Toronto in 2030 should ___
This is a trial run, and it is small on purpose. For nine days — Saturday 15 August to Sunday 23 August 2026 — anyone can say what they think Toronto in 2030 should be, agree or disagree with what other people said, and add statements of their own. At the end, we publish what the city agreed and disagreed about, and we publish how the thing was built.
How it works, in four lines
- You read a statement — one sentence about the city — and press agree, disagree, or pass. Then the next one.
- You can add your own statements. This is the part that matters. The starting set is only a floor.
- Everything anyone writes is read by a person before it appears. Nothing is published automatically. The rules that person follows are published here — all five reasons a statement can be rejected, and what to do if you think we got it wrong.
- Nobody debates anybody. There are no replies and no comment threads. You vote alone, on your own screen, which is the whole reason this produces a map instead of an argument.
The conversation
Where the starting statements came from
We wrote them. All thirty-four. Every starting statement was written by the people behind uniteTOlove from material already published on this site — nothing from private notes, nothing from unpublished drafts. We deliberately included statements we expect people to disagree with, including some we do not ourselves believe, because a starting set everyone agrees with produces a fake consensus and teaches nobody anything. We checked the set for a political lean, found one, and rewrote five statements and added four to correct it — which we describe rather than quietly fixing.
The full account — which published page each statement was composed from, why there are no participants' words in the starting set yet, and what we will and will not claim about the results — is here: where the starting statements came from.
Two ways in, counted separately
Some people were sent a personal link — through a conversation kit room, or because they had already signed up. If you got one, use it rather than this page's plain address. Everyone else uses this page as it is, which is open to anyone.
What the personal link actually does, precisely. It carries a random code. Nothing else — not your name, not your email, not anything derived from them. The code lets us tell one person's votes apart from another's, so a hundred votes from a hundred people can be distinguished from a hundred votes from one determined person. We keep no record of who was given which code, which means that even we cannot work backwards from a code to a person. That is a design choice, and it is the reason we can offer this lane at all.
What it is not. It is not identity verification. We have not checked anyone's ID, address, or eligibility, and we will not describe this lane as "verified" in the results. All it establishes is that the person was invited through a room or a list rather than finding the open link.
These are two separate lanes and we report them as two separate lines. We will never quietly merge them and present the total as though everyone had been invited. If the open lane shows signs of being gamed, we will say so in the results rather than deleting the awkward numbers.
What this is honest about
- It is small. We expect somewhere between a few dozen and a few hundred people, drawn from conversation kit rooms and people who have already signed up. A thin room will be reported as a thin room. This pilot is testing whether the machinery works, not making a claim about what Toronto thinks.
- The tool is called Polis. It is free, open-source, and has been used by governments and civic groups around the world — Taiwan's government used it at national scale, and an American county ran a version of this with about 8,000 residents in 2025. We are naming it plainly because you are about to use it, and hiding what you are about to use would be the dishonest option. It is run by its makers, not by us; your votes sit on their servers, not ours, and we do not receive your name or email through it because it never asks for them.
- We will not claim Toronto has N camps. Software like this draws groups of people who voted similarly. Those groupings are an illustration — run the same data again and they can come out differently. What does not change is how many people agreed with each statement, so that is what we will publish as the finding. We set that rule before opening the conversation, not after seeing what it said.
- This page collects nothing. No sign-up form, no email capture, no counter, no analytics, and no third-party trackers of any kind on this page — unusually for this site, we left our own visit counting off here too, because a page where you are about to say what you think should not be measuring you. The only script on the page reads the random code out of your address bar and hands it to the conversation; it stores nothing and sends nothing anywhere else. Whatever the conversation itself collects is described in its own privacy terms, on its own site.
- Nobody is running this on anyone's behalf. No organization convenes this conversation. It is an instrument, published for whoever wants to use it.
What gets published afterwards
Within roughly a week of the conversation closing, we publish:
- The agreement rates — how many people agreed with each statement, including within groups that voted differently from each other. Statements that divided people get published as prominently as statements that united them.
- The moderation log — how many statements were published, how many were rejected, and for which of the five reasons. Including any rejections we later reversed.
- How many people took part, in each lane, separately — and the days nobody cleared the queue, if there were any.
- What went wrong. This is a trial run. Its most useful output is the list of things that did not work.
Want to run a conversation of your own, on paper, in a room? → the kits. Want to send in a vision, a complaint, or a nomination outside this pilot? → the harvest.