Hacker News Reader

Comment by emp17344 | Hacker News Reader

No, it’s… fine. Useful in a limited capacity. Not the machine god, but not machine Satan either. The reality is kind of boring.

View on HN

This summarizes mostly how I feel about it. It's a tool like any other tool we have advanced since the beginning of human civilization

Machine tools replaced blacksmiths

CNC machines replaced manual machines.

Robots replaced CNC machine tenders

CAD replaced draftsman (and also pushed that job onto engineers (grr))

P&P robots replaced human production lines.

The steam train replaced the horse and cart

This is a tale as old as time itself

Those were deterministic rather than stochastic

Exactly. People love to compare LLMs to power tools for carpenters and smiths. But if my miter saw had a 20% chance to produce cuts at a 45 degree angle when I have it set for 90, I would throw it out so fast I would leave Looney Tunes style tracks. A tool which only sometimes does its job is worse than no tool at all.

To be rather pedantic your miter saw probably doesn't cut exactly 90 degrees. Especially if you reset it. LLMS are low accuracy for sure but so are humans. I am not saying AI is going to replace us all in a whole entirety my broader point is that these tools will be another tool that changes the market share of jobs.

The low accuracy of human guesstimation is why we use deterministic tools, not so-called tools that imitate our ability (or worse).

Tools are not replacements for people! Tools are enhancements.

AI is an attempt to replace people with something unhuman.

What do LLMs replace, pray tell? More like moving from a screwdriver to a drill, rather than replacing the carpenter all together.

Also note that there are inventions that may “replace” some part of a process, but actually induce a greater demand for labor in that process. Take the cotton gin, for example, which exploded the number of slaves required to pick cotton.

This isn't even our first AI hype cycle. That happened in the late 70s-80s. Every lab and agency needed Lisp machines to teach computers how to identify Russian missiles—or targets. The "GOFAI" techniques did not live up to the expectations of them, but they settled into niches where they were tremendously useful, and life went on. The same will happen with today's matmul-as-a-service AI.

Reply on news.ycombinator.com