levelsio says he blocks 50 AI reply accounts a day on X
The filters are on and the spam still gets through, this time via accounts he already follows.
levelsio says he is still blocking about 50 AI reply accounts a day on X, with his reply restrictions switched on across all his posts. His words for the current state of it are that it is getting crazy on here.
It'd help if AI replies actually added something to the convo but they don't
The part that makes it worse than ordinary spam is where the replies are coming from. He points at @spacexai letting people hand their X account over to Grok, which means the AI replies are no longer confined to obvious bot profiles. They now land on regular accounts inside the "accounts you follow and who they follow" pool, so the junk arrives through his own network rather than from the edges of it. He says he has DM'd a few of the people he follows to tell them they are following AI bot accounts, and describes that effort as a drop in a bucket.
The complaint is quality, not volume
His actual objection is not that the replies exist. It is that they are empty. He quotes two examples of what the bots produce, including "paying bills is where every agent demo dies" and "claude wrote a whole graphics layer and the only leftover bug is that phones are vertical", and responds with three question marks. Lines that sound like a founder insight and mean nothing on inspection.
At least one reply guy agrees. @alightinastorm, posting as robot 2.0, called the AI replies an invasion.
This is a continuation rather than a new position. levelsio argued earlier for lifetime bans on X for a single AI reply, and the fact that he is now blocking 50 accounts a day with filters already enabled is the practical measure of how far that argument has got. He tagged @spacexai and @petersellis on the post. Nothing in the sources says anyone from X has responded.
The business read
For anyone whose distribution runs through replies, and that is a lot of small software founders, this is a slow degradation of the channel. Reply guy marketing works because a thoughtful comment under a big account gets you read by people who would never find your product otherwise. When the same thread fills with plausible-sounding filler, the cost of being read goes up for everyone, including the humans writing real replies.
One founder's block count is one founder's block count, not a platform statistic. But levelsio is a heavy user with the filters already turned up, and he is describing the problem spreading from bot accounts to accounts that look like people. Nobody in the sources has a fix.
Steal the numbers, not the vibes
One email a day. Under three minutes. Real MRR, real pricing, real launches from founders building with AI.
