Everyone is blinded by the large number like $3 billion. That's not the number that matterS. All the companies in this market are subsidizing rides (except maybe the traditional taxis). What matters is how much each ride is being subsidized.
A while back another user did this math using leaked financials from Uber and Lyft and Lyft deserves more criticism for subsidizing rides than Uber:
It does change it a bit. If one is more efficient than the others and is improving efficiency/utilization faster than others, then eventually that company will eventually be able to compete on price while others are forced to subsidize. At that point they can just wait out the other competitor until the investors in those competitors tire of throwing good money after bad.
Subsidies don't work at scale. The bigger the ridesharing market gets, the less a company can meaningfully provide a subsidy that changes consumer behavior. Whichever company can get to breakeven before R&D and capital investments are taken into account is going to clean up because at that point they can outlast their competitors indefinitely.
So, wait it out, and then raise prices? Assuming that won't affect demand I suppose. Give me a mountain of cash and I'll start a flat rate $25 airline and put everyone under. Then I'll raise prices when it's over. Guess what happens when the mountain of cash is gone and I have to charge something that makes margin? We're back to competing because my moat was artificial. Nobody has demonstrated a profitable quarter yet in this nuevo taxi game.
A while back another user did this math using leaked financials from Uber and Lyft and Lyft deserves more criticism for subsidizing rides than Uber:
https://news.ycombinator.com/item?id=13772168