Biggest Losers of AI Boom Are Knowledge Workers, McKinsey Says | Hacker News Reader

Biggest Losers of AI Boom Are Knowledge Workers, McKinsey Says

bloomberg.com

32 pointsmirthlessend14 comments

Still waiting for the day when all of the truck drivers are put out of work and revolt as promised 5 years ago…

McKinsey et al put out these BS reports every year just regurgitating the current hype with zero value add (except for lazy journalists who want an easy story to copy paste).

The possible function of this report, as with the truck drivers, was to provide justification to companies to reduce the wages of the workers in the report. It provides a way to point the finger at a looming "threat" that they'll be replaced, so don't demand increases in wages. This is keeping with the value that McKinsey provides to companies.

You know what they say about the first 90% taking up 90% of the development time and the last 10% taking up 90% of the development time. That is likely what we are seeing with current AI efforts. They get amazingly close yet aren't good enough to be relied upon.

Yes.

My point isn’t really to comment on the specific developments in AI itself, but to assert that McKinsey has no special insight on the topic.

I'm guessing that you didn't mean to end up with 180% of development time, but that does seem about right when it comes to the software development projects that I've been part of. I also think you're correct that despite the impressive recent advances in AI, the remaining effort may still be enormous.

The 180% is purposeful (supposed to be humorous)! See: https://en.m.wikipedia.org/wiki/Ninety%E2%80%93ninety_rule

Edit: though I guess, in a sense, nobody means to put him 180% of planned effort

You know, we could save a lot on consultants if we had LLMs write the BS reports.

I am still waiting for cashiers to go out of work let alone truck drivers.

So far it seems that AI just as likely to make stuff up as it is to provide solid reliable knowledge. As long as that remains the case AI isn't going to make much progress because there's too much to lose when you can't be sure if the knowledge is bogus or not

https://archive.is/mhYIn

I wonder if it's the case.

It's certainly quicker to learn things with LLMs but they hit limitations fairly quickly.

Every project a have worked on has a huge backlog, on launch the MVP is often trimmed back to be even more minimal.

Launch and project end seem dictated by budget.

Maybe we will just end up with software that is more complete.

Did previous productivity tools such as intellisense or before it, IDEs reduce the amount of software built or the amount people were paid, I think the opposite happened.

The whole idea is absurd as if we are operating at the optimal peak level of software and the optimal peak level of knowledge work.

I would imagine there are far more people that work with databases today than there ever were people working on filing papers in a filing cabinet. Filing papers in a file cabinet was a massive constraint on what would be stored.

We are highly underestimating a boom in knowledge work from AI breaking so many constraints on knowledge work.

I don't think so. In fact I can already see a new category of jobs for knowledge workers - the "AI BS verifier". I predict millions of openings.

That is why we see so much fearmongering and neuroticism regarding AI and office workers, they are not used to being the ones whose jobs are getting shipped away.

Reply on news.ycombinator.com