Cars turned us—one of the best species in long distance running into couch potatoes.
Now llms are attacking our brains and making us stupid and insane. A species of slopheads if you will.
Cars turned us—one of the best species in long distance running into couch potatoes.
Now llms are attacking our brains and making us stupid and insane. A species of slopheads if you will.
That seems plausible, but in reality I’m not actually sure that’s true. Even before “AI” (which I fully know is a false buzzword and really should just be “LLMs”), I just stuck to the only known way of doing things if I was unsuccessful. Basically, the LLM is merely an additional opportunity to find resolution; I would’ve actually given up at the level of annoyance where I was before trying it anyway.
It actually takes finessing to understand how the model may likely be perceiving your problem, hence the entire sub-industry of experienced prompt engineers. You have to be patient in identifying why it messes up and how to guide it towards accuracy, and there are certain ways to address this.
It’s about generation of optimal solutions in the first place versus ones that don’t work. People who aren’t at least familiarizing themselves will be left behind. I speak this as someone who is wary of LLMs, is fully aware of their copyright disputes, and tries to use them less than once/week.
They will emerge from the experience of not relying so heavily on LLMs for very complex matters as opposed to simpler ones.
Just my suspicions about where this is all going…
Fair point that I could have elaborated on: I know people using interconnected agents to shrink 5 hours of work into half an hour, like it or not. Let me share what he said:
How can you fight this? Ethical or not, we will fall behind if we shun it like luddites. At the same time, though, I think the bubble may burst for extremely complex operations revealing faults, which would pull employers back to using it in lower-level capacities; either that or else the killer may be permanent ecological devastation. Either way, we have certainly opened Pandora’s box and it’s going to come to some breaking point. If on the off-chance that none of these calamities comes to pass, though, then we will really fall behind all the more acclimated.
If AI requires so much patience and persistence to use properly, how do you expect to achieve anything with it? You, who by your own admission, quits when you are initially unsuccessful?
Is “less than once a week” of time investment your official recommendation of how much is needed to “not get left behind” in developing our “prompt engineering” skills that will be in such high demand?
To be honest I don’t understand why I’m supposed to be anxious about your random office worker buddy using AI set up some reminders and calendar invites. Does it not strike you as odd that literally nobody can come up with specific, concrete examples of how the technology has improved their efficiency as a matter of fact? Like it’s all just a vibe they have followed by the bare claim that they are working 20 times faster, but that reality never seems to materialize in a way that can be measured by anyone else.
He also describes it as “soooo easy”, so again, where is this investment of skill that I should be worried about not doing? Like are there any “prompt engineering” skills that take more than a few minutes to learn?