What's going on Canada?
🍁 Meta
🗺️ Provinces / Territories
🏙️ Cities / Local Communities
Sorted alphabetically by city name.
🏒 Sports
Baseball
Basketball
Curling
Hockey
Soccer
💻 Schools / Universities
Sorted by province, then by total full-time enrolment.
💵 Finance, Shopping, Sales
🗣️ Politics
🍁 Social / Culture
🌐 Other Relevant Communities
Rules
Reminder that the rules for lemmy.ca also apply here. See the sidebar on the homepage: lemmy.ca
It's important for citizens to understand the relative strengths and weaknesses of the Canadian and US economy before they decide that the country needs a 'deal'. I've brought together a bunch of background info that I haven't heard much of from the legacy media. I think it suggests Canada is doing a whole lot better than America right now. https://billhulet.substack.com/p/my-hot-take-on-the-current-trade
I think your read on AI investment is quite far off. We are in the midst of a new industrial revolution, and the two players leading it are the US and China. Canada is basically on par with the US as a leading country in terms of uptake of AI and usage by consumers, but Canadians aren't using Canadian AI. How many people are using Cohere on a daily basis as compared to OpenAI or Anthropic? How many unsolved math equations are being solved by Cohere? There won't be any left for Cohere to solve because they're getting solved at the frontier and we are not there. This gap is not shrinking. Cohere's top model is on par with Haiku, not Fable. It's so far down it doesn't even appear in most rankings unless you look at very specific use cases, like having low hallucination rates on RAG. So, sure, we have exports like aluminum, O&G, REEs, and electricity that the US needs, but the frontier of innovation where all the new economic value is being generated is down south and in China, and a huge amount of the AI investment in Canada on AI data centres is with US capital and US tech. So, our relationship is not changing much in that regard. We have long been an exporter of inputs to higher value processes, and we look to be continuing as such while the US continues to drive the frontier of value creation and US corporates own the IP and the infrastructure.
And, btw, Chinese AI labs are not just focused on niche applications and there is massive investment in China towards building frontier chips and infrastructure, not to mention physical AI, and we should be glad of that. If not for Chinese labs pumping out near-frontier open models, the US labs could basically lock up the market right now. As Chinese chip fabs improve, their capacity to compete on chips will do the same in making hardware markets more open and competitive, and we should be very glad of that, too. We already face enormous pressure to just be a supplier of inputs to a US-controlled tech revolution. We are really the ones who are trying to compete on niches, like some quantum and narrow application AI models, while predominantly feeding inputs into the general purpose tech development driving demand.
If Canada invests in AI intelligently, then the few investments we make will pay off significantly. It is all about picking the winning horse, as compared to the American approach which is to bet on ALL of the horses, hoping that you have bet on the one that wins. That produces a LOT of losers, only one winner. The dot.com bubble all over again.
The winning horse, I'm sad to say, is not coming from Canada. Unless Rich's new lab disproves his bitter lesson, we are not in the race. Cohere is valued at $7B. Anthropic is expected to go public at $2T. Cohere can make a lot of money by focusing on strengths they have that are practical business applications, but there's a reason Anthropic's valuation is coming in at 285× Cohere. And, we don't have much else while the US has OpenAI, xAI, Google DeepMind, Meta and more, not to mention owning the hardware layer with Nvidia. It's not even close. Even if a lab taking a divergent approach achieves an incredible breakthrough, where do you think the capital to scale will come from? We don't have the depth of capital markets domestically to compete at the top end of this. It is a two horse race because there are only two horses with the combined frontier tech and capital to run the race.
And, wrt to the dotcom bubble, it popping didn’t stop the internet from being revolutionary and it didn’t stop the US tech from dominating the internet. We are very likely heading the same direction, but with China as a competitor this time and a more fragmented global market.
We don't need the killer app. we just need one that holds its own. $7B is still a lot pf capitalization. If we grow that, it will suffice for the Canadian market. But apart from making it big in the investment market (like Red Hat Linux did in the dot.com era) your AI megaliths do not show a lot of financialization. They don''t seem to be producing a solid return on the investment dollar. For all its hype, very few have demonstrated how to actually make money by using it. Great for aggregating research, but how do you make money from that research? Facebook makes their money by paid advertising and paid promotions. How do you translate that model into AI? For META, Google, and Amazon, it is just a different way to do what they are already doing.
This isn't really a matter of killer apps. It's a general purpose technology. And, to be clear, Cohere is a good business and they're doing well focusing on enterprise. I'm not knocking them. Still, they're on a similar timeline to profitability as Anthropic, at wildly different scale. Cohere was around $250M ARR last year. Anthropic is on track for something like $60B ARR this year and $140B ARR next year. The idea that profitability isn't there is just wrong. And that's just the model providers, but if you look at Cohere's business model that has enterprises invest in sourcing their own instead of getting inference from Cohere, who are they getting the compute from? They're either buying GPUs/TPUs made by a US tech giant, or they're renting compute from a US tech giant, or if they're actually renting compute from a Canadian tech giant then that tech giant is buying GPUs/TPUSls from a US tech giant. No matter how you slice it, US tech is winning.
By that logic, the end benefit goes to Taiwan, because that is where Nvidia outsources their chips to.
Of course, TSMC is also doing incredibly well out of manufacturing, but not as well as Nvidia which still has like 75% margins on their GPUs and nets like 50% on them. Not sure what Google's margins are on their TPUs, but I'm sure they're also high and they are making huge returns on serving inference.
As to the Canadian compound semiconductor fab, it seems great and may be very worthwhile, but again it's pretty small potatoes in the industry:
Nvidia's profits last year were 10× the upper bound of what the potential Canadian fab is promoting as a possible 10-year gross impact on GDP.
The industry is a point where capital demands and returns are so extreme that GPU access is having to be sold like bonds to the biggest asset managers on the planet just to feed the CAPEX to meet demand, and that's a big part of why the frontier is a two-horse race that we're not in. Again, that's not to say Canadian businesses can't make money during all this, but the players that are going to dominate this space are not coming from Canada.
An old adage in the industry says that the first GPU out of the fab plant costs $1 billion dollars but the nest one costs $5. The actual making of the chip itself, since so many of them can be made from the same die, is not that costly. The total cost of designing and developing the GPU is enormous and a significant chunk of the profit from each GPU sold goes to developing the net model. Although Nvidia's balance sheet looks good, it is only as good as the next iteration.
I think it's important to understand what happened during the Industrial revolution. It involved a lot of booms, busts, and a tremendous amount of wasted investment. In Ontario, where I live, there are pieces of extremely expensive to build, but ultimately worthless infrastructure projects all over the place from this time. There was a tremendous over-build of railways and also some canals, for example. In Guelph Ontario, for example, the city owns a railway that it took over when it's parent went bankrupt and stranded local industry that depended on it. And in Perth Ontario, there's a dead canal that goes through the downtown and is a remnant of another failed infrastructure project. These are the 19th century equivalents of LLM data centers.
It's not a question of AI being a bubble or a new industrial revolution. It can be both. And governments have a responsibility to navigate this stuff to minimize risk to the ordinary citizens. Instead, it looks to me like the US is terrified by Fear Of Missing Out (FOMO), so it's putting the deed to the farm on the table. This is a profound misallocation of the nation's resources.
Follies and failures are a normal part of any period of explosive change and growth. That's the nature of markets. What matters long-term is who comes out on top in the end. Whoever comes out on top in an industrial revolution also comes out with enormous economic and military leverage over anyone who lagged behind. Now, we can take a cautious approach where we try to never set a foot wrong by regulating everything to a T (and it would have to be done via massive regulatory restriction, because this is a market-led revolution and not a state-led revolution), but that won't make us win in the long-run. So, if you want to think of what happened during the industrial revolution, consider what happened to countries that fell behind during the industrial revolution. That is a big part of what's missing in your analysis of how Canada is positioned relative to the US. The US is in a period of instability as they revamp their whole industrial and financial policy approach, but US tech is positioned to be winners of this industrial revolution and to have massive leverage over Canada in the long-term because of it.
To be clear, I don't like this and if it were up to me we would slow all AI development down until we could provide the power supply from renewables while making social and political adjustments to ensure the benefits could be widely distributed. But, it's not up to me and I can only call it as I see it.
The uptake (successfully getting value put of llms) is probably the bigger opportunity than the models themselves at this point - as model performance gains plateau and converge, the possible margin per token may get close to nothing.
Canada punches above its weight on a research pov, but misses on capitalization. That said I think US market is going to take a massive hit from over investment. Related sometimes Canada's caution is a good thing! (See financial crisis).
On the other hand... The opportunity for the frontier labs is less on research / capital in my opinion, but whether network effect becomes a material advantage.
I agree the value collapsing to a layer at which we can compete is probably our best hope of staying competitive, but counting on network effects is not something that leans in our favour. The US labs and the existing US tech platform giants already dominate on network effects. And, on research, we have traditionally punched above our weight but research itself is increasingly going to be dependent on frontier model access, which means we're either looking to big Chinese models or big US models.
Even if the US market takes a big hit and we fare better due to prudence, I'm not sure what edge that ultimately gives us. We still wouldn't be the ones who own the rails on which recovery and growth would be built, and just like we saw the US recover to dramatic growth faster after 2008, we would likely see the same again as capital keeps chasing the frontier and betting on big platforms that already own the infrastructure layers.