Thoughts on LLMs
This is a page where I try to suss out my thoughts on LLMs.
At a high level: I’m skeptical.
I’m not skeptical of the technology, but of how it is being deployed.
Two related issues:
- that these tools prioritize generation over explanation
- that these tools are tightly coupled to corporate interests
I think part of the land grab we are seeing right now, where many groups are throwing all kinds of noodles at all kinds of walls is because corporate interests see value in controlling the stack, soup to nuts.
Turning engineers into “prompt engineers” is the most direct way they’ve ever had to do this — the backing systems (GitHub, TypeScript, NPM, etc.) stop mattering when you give folks a tool that just does the thing when asked…and slowly folks become wholly dependent on the system asked, the system you control. By focusing on generation over explanation you don’t help folks learn how to do a thing, you make them dependent on the tooling you provide to do that thing.
A big ol’ give a person a fish vs teaching them to fish.
The reason folks like Microsoft are going so hard on AI is they’ve got an opportunity at turning whole classes of professionals into “prompt engineering” supplicants, redirecting funding and employment into business models conditioned on wedging AI-anything into it somewhere…they’re not automating people out of the workplace, they’re adding a new dependency.