Using Signal 2's AI performance as an excuse is pathetic

It's true that it's lacking at this point, but looking at the pace of development, it will be capable of reaching a level that feels seamless in two years.
There is already a significant time gap from version 1, and since it's already been delayed, people will be understanding even if it's delayed further. In fact, they might even gain some sympathy, so I wonder if it's really necessary to take such a reckless risk.