Dear reader,

Sometime in the early 2000s, while I was in college, Gmail came to me wearing a small grey label that read beta. I had met the word before, in a hazy way, in the old engineering sequence of alpha, beta, gamma—the stages a thing passed through before it was deemed fit to face the world. Gmail fixed the word in my head for good, and I expect many of you remember it too. The label sat in the corner of the screen, hour upon hour, a soft confession that the instrument into which I poured letters to friends, recruiters, and editors had not in fact been finished. It might err. It would learn from its errors. It would, one was given to understand, improve.

What surprised me even then was that the confession remained steadfast. Gmail launched on the first of April, 2004, and wore its beta tag until the seventh of July, 2009—more than five years of being officially incomplete while a hundred million entrusted their correspondence to it. When Google at last took the label off, the engineers gave a reason of almost touching banality: cautious business customers disliked committing to anything that sounded unready. The code had scarcely altered. Only the word was gone.

I remember wondering how this could be permitted. A service reaches the public before it is done, and the public says thank you. We forgive it because it is free, and because no calamity follows a late email. But would one board a beta aeroplane? Hand over our body to a beta surgeon? Cross a bridge that was described, with a shrug, as a work in progress? The questions sound foolish, and the foolishness is instructive. It marks the precise corners of life that we agreed to treat as provisional.

Tim O’Reilly, who all but named the age when he gave us “Web 2.0”, had a phrase for the arrangement: the perpetual beta. By 2005, the old object you carried home in a box and installed once was already a memory; software had become an always-on service, forever revised. The dictum “release early and release often”, he wrote, had hardened into something more radical—the product developed out in the open, never closed. And the cunning of it sat in a single line of his: users must be treated as co-developers. We were not customers taking delivery of a finished good. We were the unpaid testing department, our grumbles and our crashes feed the next version. We had been put to work, and no one had mentioned the appointment.

For a time this did no great harm, confined to the free tools of the web with low stakes. But it bred a certain cast of mind. The generation that shrugged at Gmail’s grey label has, 20 years later, taken up ChatGPT (and other AI tools) and accepted all their confident fabrications—even though they are all still very much unfinished.

Does this mindset reach deeper than the software? Has the beta worked its way into the human temperament itself, until we give up waiting for the finished version of anything at all: an institution, a marriage, or nature itself? Zygmunt Bauman, the Polish-British sociologist, caught the shape of this from another direction. Ours is a liquid age, he said, in which every settlement is a passing one, held—here is his lovely, chilling phrase—“until further notice”. A century ago, to be modern was to chase a final state of perfection; now it is to pursue an infinity of improvement, with no final state in sight and none desired. Read that with Gmail’s grey label in mind. No final state desired. The beta is not a stage on the way to somewhere. It is the place where we live.

The mind that takes the world as beta is easy, careless, irresponsible. It imagines there’s a fix for everything. When Facebook came to Myanmar, it sped hate through the population at a velocity no one had witnessed. In 2018, the UN found Meta had played a “determining role” in the slaughter of the Rohingya; Meta itself confessed finally that it had been too slow to respond. Some 7.5 lakh Rohingya people were driven from their homes. There was no patch for it in the next Meta release.

The deepest deceit of the beta is the promise tucked inside it: that whatever goes wrong can be set right in the next release. Thus, the gig economy, which promises instant deliveries on an app that is still trying to become optimal. Meanwhile, the man delivering groceries in the rain is not work in progress; he is wet, poor, and as real as the rain. The app can wait to be fixed, but the humans depending on it are already hurting, with no fixes in sight. The damage already done to him cannot be taken back.

If you expand this idea, then you see that the grander the irreversibility, the thicker the language of provisionality laid over it. Nowhere is this clearer than in the great works pressed upon the land in the name of development.

Take the Great Nicobar mega-project. On an island in the Andaman Sea, the government means to raise a transshipment port, an airport, a power plant, and a township for some 3.5 lakh settlers on an island that today holds about 8,000 people. As many as 7.11 lakh trees are to be cut. Read the official defence and you are reading an app release note. The felling will be done in a phased manner. An Environmental Management Plan will mitigate the impact. An overarching committee will monitor and course-correct. The felled forest will be made good by another grown to order 2,000 kilometres away in Haryana. Even the project’s reason for existing has been fixed with a retrofitted patch: an official review in 2024 found the port had no defined strategic purpose; then the Defence Ministry fixed “national security” to the official review, like a feature added in the next build. It is the grammar of the beta—build now, monitor, improve—on a thing that cannot be iterated or repaired.

A rainforest is not software. The Shompen, who have kept their distance from the modern world for longer than the modern world has existed, do not live in the beta. In February 2024, 39 scholars of genocide wrote to the President of India to say the project could amount to a death sentence for the tribal community. There will be no later version of a people once they are gone. No committee can reverse the felling of a forest once it is gone.

And the island itself sits on a fault line—in 2004 its southern tip dropped some 14 feet into the sea in a single afternoon. A beta project built on a tectonic fault line. What’s the fix?

The Great Nicobar project is something utterly final. The shifted turtle reserve, the translocated coral banks, the afforestation—these are beta labels slapped on some very terminal outcomes. Labels used as though the consequences can be rolled back for the next release.

Frontline has been following the GNI story at every step as it unfolds—the rash of clearances, the science set aside casually, the tribes whose consent was faked. Read those stories here. They will make you see what this unfinished beta project will finally cost in terms of real-life damages.

Wishing you a week pondering over the betas in your life,

Jinoy Jose P.

Digital Editor, Frontline