Thing is, my whole life has been an inevitablism of people telling me to stop being obstructionist and get with the program. I’m now validated that people were projecting how they think someone bearing warnings should also bear intent.

Don’t look at how people treat LLMs for your signs of adequate social understanding, look at whether they talk to other PEOPLE like chatbots. Are there people who suddenly seem terse because they don’t think they need you (or anyone) on side or not?

Don’t just judge from afar, interact and find out. Whatever you think of this tech, if you think the future doesn’t involve people then ask yourself what business, what cybersecurity is even *for*.

I don’t care much whether you reach the same conclusions as me so much as you actually take responsibility for your interactions rather than blaming norms formed from misaligned incentives.

I'm working on a project to try to mitigate some of these risks and I affectionately call it the "loaded footguns" project.

All this risk that Google put paying customers through that frankly just boggles my mind. It's like they made it bad on purpose to drive you toward something but that other something just doesn't exist...

All I want is to be able to configure blocks and rules in an API. I would happily write the terraform provider myself.

Want to quarantine emails, block senders or create content compliance rules? Is there an API for that? No, go fuck yourself.

You have to click around in their *awful* inconsistent web interface, so obviously you can't do any configuration review.

We nearly had a fuckup the other day because the compound conditionals on the content compliance rules have to be selected from a drop-down and ANY looks a whole lot like AND and nobody could review the change before it went out.