The speaker discusses SpaceX's plan to start launching Starlink payloads with Starship once they achieve sufficient confidence in its reliability. They suggest that even if "one or two rockets don't quite make it" leading to "hundreds of millions of dollars" in losses, it would be considered "no big deal" because Starlink payloads "pay for themselves almost instantly," allowing them to "just launch another one."
The speaker believes it won't take "too long" for SpaceX to gain this confidence and anticipates they are "probably not going to lose the payload." Beyond Starlink, the next step would be to "start launching some of these orbital data centers" and "figure it out" iteratively. Ultimately, the pace of these operations will depend on "how quickly they can produce Starships" and "how frequently each Starship can fly," including the efficiency of "refurbishment in between flights."