We have written a review of how Polkadot is reorganising itself in 2026, while 1954 (the DOT DAO Strategy Mandate) and 1955 (PCF as onboarding operator) are both deciding.
It does not argue for or against either vote and makes no proposals. It is mainly a record for ourselves, of how we understand the change, so we can weigh our own work at Birdbrain against it.
We are sharing it because the register may be useful to others. It tries to show the shape being voted on, which was assembled across thirty referenda over two and a half years and is hard to see from inside any one of them.
Full paper, with every figure and a register of the referenda: publications/content/articles/polkadot-changing-shape/index.md at 5fef3d0d67f3dfb1b4cb27f0af26e3d212f1f0ed · Birdbrain-wtf/publications · GitHub
TL;DR.
- The direction in 1954 counts people before total value locked, pays individuals rather than agencies, and leaves privacy trade-offs to the user. Much of that matches what we have argued for in our own work.
- The bigger change is underneath it. The PCF was founded (730) as “an unopinionated entity with no agenda, roadmap, or business purpose except to take actions directed to it via OpenGov referenda”. It now operates the Polkadot App, proposes strategy and stablecoin referenda, holds the key that would attest new users onto the People chain, and admits privately funded units. Each step had some authorisation. Together they are a change of role no single referendum asked voters to approve.
- The widest step rode on the smallest budget. 1900 asked for $1,500 for a website, and is now called “the PCF 2.0 mandate”. One account voting 5M DOT at six-times conviction made up about 99% of its aye side, and abstentions supplied 93% of its turnout.
- The checks in the founding design were never switched on. Both DOT-holder board seats have been empty since 2024 after two failed elections (1737, 1749), and we found no published transparency report.
- Units, in plain terms: a team whose budget never goes through the treasury and never faces a referendum works under the PCF’s name, admitted by the directors alone. The founding design made treasury funding the DAO’s main lever over the PCF. A unit’s money comes from elsewhere, so that lever does not reach it. The one unit so far, the Human Web Initiative, “brings a renewed initiative with a clear direction for the PCF”. Its funder is not named, and a question on this forum about whether Polkadot App work sits inside it is unanswered.
- Per-person incentives make verification the payroll. Project Individuality is designed so that no entity has discretion over disbursement (1783). The app’s user terms let the operator add “any other ‘Proof of’ verification system”, and 1955’s lite person shows an attested app install, “not that it belongs to a unique human”. The mandate does not say which one its per-person limit counts.
- dotUSD’s primacy rests on phase two. Today it mints one for one against USDT, the kind of third-party-backed coin the mandate sets it against.
- The direction is being set by very few votes. 1954 had 419,551 DOT of support on 19 October, about 0.03% of issued DOT.
Why we wrote it. We run small versions of the same machinery on Kusama: per-person incentives, passkey joining and a dollar unit on Asset Hub. Section 8 of the paper sets each of Polkadot’s pressure points against its counterpart in our own design, including the operator key we hold and the gap between a passkey and a unique human. That is the part we will keep coming back to.
Disclosures. Birdbrain is a collective on Kreivo. Decent Partners, which supports us, has a commercial interest in dUSD, a custodial stablecoin, and is an investor in Bloque.
Appendix B of the paper lists six questions we could not answer, and we would be glad of corrections to any figure.