If your company actually became vastly more productive, because of "AI" or whatever, you wouldn't lay off people: you would take on new efforts, make higher profits, have more things going on.

Layoffs are always - always - a sign of bad decisions, misallocation of priorities. Gross mismanagement. Always.

@RonJeffries
I think in general, even where generative AI kind of works, there's no productivity gain. It allows you to produce a great *volume* of documents, code, etc but not more *value* because that isn't in the text but in the ideas, understanding, morale etc of people in the business.

If you lay them off, you're admitting you didn't know what to do with that value or know that's where the value was.

@petealexharris @RonJeffries

A developer who just writes code, without understanding the system they’re working on deeply, is a liability. An engineer who understands the system can prevent risks.

The recent production outages at AWS, supposedly induced by AI-generated code, tell you that they’re using it the wrong way, in this context.

But there’s a reason I don’t hire people who just write code. I need systems thinkers who love to collaboratively innovate.

@systemalias @petealexharris @RonJeffries

The same principles of software quality assurance appy to AI generated code as human generated code if not moreso. Sure AI can output classes and facory patterns all day long. How is AI at version control or system integration? Enterprises can scarcely employ enough systems/data architects who can sufficiently communicate with their human counterparts let alone AI. How the hell can AI be trained to consider/evaluate/adjust to any company's complex (oh yes, they ARE complex) tech landscape?

AI is the emperor without any clothes. Delivering quality software and systems requires a farm to table/supply chain model. Quality outcomes come from quality ingredients. Companies didn't want to pay for the requisite resources before AI and it appears they're less inclined to do so with AI.

@greg_b @systemalias @RonJeffries
Yeah, you can't do quality assurance on what you don't understand, and systems thinking is built on understanding the components of the system, including algorithms and data flows. You come to that understanding best by building such components. Those who try to shortcut with code generation tools do not acquire, hone, or retain that understanding.
There may be a right way to use those tools, but it's definitely not to replace "people who just write code"

@petealexharris @RonJeffries Underrated comment. 🔼

Value isn't just more output.

@RonJeffries by that argument, shouldn’t a board seek a new CEO within a quarter, after each layoff?

@RonJeffries @systemalias

Not necessarily, but the stockholders should demand a new board.

@systemalias @RonJeffries yes, but with some wiggle room for if the mistakes were the previous CEOs domain.

@RonJeffries
Where I work we made gains with ai. Also hired a few people, juniors and seniors.

Layoffs can also be a sign of a downturn.

@Noisecolor @RonJeffries CEOs are being paid, quite handsomely, to forsee downturns and adjust without massive layoffs. If the AI is making the talent pool that much more productive, why would you get rid of them? So you can hire again when you expand because you're so successful, locking in a sunk cost of training new people? The original assessment of their incompetence is accurate. The only group that AI should be replacing is the executive group.

@Karnbot13
Maybe they are doing their job and predicting a downturn?

Ai is not causing the layoffs. It's just a handy excuse.
@RonJeffries

@Noisecolor @RonJeffries It looks like we agree that the CEOs aren't doing their jobs. I agree that AI is being used as an excuse to downsize because they overestimated the needed staffing levels for their actual share of the market.

@RonJeffries You are implicitly advocating for companies to grow without bounds. Which is what investors want, but is in reality impossible, unless other companies shink. But which the public should not want, because it creates monopolies, which are always bad, unless they are socially owned.

A company is a collection of divisions, each providing a single product or service. Each division has a natural size limit, defined by the market’s ability to consume. If a company always reinvests its financial resources, it must eventually create new divisions, eventually becoming a behemoth conglomerate. In practise, this usually happens not through new innovation, but by mergers and acquisitions.

Things are more complicated than your toot admits. The market does not have infinite capacity to consume products, thus companies cannot grow without bounds. There is always a limit.

@RonJeffries as a manager: strongly agree.

(That’s before I say that the productivity gains I hear about, that I believe, are in the 10-25% range, which isn’t nothing but unless somehow you have zero ability to grow revenue … and if you have zero growth options see your point about bad decisions etc)