Die KI-Gegenreaktion kommt. Die Kanadier reden nicht annähernd genug darüber.

I’m a midlevel programmer at a brand name tech company, I work with ai everyday. Claude opus 4.6 can do my job about 80% as well as I can. It’s one of the most expensive ai models but it’s qualitatively different from other ais which produce very uneven work and spin out into goose chases if you take your eye off them.

Using it means all the drudgery is gone from job and the bottleneck for shipping new things is coordinating with other people and having enough ideas for things to ship. The problem is, you hire interns and juniors to teach them and to give them shit work to alleviate the drudgery of midlevel and senior programmers.

If I had a junior or an intern on my team I have no idea what I would have them do. There is nothing the early-career version of me could do that I couldn’t quickly do with ai now. That saves the company money but it cuts off the supply of new talent.

No company has figured out the solution here, but it definitely isn’t cutting off or over-regulating AI. My company wouldn’t be competitive at all if it couldn’t use AI in Canada to reduce headcount since all its international competitors are currently doing that.

We’d be in a much poorer world if the Luddites had held back mechanization. It’ll be the same if the anti-AI people win. We should have had smart government policy back then to ease the transition and we should definitely make sure we do it this time.