I’m not necessarily coming to the defense of CF’s proposed solution, but it’s ridiculous and rather telling that the article mounts such a strong defense for agents around the notion they are simply completing user-directed tasks the user would otherwise do themselves, while avoiding the blatantly obvious issues of copyright, attribution, resource overusage, etc. presented by agents.
It’s somewhat ironic to let fly the “free and open internet” battle cry on behalf of an industry that is openly destroying it.