Cohere co-founders Aidan Gomez and Nick Frosst joke on social media about the reliability of Canadian LLMs
Story Overview
Cohere co-founders joined a quick round of Canadian-themed humor after Wes Bos posted that Canadian LLMs would never do 'this' to you. Nick Frosst quote-tweeted a simple 'We won’t,' while Aidan Gomez added his own light jab about models from north of the border steering clear of whatever behavior was implied. The exchange stayed short, playful, and free of any technical claims.
The reference stays vague
Neither founder spelled out what 'this' actually meant, leaving the joke open to whatever controversy readers had in mind.
No benchmarks attached
The thread contains zero test results, model comparisons, or performance data—just the same casual social-media tone that started it.
Positive users praise Cohere for frontier AI work and open-source releases while negative users dismiss Canadian LLMs as nonexistent jokes due to brain drain and lagging tech.
Most Activity
@wesbos Yes, because by the time they reach that level of competence, open-weight models have already caught up.
@wesbos they exist?
@wesbos "You betcha, bud — you’re absolutely right, eh. That’s a great question. It’s a bit more nuanced than it first appears, so let’s break ‘er down over a double-double, because both things can be true. No worries, we’ll get ’er sorted. Sorry."
@wesbos Hard to get banned from a model that does not actually exist
@wesbos Canadians would destroy it themselves first
@nickfrosst I like you guys and I hope you succeed (thanks for releasing open source!!), but AFAICT the main reason this is probably true is that there won’t be a frontier model coming out of Canada :/
@wesbos they would instead freeze everyone's bank account.
@wesbos Gotta exist first before you can even think about banning.
@wesbos Yes. @cohere is one of the top companies in the world developing frontier level AI models both open source and proprietary and its foundation is sovereignty. Based in Toronto. Look it up.
Canadian conservative hate their country.
@wesbos Where do they get their data to train the models? Distillation?
@wesbos
@wesbos Canadian LLM is an oxymoron. 😂
@wesbos only id they’re opensource. otherwise ANYONE can pull shit like that. country/company doesn’t matter.
@wesbos Me after asking my local LLM to fix my code: "bro just one more error and I uninstall everything"
@wesbos Ya but they’d only work at half the capacity 😂
@wesbos Canadian LLM feels hurt
@wesbos Canadian LLMs would never do anything to you.
Because they don't exist.
@wesbos Do what?
@wesbos Time roll one