The ubiquity of “AI understood as something that can do your job” is belied by the name itself. There is literally no reason to refer to LLMs or diffusion model

The ubiquity of “AI understood as something that can do your job” is belied by the name itself. There is literally no reason to refer to LLMs or diffusion models as Artificial Intelligence other than to convince people of their capacity to… do your job. This entire hype cycle is based around that concept. Billions of dollars have been invested in that idea. Every day consumers and investors are pitched an Artificial Intelligence that can do your job, and in order to keep the money flowing the purveyors of this software blur the line between today’s capabilities and some vague greater future capability. If your contention is “my friends and I are completely immune to marketing and this means that everyone else is too” then I’ve got kind of bad news for you about both the world writ large and your particular corner of it. I am not sure if you saw any of the several very public examples of people using these tools in exactly the way they’ve been pitched to, so I’ll link to my post above: https://news.ycombinator.com/item?id=41440126 https://news.ycombinator.com/item?id=41440126 And again I still have no idea what you’re talking about re: agents and the publishing industry in general. The industry was able to work with a signal-to-noise ratio of [x] pre-AI and that means it should work with a ratio of ∞x because agents exist? This is like saying that because the human body has a liver and has processed 2-3 units of alcohol per hour for all of history that humans can survive on a diet of 1,000 units per hour. I don’t mean to sound rude but “the influx of garbage has no impact” reasoning is indistinguishable from gibberish to me. It may as well be a cat typing out an equal number of letters by walking across a keyboard in terms of how it makes sense in the real world.