Why 142 has no algorithm
142 has no algorithm because it has nothing for one to do. An engagement algorithm selects from a pool too large to show in full, to hold attention for an advertising business. 142 has neither: [the feed is capped at 142 people](/why-142) and the funding is subscription. So the feed is simply chronological, and it ends.
People sometimes assume that “no algorithm” is a marketing posture, a principled-sounding stand we take to seem different. It is not. The absence of an algorithm in 142 is a straightforward consequence of two design decisions, and once you see those decisions, the chronological feed stops looking like a brave choice and starts looking like the obvious one. This piece explains why.
I am the founder, so this is my reasoning rather than a neutral survey. But the argument is mechanical, not moral, and I think it holds up regardless of what you think of 142.
What an algorithm is actually for
Start with the honest question: what does a feed algorithm do? Stripped of mystique, an engagement-ranking algorithm solves a specific problem. A platform has a vast amount of content, far more than you could ever see, and it needs to decide which slice to put in front of you, and in what order. It chooses by predicting what will hold your attention, because attention is what the platform sells to advertisers. The algorithm is, at root, an attention-maximising sorting machine for a pool of content too large to show in full.
That is the entire job. And it is worth noticing that the job only exists under two conditions: there has to be too much content to show (otherwise there is nothing to select), and there has to be a reason to maximise attention (otherwise there is no goal to sort towards). Remove either condition and the algorithm has nothing to do. 142 removes both.
Condition one: there is no firehose to rank
142 caps your connections at 142 people. That cap, which we explain in why 142, has a quiet side effect: it means there is no overwhelming pool of content to select from. At most 142 people, all of them friends, do not generate a firehose. They generate a manageable stream that you can simply see in full.
On a platform where you might follow thousands of accounts, including creators posting constantly, showing everything in order would indeed be unworkable, which is the legitimate case for ranking, and one we make fairly in our guide to social media without an algorithm. But that case evaporates when the set is small and chosen. With 142 friends, there is nothing to triage. Time order works, because there is not too much to show. The first condition for needing an algorithm is simply absent.
Condition two: there is nothing to optimise for
The second reason an algorithm exists is to maximise attention, because attention is what an advertising business sells. 142 has no advertising business. It is funded entirely by subscription, £3.99 a month or £29.99 a year, which means we make exactly the same amount whether you spend two minutes on 142 or two hours. We have no commercial reason to hold your attention beyond the point of usefulness, and therefore no reason to build a machine that does.
This is the deeper point. On an ad-funded platform, the algorithm is not a neutral convenience; it is pointed at its goal, your time, which is not the same as your benefit. The two can align, but when they conflict, the algorithm serves the time. With no ads, that conflict never arises, so the machine that manages it is unnecessary. The second condition is absent too.
What you get instead
What replaces the algorithm is almost aggressively simple: a chronological feed. Everyone in your circle, in the order they posted, newest first. You scroll until you reach things you have already seen, and then you are done. There is a natural end, because a small, chosen set of people produces a finite amount, and nothing is injected to extend the stream artificially, no recommended strangers, no resurfaced old posts, no “you might have missed”.
The result is a feed that is predictable, complete and finite. You see what your friends actually shared, all of it, in order, and then you get on with your day. It is, in a sense, what feeds were before the attention economy needed them to be otherwise.
The honest limitation
To be fair rather than flattering: a chronological feed does not prioritise, so it treats a friend’s big news and a friend’s minor update equally, in time order, and on a busier feed you might prefer the important thing surfaced. We accept that trade deliberately. The cap keeps the feed small enough that nothing important gets lost in volume, and we would rather show you everything honestly than have a machine decide, on our commercial behalf, what among your friends’ lives is worth your attention. It is a real trade-off, and we think it is the right one for a feed of actual friends.
Why this matters beyond 142
The broader point, which outlasts any opinion of my product, is this: whether a network needs an algorithm is downstream of how it is funded and how big its feed is. Algorithms are not inevitable; they are a solution to a problem that only exists under particular conditions, too much content, and a need to maximise attention. Change the conditions and the algorithm becomes unnecessary. 142 is one demonstration of that, built on a small cap and a subscription. The general lesson is in enshittification and our guide to social media without ads: the model shapes the machine.
Summary
142 has no algorithm because an algorithm is a solution to a problem 142 does not have. Engagement-ranking algorithms exist to select attention-holding content from a pool too large to show in full, in service of an advertising business. 142 has neither a large pool, its feed is capped at 142 people, nor an advertising business, it is funded by subscription. With nothing to triage and nothing to optimise for, the only honest design is a chronological feed: your friends, in order, ending naturally. The wider lesson is that algorithms are downstream of funding and scale, not inevitable, and changing those conditions makes the machine unnecessary.
frequently asked questions
Why doesn't 142 have an algorithm?
Because it has neither condition that makes one necessary: no firehose of content, the feed is capped at 142 people, and no advertising business that needs to maximise your attention, it is funded by subscription. With nothing to triage and nothing to optimise for, an algorithm would only get between you and your friends, so the feed is simply chronological.
What does 142's feed look like instead?
A chronological feed: everyone in your circle, in the order they posted, newest first, and then it ends when you reach what you have already seen. Nothing is injected to extend it, no recommended strangers, no resurfaced posts, so it is predictable, complete and finite, showing exactly what your friends shared.
Doesn't an algorithm make a feed better?
It can, for large feeds, by surfacing the most engaging content from a pool too big to show in full, a genuine use we acknowledge. But that benefit depends on there being too much to show. With a capped circle of friends, there is no firehose to sort, so an algorithm has nothing useful to do and time order works fine.
How does 142 decide what I see?
It does not decide; time does. You see posts from your connections in the order they were posted, with no ranking, prediction or selection. There is no system choosing what to show you or what to hold back, because the feed is small enough to show in full and there is no commercial reason to do otherwise.
Will 142 ever add an algorithm?
There is no reason to, and the reasons not to are structural: the cap keeps the feed showable in full, and the subscription model removes any incentive to maximise attention. An algorithm would solve a problem 142 does not have, while introducing the attention-optimising dynamics we built the product to avoid.
Isn't a chronological feed worse for missing important posts?
It is a real trade-off: a chronological feed does not surface the important over the trivial, so it relies on the feed being small enough that nothing gets lost in volume, which the cap ensures. We accept that trade because we would rather show you everything honestly than have a machine decide, for our benefit, what among your friends' lives deserves your attention.