, too, a fair scepticism about the AI repositioning itself: rebranding a settlement layer as an AI network is easy to announce and much harder to prove, and the industry has seen enough opportunistic narrative-chasing to warrant caution before taking such claims at face value.

The bottom line

NEAR Protocol occupies an unusual position: technically credible enough to earn the respect of serious engineers, yet still searching for the kind of decisive adoption that would settle the argument in its favour. Its market capitalisation, currently above two and a half billion dollars, reflects a network that remains relevant without having conclusively won its category. Whether sharding and an AI pivot prove to be the right long-term bet, or simply the latest chapter in a long search for relevance, is a question only time and usage, not conviction, can answer. This is a work of journalism, not financial advice.