Meta says Goodbye to Those Who Won't Use AI

Thursday, May 28, 2026

Meta is the latest company to trim its workforce as a result of the growing use of AI within the industry.

Posted by LampLighter at 03:31 PM | 26 COMMENTS | permalink | Comment on This Entry |

Last week, Meta laid off 8,000 employees and reassigned another 7,000 to train AI models. So when a software engineer posted a farewell parody video to the tune of "American Pie" in an internal message board, staff thought it perfectly captured how the company's culture had fundamentally shifted.

[image or embed]
-- Mother Jones (@motherjones.com) 3:30 PM Â· May 24, 2026

Comments

Admin's note: Participants in this discussion must follow the site's moderation policy. Profanity will be filtered. Abusive conduct is not allowed.

More from the article ...

... "AI is the most consequential technology of our lifetimes," Zuckerberg said in a memo that he sent to employees, informing them of the cuts. "The companies that lead the way will define the next generation," he added.

The company has not revealed too much detail of the changes in the workforce, but it's clear that jobs focused on AI infrastructure will be protected.

Meta is not alone in announcing cuts. In a blog this month, Cisco said it was cutting 4,000 jobs and Microsoft is looking at inciting employees to take voluntary retirement for the first time.

The Meta reorganization is following the trend that businesses that don't adapt to AI usage will struggle. Earlier this year, PwC US CEO Paul Griggs caused consternation when he suggested that executives who failed to get to grips with AI had a limited future in the company.

While workforces are increasingly dependent on AI as a path to progress, IT departments are not necessarily on top of the game. A Dataiku survey earlier this year revealed that 74% of CIOs were fearful that their career paths were dependent on AI outcomes. ...

#1 | Posted by LampLighter at 2026-05-27 12:26 AM | Reply

AI is just the excuse, not the reason.

The reason is maximizing short term profits even at the cost of a worse product.

#2 | Posted by Sycophant at 2026-05-27 01:01 AM | Reply | Newsworthy 3

@#2

I do not disagree.

But another aspect of AI that needs to be discussed, imo ...

Prior ~innovations~ in the workforce have mostly been targeting blue-collar labor. That is, eliminating blue-collar jobs.

But AI is looking different.

AI, as it is being moved into the workforce seems to be targeting white-collar labor. That is, eliminating white-collar jobs.

Perhaps that may be a reason why there seems to be such a resistance to AI moving into the workforce?

#3 | Posted by LampLighter at 2026-05-27 01:08 AM | Reply

White collar ain't Special. Just overpaid.

They have a big Wake Up coming.

They're Expendible Too.

Labor movements might benefit from this.

If the Tech Bros Wake Up in time.

I'm thinking they're too Entitled and Complacent to See the Writing on the Wall.

Be Afraid, Be Very Afraid.

The Jobs being lost will be Good ones.

More money for the Ownership Class, none for the Serfs.

Like Trump Intended.

#4 | Posted by Effeteposer at 2026-05-27 11:47 AM | Reply | Newsworthy 1

From the cited motherjones.com sub-summary link ...

... The tech giant made thousands of engineers train their AI replacements"then fired them. ...

This week, Meta laid off 8,000 employees"10 percent of the company's staff"and reassigned another 7,000 to train AI models. Fear of the layoffs had been building around the company for weeks, compounded by the way that Meta has taken a sharp turn from a company built by coders to a company that has staked its future on AI. So when a Meta software engineer named David Frenk posted a farewell parody video to the tune of "American Pie" in an internal message board, staff thought it perfectly captured how the culture of the company has fundamentally shifted. They begged him to post it to YouTube, making their plight inside the company public.

"There's a bit of a disconnect," one former employee who asked not to be identified told Mother Jones, "This is a company of really smart people who work really hard"coders, engineers, designers"people whose creativity and intellect is a part of their job. And you are being told that this AI agent can do it better than you, and you are being asked to train it." ...

#5 | Posted by LampLighter at 2026-05-28 03:47 PM | Reply

The Singularity is Nearer..

#6 | Posted by donnerboy at 2026-05-28 03:58 PM | Reply

Cutthroat capitalism at its finest...cutting jobs, depreciating lives, ruining family hopes...to ally the super rich with the robot called Power, America's future of wreaking havoc for profit.

#7 | Posted by Hughmass at 2026-05-29 06:46 AM | Reply

I may have mentioned this before. But the thing about singularities ... Once you realize you have crossed the event horizon it's too late. And the problem is you cannot tell where the event horizon is exactly. So. In fact. We may have already crossed that horizon and there is now no turning back.

So best hang on and try to enjoy the ride. And yes. Those who won't use AI will be left behind.

#8 | Posted by donnerboy at 2026-05-29 11:55 AM | Reply

Hugh you think AI is an American driven issue alone? This is a global shift regardless of the political, societal, or ideology leanings. China is driving forward with AI just as much. Blame on stuff, call it evil, etc... but it's not going away. Get on board or get left behind.

And i'm sure every key point in history had someone grousing and carping about the loss of jobs to a new technology.

Just look at society since the blacksmith has been put out of work, evil capitalism!

#9 | Posted by kwrx25 at 2026-05-29 01:05 PM | Reply

- grousing and carping

aka starving and dying.

#10 | Posted by Corky at 2026-05-29 01:08 PM | Reply

"AI is the most consequential technology of our lifetimes"

That's what he said about the metaverse. AI is great for many things but these greedy AI billionaire bros need to realize that we don't really want to be in some sci-fi dystopia where our lives are controlled by GPU's.

#11 | Posted by Derek_Wildstar at 2026-05-29 01:39 PM | Reply

Just look at society since the blacksmith has been put out of work, evil capitalism!

POSTED BY KWRX25

Blacksmiths were able to integrate into specialized manufacturing, farrier work (shoeing horses), or factory jobs. Today, many modern smiths work as custom metal artists, fabricators, and historical restorers, using both traditional hammers and advanced machinery. There are still blacksmiths today.

Also there are only ever about 100,000 blacksmiths in America ever.

Where will all the workers go who are currently projected to be displaced by AI?

Many will be able to integrate into adjacent fields but many will not. So it wouldn't it be great to have an actual plan for them don't you think?

Because right now it is estimated that 85 million jobs globally are projected to be displaced by AI and automation. That's a lot bigger than 100,000. And because companies rarely list "AI" as the official legal reason for a layoff, the exact number of people currently out of work is not really known but tracked through economic modeling and corporate outplacement data.

That's a lot of disruption. And it appears to be escalating rapidly.

#12 | Posted by donnerboy at 2026-05-29 01:39 PM | Reply

If Idle Hands are the work of the Devil, then what will our traditional Me First! society become when 2/3 of our fellow working class travelers are given no place/purpose in that AI uber alles scenario? Let them eat 1's & 0's? It would appear the Word O'the Day is...Dystopian.

#13 | Posted by dutch46 at 2026-05-29 03:54 PM | Reply

A product so good you have to force your employees to use it...

I've been less than impressed by my experiences using it (mostly ChatGTP and Gemini), especially the frequency with which the models simply make s*&^ up.

#14 | Posted by jpw at 2026-05-29 08:02 PM | Reply

Those who won't use AI will be left behind.

#8 | Posted by donnerboy

It's more nuanced than that.

People who overuse it will likely get left behind as well as it generates garbage.

Those who are going to be best off are those who know how to effectively manage/use AI and can effectively judge the tasks where it's going to be most effective.

#15 | Posted by jpw at 2026-05-29 08:10 PM | Reply

Would you trust an AI model that you didn't train?

I suppose it varies greatly by context, but generally speaking, there is significant risk here.

There are significant known and unknown and thus un-risk-scorable risks in deploying a machine that you don't really know how it works.

#16 | Posted by snoofy at 2026-05-29 08:13 PM | Reply

Those who are going to be best off are those who know how to effectively manage/use AI and can effectively judge the tasks where it's going to be most effective.

#15 | Posted by jpw

Field with lots of data that's supposed to be cohesive will do well.
Mathematics. Law. Medicine. Drug discovery. Anything that already has a well defined and robust ontology with history behind it. And, crucially, in these fields there's quite often a True and a False, and it's testable.

Fields where people lie constantly to get ahead --like business and politics and relationships -- being "best off" will probably come at some incalculable and unforeseen cost.
There's no universal true or false: My true is your false, your loss is my gain.

Sort of like how it's cool that we have cars and sporks.
But the ice caps are melting and microplastics are in every living cell.

#17 | Posted by snoofy at 2026-05-29 08:18 PM | Reply

- microplastics are in every living cell.

"One Word: Plastics"

www.youtube.com

1 min

#18 | Posted by Corky at 2026-05-29 08:24 PM | Reply

Corky I liked the Doonesbury on that, way back when.
Instead of Plastics, it was... Handguns.

#19 | Posted by snoofy at 2026-05-29 08:34 PM | Reply

#20

And eventually they got together for the White Nationalist Special, the Plastic Gun!

#20 | Posted by Corky at 2026-05-29 08:40 PM | Reply

On redt they've got bots making posts and more bots commenting on those artificial posts, make-believe garbage.

#21 | Posted by grumpy_too at 2026-05-29 10:55 PM | Reply

Anyone saying that it produces mostly garbage hasn't used the newer models that came out in the past 3mos to write code.

I wrote an app at last week in 15 minutes that continuously updates an html file on disk for me that autorefreshes in a browser. It scans my outlook, webex, Jira, and github and creates a running list of issues of note for me to act one, and a todo list based on other signals from those tools that I most likely need to take action on above awareness.

Those 15 minutes of time prompting the AI has saved me hours everyday since I wrote it. If you call that "garbage" then I think your bar is set way too high. The past few months have shown an exponential improvement in the coding quality in the newer models. Opus 4.8 just came out, maybe even just yesterday, so we'll see if the improvements continue to improve at this rate.

Start learning how to use Gemini, Codex, Claude, to write skill files to start cloning your abilities in different parts of your job, or someone else will be walking you out the door.

#22 | Posted by kwrx25 at 2026-05-30 02:24 PM | Reply

Given the advent of all I said above, One poster here definitely will be proven right... it's time to take a hard look at Universal Basic Income. The only alternative is that like the Blacksmith example... If you look at the jobs that existed then vs the jobs that exist now the vast majority did not exist back then. The jobs of tomorrow, probably don't exist today.

#23 | Posted by kwrx25 at 2026-05-30 02:28 PM | Reply

it's time to take a hard look at Universal Basic Income.

Absolute pipe dream.

#24 | Posted by ClownShack at 2026-05-30 02:54 PM | Reply

#22 I was using it for job search related tasks.

Resume optimization for specific job listings. Finding job listings across my preferred locations from all sources (job boards, company career sites ect) based on keywords for title, responsibilities, associated skills and several other tasks.

The resumes and cover letters required significant revisions to remove stuff the AI put in that wasn't true. Job listings were pretty good but often were outside of my field an didn't really match based on skills/keywords but had a title that was close.

It saved me a lot of time, but there was still the absolutely necessary need for human oversight and curation of the results. That was my point.

I haven't had the ability to use it professionally because the nature of my work prohibits the use of open source models. It would be considered a major data breach if I plugged my info into Gemini or Claude. You either have to internally train a secure, closed model or obtain one and the development of them for the field I'm in is in early stages.

#25 | Posted by jpw at 2026-05-30 03:02 PM | Reply

Would you trust an AI model that you didn't train?

#16 | Posted by snoofy at 2026-05-29 08:13 PM | Reply | Flag:

What kind of model? I've trained 10 million in the last 14 days. If you can make a context aware bigru model with a better accuracy, f1, log loss, and precision than mine, and it passes real world validation, yes.

#26 | Posted by sitzkrieg at 2026-05-30 03:40 PM | Reply