[asadotzler]

What I'm seeing is marked decrease in standards and quality everywhere I see LLMs (and diffusion models, any of the transformer-based stuff.)

Sometimes that decrease in quality is matched by an increase in reach / access, and so the benefits can outweigh the costs. Think about language translation in web browsers and even smart spectacles, for example. Language translation has been around forever but generally limited to popular books or small-scale proprietary content because it was expensive to use mult-lingual humans to do that work.

Now even my near-zero readership blog can be translated from English to Portuguese (or most other widely used languages) for a reader in Brazil with near-zero cost/effort for that user. The quality isn't as good as human translation, often losing nuance and style and sometimes even with blatant inaccuracies, but the increased access offered by language translation software makes the lower standard acceptable for lots of use cases.

I wouldn't depend on machine translation for critical financial, healthcare, or legal use cases, though I might start there to get the gist, but for my day-to-day reading on the web, it's pretty amazing.

Software at scale is different than individuals engaging in leisure activities. A loss of nuance and occasional catastrophic failures in a piece of software with hundreds of millions or billions of users could have devastating impacts.

[crabsand]

What I have in mind as an analogy is house cleaning or laundry. In the past these would consume so much time and when washing machines appeared, they probably didn't wash as the "best washer" but they let a large percentage of population to spend less time with laundry and people began to see having clean clothes as a standard.

As you said the current state of LLMs is not dependable but we'll learn how to deal with their poor quality and come up with solutions. The overall secondary effect may be increased standards for the knowledge work everywhere. No one would bother to read TOS for a SaaS, now you can ask for a summary from LLM, or ask it to write a privacy notice to your blog.