Conceptually, this is correct, but you must account for the fact that some problems are so hard to explain in English and have so many restrictions and strict rules that you end up writing a much longer and much more brittle program.
In the end, you need to explain the real world problem as without it, the AI does not know what to do, however smart it is.
This is a bad take. Large companies have often team collaborating across the globe so they're used to partial remote already. My company (Microsoft) turned mostly remote, teams that started through COVID have never seen office. Also the company doubled through COVID so a large portion of the company is remote first.
Also.. Offices cost a lot whether you're big co or a start-up.
It is amazing for typing out mock data. Say you're testing parsing of XML - it can easily suggest the the assertions over the data parsed from the XML.
Example test that was 95% coming out of Copilot:
https://github.com/dotnet/arcade-services/blob/61babf31dc63c...
It also predicts comments and logging messages amazingly well (you type "logger." add 7/10 times get what you want, sometimes even better), incorporating variables from the context around. This speeds up the tedious parts of programming when you are finalizing the code (adding docs + tracing).
Honestly, Copilot saves me so much time every week while turning chores into a really fun time.
I didn't appreciate developing Azure Pipelines in YAML so I've created a library that lets you use C# instead.
Aside the apparent advantages of the strong typed environment, I was able to bake in many more features that make your life easier. Code reuse is also super easy.