jjj

Great new(/s)! Ford has patented this already: https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/20230055958

This kind seems like it’s basically what I said?

My mistake, I associated what you said as

(machine learning) is not the same as (“AI” (LLMs, presented as beings capable of thought))

(aka LLMs aren’t ML) rather than

machine learning is not the same as “AI” e.g. (LLMs, presented as beings capable of thought)

(aka LLMs are ML, and when presented as a replacement for humans they are AI)

Either way, I believe there is some value in comparing even simple decision making technology to the human brain. After all, we are capable of making those sorts of decisions as well. Tbf I may be biased by my internal mental model of coding, where I imagine describing a set of steps, relations, abstractions, etc to an entity with an extremely direct manner of interpretation.

My personal moral of (seemingly malicious) language ambiguity is to avoid the term AI altogether.

Thanks for the discussion!

I appreciate that sentiment, and much more the indication that I’m not mistaken to choose the “never deliberately run the output of an LLM” path.

machine learning is not the same as “AI” (LLMs, presented as beings capable of thought)

(insert nerd face) LLMs are a kind of machine learning algorithm. AI is actually a superset of ML, in fact pretty much all computer programs count as AI. (a lot of old computer science departments were once called AI research departments in stead)

The phrase “AI”, as promoted by tech bros, however, should be considered seperately to the correct definition: I agree that software is generally harmful when presented this way by its creators.