Amazon went 14 years before posting a profit. Not that I think Twitter was necessarily investing revenue correctly, but a business can go pretty far even in the red.
Because Amazon was spending on R and D. AWS came as a result of that. Amazon's core business was known to overspend(still do) just for customer's satisfaction, and it was very clear Amazon could turn into profit any day they want. Same is not the case for Twitter.
Difference is Amazon spent a lot of it on durable infrastructure (datacenters, servers, fulfillment centers).
I don't know how Twitter spent its money, but seemingly on undifferentiated software (i.e. not durable) or moderation? Both of which are recurring expenses.
The difference is that Amazon made plenty of cash, and chose to reinvest all of it. For nearly all of those 14 years they could have just decided to be profitable, if it had been advantageous to do so.
I don't think Twitter is in the same position at all