Snarky but serious question: How do we know that this wave will disrupt labor at all? Every time I dig into a story of X employees replaced by "AI", it's always in a company with shrinking revenues. Furthermore, all of the high-value use cases involve very intense supervision of the models.

There's been a dream of unsupervised models going hog wild on codebases for the last three years. Yet even the latest and greatest Claude models can't be trusted to write a new REST endpoint exposing 5 CRUD methods without fucking something up. No, it requires not only human supervision, but it also requires human expertise to validate and correct.

I dunno. I feel like this language grossly exaggerates the capability of LLMs to paint a picture of them reliably fulfilling roles end-to-end instead of only somewhat reliably fulfilling very narrowly scoped tasks that require no creativity or expertise.

▲

closewith 5 days ago | parent [-]

> instead of only somewhat reliably fulfilling very narrowly scoped tasks that require no creativity or expertise.

This alone is enough to completely reorganise the labour market, as it describe an enormous number of roles.

▲

badestrand 4 days ago | parent | next [-]

How many people could be replaced by a proper CMS or a Excel sheet right now already? Probably dozens of millions, and yet they are at their desks working away.

It's easy to sit in a café and ponder about how all jobs will be gone soon but in practice people aren't as easily replacable.

▲

7952 4 days ago | parent | next [-]

For many businesses the situation is that technology has dramatically underperformed in doing the most basic tasks. Millions of people are working around things like defective ERP systems. A modest improvement in productivity in building basic apps could push us past a threshold. It makes it possible for millions more people to construct crazy excel formulas. It makes it possible to add a UI to a python script where before there was only a command line. And one piece of magic that works teliably can change an entire process. It lets you make a giant leap rather than an incremental change.

If we could make line of business crud apps work reliably, have usable document/email search, and have functional ERP that would dissolve millions of jobs.

	▲	lotsoweiners 4 days ago \| parent [-]
		> If we could make line of business crud apps work reliably, have usable document/email search, and have functional ERP that would dissolve millions of jobs. Why would that be your goal? I’d prefer millions of people have gainful employment instead of some shit tech company having more money.

▲

pjmlp 4 days ago | parent | prev | next [-]

I can tell that those whose jobs depended on providing image assets or translations for CMS, are no longer relevant for their employers.

▲

PleasureBot 4 days ago | parent | prev [-]

A lot of jobs really only exist to increase headcount for some mid/high level manager's fiefdom. LLMs are incapable of replacing those roles as the primary value of those roles is to count towards the number of employees in their sector of the organization.

	▲	closewith 4 days ago \| parent [-]
		Unless AI spend overtakes headcount as the vanity metric du hour, which it already has.

▲

krainboltgreene 4 days ago | parent | prev [-]

I promise you that your understanding of those roles is wrong.