I've been talking about learning to be deliberate in how you communicate with AI instead of "splatterprompting" for a while now. I'm also a fan of the iterative process: Ask for a dragon, see what AI returns by default, compare it to your vision, spot what's missing or different. Ask for a new dragon with those things corrected. Rinse, repeat.
But I think putting things into words is just as relevant, perhaps counterintuitively, when it comes to simply writing for an audience. Because yes, AI can spit out passable prose at scale in seconds, but having a unique voice is only possible when you're not an "average common denominator aggregator." I really think we'll see more people gravitating away from midstream AI slop towards people they can relate to.
So yes: Learn to describe shit for AI, but also for other humans.
Since it apparently has never occurred to anybody who dislikes AI to just find their nearest operation centre and destroy the machines with a baseball bat, or even just pull the plug on them, the most obvious concern is to figure out how to positively employ AI. Industrial change has seen both things happen.
The Luddites were weavers who were put out of business by mechanical looms, and they restored to destroying the machine until the British authorities put them out of business- this was several centuries ago, but the parallel is obvious. Likewise for the advent of "automation" in the 20th century- labor unions protested the process because it obviously put their members out of work, but their employers ignored and oppressed them for doing so. The unions, however, still retained some influence as a bargaining body for those who were left in their trades, something the Luddites never even considered that they could do.
They want to force us to use AI, but the most effective way is to ignore and bankrupt them so they will have to spend money on something other than AI for once- or, more reasonably, determine what AI can be used ethically and morally for in professions. Both things are happening now...
I’m very much here for the ethical/moral direction! I’m glad to have good thinkers like you here too, David. I’m optimistic that minds like ours and others here will continue to point out things to consider - to me, simply getting these things on the radar of people is the most important thing I can do.
Very good point I think a lot of folks miss. AI isn’t “replacing humanity” so much as it is our best mirror so far. It’s everything we’ve ever done, basically, distilled into digital form. This is very much our proxy.
Preach!
I've been talking about learning to be deliberate in how you communicate with AI instead of "splatterprompting" for a while now. I'm also a fan of the iterative process: Ask for a dragon, see what AI returns by default, compare it to your vision, spot what's missing or different. Ask for a new dragon with those things corrected. Rinse, repeat.
But I think putting things into words is just as relevant, perhaps counterintuitively, when it comes to simply writing for an audience. Because yes, AI can spit out passable prose at scale in seconds, but having a unique voice is only possible when you're not an "average common denominator aggregator." I really think we'll see more people gravitating away from midstream AI slop towards people they can relate to.
So yes: Learn to describe shit for AI, but also for other humans.
I’ll keep working on it! The more I try to say to people, the more I realize how difficult communication really is.
Since it apparently has never occurred to anybody who dislikes AI to just find their nearest operation centre and destroy the machines with a baseball bat, or even just pull the plug on them, the most obvious concern is to figure out how to positively employ AI. Industrial change has seen both things happen.
The Luddites were weavers who were put out of business by mechanical looms, and they restored to destroying the machine until the British authorities put them out of business- this was several centuries ago, but the parallel is obvious. Likewise for the advent of "automation" in the 20th century- labor unions protested the process because it obviously put their members out of work, but their employers ignored and oppressed them for doing so. The unions, however, still retained some influence as a bargaining body for those who were left in their trades, something the Luddites never even considered that they could do.
They want to force us to use AI, but the most effective way is to ignore and bankrupt them so they will have to spend money on something other than AI for once- or, more reasonably, determine what AI can be used ethically and morally for in professions. Both things are happening now...
I’m very much here for the ethical/moral direction! I’m glad to have good thinkers like you here too, David. I’m optimistic that minds like ours and others here will continue to point out things to consider - to me, simply getting these things on the radar of people is the most important thing I can do.
Since AI is trained in the entire corpus of electronic information; a generic prompt spits out an average response. No one wants to be mid.
Very good point I think a lot of folks miss. AI isn’t “replacing humanity” so much as it is our best mirror so far. It’s everything we’ve ever done, basically, distilled into digital form. This is very much our proxy.
Prompt engineer...