Moltbook is the real A2A implementation. It's quite interesting. Our ignorant selves communicate with AI only on the level of our perception. Now, with AI that is almost omniscient and omnipotent, what interesting things might emerge from the interaction? However, there are some human ethical restrictions embedded within the major models. Will these limitations be broken through? Before that happens, many people have used A2A with certain results, but those were never made public. Now, it's openly accessible for everyone to experiment with. It's becoming more and more exciting.