There's constantly threads like this on the EVE forums.

One could argue that an improperly setup interface for ISBoxer can lead to mistakes as well but that is more of a mechanical setup issue and not a human mental error issue. Once perfectly setup, multi-boxing can still allow for human error, ISBoxer cannot.
How does ISBoxer not "allow for human error" exactly? It's not doing magic. It can't tell the client to perform a specific action the way bots can. Ask an EVE multiboxer if they've had any human errors...

That's his only argument and it's not very convincing. Nothing to see here...