Comment on '3d-printing a screw' is a way to describe how AI integration is stupid most of the time
Colonel_Panic_@eviltoast.org 2 days ago
You “nailed down” screws? I think that’s your problem. You are supposed to hold the screw with a wrench and then rotate the entire piece of wood around it. What an idiot.
On a serious note. That is absolutely how I feel about LLMs. For one, LLMs are a specific subset of ML, which is itself a subset of a full “AI”. I hate that LLM has been branded AI when that is completely misleading at best and a flat out lie at worst. Sure it’s a part of AI, but it’s like marketing a steering wheel as a whole car. It’s very useful as a tool for a specific job, but saying it’s a car makes people think it’s all they need to travel somewhere.
LLMs are really just “Google Search 2.0” it just scrapes all the sites and combines them into a somewhat more concise and useful answer, but since anyone can post anything on the internet and those things can be true or false or anything in between, so too will the answers LLMs provide you. It’s risky to put ANY stake in the answers without knowledge yourself of what it means.
prime_number_314159@lemmy.world 2 days ago
A 1.5" long number 8 wood screw has slightly more than half the holding strength when it is driven by a hammer into 2 3/4" pine boards, relative to the traditional rotating the workpiece approach, but takes only about 1% of the time to install. You can very easily add a second screw to achieve the desired strength, while often gaining resistance to torque in the workpiece.
Meanwhile, 3D printed screws usually break when struck with a hammer, so they’re no good for this approach.
Burninator05@lemmy.world 2 days ago
Why would I want torque resistance at work? Getting torque is the only reason I have a job.