Several posts are saying this would be a good thing. Maybe, if implemented like the meme shows, but I worry ads in LLM chatbot output have the potential to be the most insidious and persuasive form of advertising we’ve ever seen. Chatbots are – by their nature – conversational and pushed in forms that are meant to eliminate any sort of thinking on the part of the user. And more and more, people are using them as their closest, most trusted confidants. Who’s to say a (e.g.) political candidate won’t pay to control the narrative of a chatbot, and have their propaganda (“ads”) appear as a private conversation between an individual and their trusted “friend” or “partner”?

To add to this, the Phillips screw was designed to be self centering for use with motorized screw drivers on assembly lines, with an added benefit of easy cam out that prevented over tightening by said screw drivers. All that is kinda obsolete now and actually disadvantageous, so Phillips is probably the most entrenched and enduring corporate propaganda screw.

Although it may not be ideal for a large document, I have used the Coblis for individual figures that I’m worried about. I usually find small tweaks to the colors can make them much more readily differentiable in the simulator.