# What if Cory Doctorow was Canada’s AI minister? > Source: > Published: 2026-08-13 12:00:00+00:00 What would Canada look like if tech journalist and author Cory Doctorow were AI minister? For starters, there would likely be less sloganeering and more cursing coming out of Canada’s innovation ministry. It might also look, however, like a Canada that’s more skeptical of big, southern tech firms, more comfortable outside the confines of US copyright law, and less AI-pilled than the government is today. Doctorow, who published his most recent nonfiction work, *The Reverse Centaur’s Guide to Life After AI, *earlier this summer, joined *BetaKit *editor-in-chief Douglas Soltys in the podcast studio last month, where he talked about repealing Canada’s anti-circumvention law Bill C-11, why jailbreaking big tech is a path to sovereignty, and how he fears the government’s “all-in” approach to AI could turbocharge 21st century fascism. You can read an excerpt from that conversation below. *The following interview has been edited for clarity and length.* **In your essay in BetaKit Most Ambitious, you talk about this idea that for Canada’s digital sovereignty to mean anything, we’ve got to jailbreak this country. Can you tell our audience what that means?** First, let’s talk about three ways that Canada is not digitally sovereign. We’re subject to raids on our privacy and on the functionality of the tools we use. We try to get Facebook to pay for news, now we just don’t have news on Facebook, and so on. Another is market opportunity. You could jailbreak the Facebook app. You could offer people all of the services that Facebook has, but without having to use Facebook. There’s plenty of Americans who would buy that if they could, and there’s plenty of people all over the world who would buy that if they could. The third way is that Microsoft could turn off M365 for any ministry or structurally important firm and brick that firm or branch of the government. So the question is why can’t we just fix this? The answer is that in 2012 we passed a wildly unpopular [law](https://www.parl.ca/LegisInfo/en/bill/41-1/c-11) that says it’s illegal for Canadians to modify American technology if it has digital rights management (DRM), and you can’t break DRM even for lawful purposes. **Apple won’t let you replace a screen or add more RAM unless you go Apple-certified and even then they might not let you. ** Jailbreaking is a way we defend our human rights. It’s the way that we capture those markets and turn their trillions into our billions. Jeff Bezos tells every supplier, “your margin is my opportunity.” When he does it, that’s not progress. And, if we do it to him, it’s not piracy. That margin is our opportunity. It’s also how we get our digital sovereignty because while we can build equivalents to M365, no one is going to copy and paste a million documents and data structures into your Canadian cloud. We need to do something like virtualize a phone and iterate through all the records and be able to extract the data that way which we can only do if we repeal this wildly unpopular law, become a dis-enshittification nation, turn their trillions into our billions, become digitally sovereign, protect the privacy and money of the people in Canada and set up an export market where everyone in the world, including Americans, become our customers rather than doing something like boycotting American products and drinking shitty Wayne Gretzky rye instead of delicious American bourbon, and whacking some poor farmer in a state that begins and ends with a vowel because we’re not buying his corn anymore. **By the time this episode comes out, BetaKit will have sat down with ** **Evan Solomon****, Canada’s AI minister. My question for you is: if you were AI minister, what would you do?** There’s a story about a system admin who’s told they have to build a server farm that can process a certain number of flops within 10 years. When asked what their plan is, they say, “I’m going to do nothing for the next nine years while Moore’s Law does its work, and then I’m going to build a data centre for 10 percent of what it would cost me to build it today”’ So, as minister I would say: I am not investing in GPUs, data centres, or AI, except to find efficiencies in open-source models and to do scientific research, until and unless the bubble bursts. I’m not buying [Aeron chairs](https://www.businessinsider.com/the-herman-miller-chairs-everyone-wanted-headed-for-landfills-2023-7) in March of 2020 because I’m not dumb. That would be my answer. **Our audience has seen Canada fail to commercialize all of the innovations it helped to R&D. I think we’re looking for a way to win for once. ** But what would winning mean? Winning the race to build the most money-losing technology in the history of the human race? That is not a contest anyone wants to win. If you want to capitalize on this, then improve the open-source models and find ways to do local compute so that they’re not dependent on these big firms.**Isn’t that Cohere? ** Sure. Sounds good. What you don’t want is vendor lock-in, and what you don’t want is to shovel asbestos in the walls by taking these chatbots where you’re buying them at one percent of the operating expense and then making them structurally important to your firm, with the understanding that these companies are not long for this world. That’s crazy. That’s not how you build a Canadian future; that’s how you take a bunch of Canadian businesses and expose them to risk. **So it’s even worse when the government buys into the inevitability fallacy?** Governments hate civil service in the same way that bosses hate employees. You show up and—especially if you’re Trump or Musk—you’re like, I would like to do a bunch of illegal things and everyone who works for the government says, “yeah, but it’s illegal.” You get rid of that civil service and replace them with chatbots, and you just say a thing and it happens. It allows you to live this solipsistic existence where you have an idea and it turns into policy and you don’t have to talk to experts who know how to do things you don’t know how to do. **The subtitle to your new book is How to think about artificial intelIigence—before it’s too late. What does “too late” look like?** The “too late” is when we fire a bunch of people who know how to do things; we vaporize their process knowledge, which we will spend decades trying to get back, and we replace them with chatbots who are bad at doing their job and cannot absorb their process knowledge because it can’t be integrated into training data. Then, the companies that run those chatbots go under, workers have scattered, and even if they do come back, they’re missing the other parts of distributed knowledge of how things work. And then, when 35 percent of the stock market is suddenly vaporized when the Magnificent Seven go away, we do what we’ve always done in this century when we have a market crisis: austerity, which we know drives people into the arms of fascism. That’s too late. *Watch the full conversation from *The BetaKit Podcast* here:*