With regards to talent, there's no particular reason why software centers couldn't be in any major established city in the world. It's not like it takes billions of dollars on a highly uncertain bet like creating a car company, rocket reuse company, or a CPU company.
A small crew of people could potentially build the next WhatsApp. On Erlang.
There are definitely many good programmers all over the world, but there are more in the US, because that's where all the best companies are. So if you're trying to make a good company and you want good programmers, where do you go?
In the case of AI it absolutely does take billions and billions of dollars on an uncertain bet. They bet that throwing more data, more hardware, more GPU cycles at the problem would yield results and it has.
That statement applies to most industries. Tons of areas have the potential for an industry boom, but silicon Valley in is California. Or semi conductors are in Taiwan.
For many reasons, only some areas succeed whilst the rest fail. In this case, Canada doesn't have silicon valley, nor do they have a high amount of start ups.
A small crew of people could potentially build the next WhatsApp. On Erlang.