Meta’s Mark Zuckerberg has seemingly pushed the company to the brink, squandering billions on harebrained tech, destroying the engineering culture that built his Facebook empire, losing badly in the AI race, and enshittifying their core products.
Meta’s Tea Spills All Over the Media
It’s been a season of remarkably bad press for Mark Zuckerberg and Meta.
First there’s this obituary for his core products written by Julia Angwin for the New York Times:
There is a moment when internet companies get the stink of death on them. For AOL, it was 2003, when it became clear that its users were abandoning its clunky dial-up internet service for far-faster broadband. For Yahoo, it was 2015, when its last-ditch acquisition spree failed and it sold itself to Verizon.
For Meta, that time is now. I believe the company — one of the most powerful media organizations in the world and one of the most valuable members of the S&P 500 — is at the start of a long, slow decline that will trigger aftershocks to our economy and our society.
…Meta’s earnings are starting to show the strain from years of growing consumer disaffection and reckless spending. The latest earnings, released on April 29, revealed a dip in user numbers for the first time since it started reporting these figures. And the slumping stock confirms what we have all known in our guts for a while: This is a company entering its zombie era.
Death is different on the internet. Lifeless companies like AOL and Yahoo are still technically with us. You can visit their websites. They have customers. They may even be profitable, as they cut staff and monetize their last remnants of traffic. But they are, as the kids say, peak cringe. Many teenagers wouldn’t be caught dead with an AOL account, a Yahoo email address — or a Facebook profile.
Then there’s this dramatic headline from Wired:
— Nat Wilson Turner (@natwilsonturner) June 17, 2026
Turns out there is such a thing as bad publicity.
Engineering Culturecide
The Wired piece focuses on the aftermath of Zuckerberg’s recent AI moves and mass layoffs:
growing frustration inside the company’s Applied AI team, which was formed in March to support the work of AI researchers at Meta Superintelligence Labs. Three current employees tell WIRED there is widespread dissatisfaction with how Meta assembled the unit of about 6,500 engineers and product managers and the drudgework they allege they have been assigned to improve AI models.
…
“It’s literally the gulag,” one of the employees claims. “You have zero purpose in life all of a sudden, you barely interact with anyone, you just have these tasks every week.”Another employee describes some of the tasks—generating puzzles to test how reliably AI models from Meta and other companies can solve them—as easy compared to the software development work they had been doing previously. But the new projects feel menial, and “almost all” employees seem unhappy, they say.
“Most people find the work soul-crushing,” the third employee says.
Gergely Orosz at The Pragmatic Engineer has more, asking “Why is Meta destroying its engineering organization?“:
For two decades, Meta had a unique, high-performance engineering org; right up until around April of this year. For the first 20 years of the company’s existence, it had a “move-fast-and-break-things” culture, and in the early 2020s this shifted to a “move-fast-with-stable-infra” one. Engineers I know at the company were empowered to do good work, focus on impact, and to balance business interests with solid engineering.
But in the past few weeks, all that has changed, as if the leadership has been following detailed blueprints on how to demolish a proven, successful engineering culture in the most ruthlessly efficient way possible.
…what’s going through the minds of leadership who are reducing software engineering there from the profit center that it was between 2004 until very recently, to the disdained cost center that it has become in just a few weeks.
And given that Orosz is the scribe of the most in-depth analysis ever done of Facebook’s engineering culture at its peak in 2022, his eulogy carries some weight.
He diagnoses what may be motivating Zuckerberg, he failed to build a mobile operating system in the 2010s and is the only one of the hyperscalers not to own his own platorm.
So in 2022, having burned up to $83.6 billion on the Metaverse — a patently ridiculous virtual reality concept — Zuck was determined to not to miss out on AI:
When it became clear that AI would become a mega-trend in 2022, Zuckerberg didn’t miss it: he assembled the internal FAIR group (Fundamental AI Research team) as well as a GenAI product organization and released a series of open-weight AI models:
- Llama 1: released in Feb 2023, three months after ChatGPT, built by FAIR
- Llama 2: in June 2023, built by the GenAI product organization (as well as all subsequent Llama models)
- Llama 3: in April 2024. This model was Meta’s most competitive LLM of all, and gained momentum in adoption across the industry
- Llama 4: in April 2025. This model was deeply disappointing
In June that year, Meta acquired a 49% stake in Scale AI to reboot its AI efforts for a whopping $14.8B, and brought in Scale AI’s CEO, Alexandr Wang to take over Meta’s AI strategy. The acquisition of Chinese startup Manus AI for $2B is currently in question after China blocked the deal from being completed.
Based on the investment made into Scale AI and Wang, it’s pretty clear that Meta – and Zuckerberg – is determined to build a state-of-the-art LLM that can be competitive with the latest versions of Claude and ChatGPT. But Meta has to start pretty much from scratch, and it’s up to Alexandr Wang to deliver.
Most recently Zuck has made what may be a fatal mistake, he followed massive layoffs with mandatory requirements that top engineers spend time performing menial tasks designed to train AI — likely to replace them.
Here’s how the layoffs were handled in May, via NDTV:
Meta began laying off roughly 8,000 employees Wednesday — about 10 percent of its global workforce — as co-founder and Chief Executive Mark Zuckerberg pushes to redirect resources toward an ambitious artificial intelligence agenda.
…notifications went out beginning in the early morning hours, with Singapore-based workers among the first to be informed.
In addition to the cuts, Meta said in April it would cancel plans to hire 6,000 people and shift 7,000 other employees into AI workflow-related roles.
In a memo to staff Wednesday…Zuckerberg expressed thanks to departing employees and sought to reassure those remaining.
“It’s always sad to say goodbye to people who have contributed to our mission and to building this company,” he wrote. “I feel the weight of that.”
How bad were the new roles? Reuters described the jobs-turned-humiliation-rituals:
Meta (META.O), opens new tab is installing new tracking software on U.S.-based employees’ computers to capture mouse movements, clicks and keystrokes for use in training its artificial intelligence models, part of a broad initiative to build AI agents that can perform work tasks autonomously, the company told staffers in internal memos seen by Reuters.
The tool, called Model Capability Initiative (MCI), will run on work-related apps and websites and will also take occasional snapshots of the content on employees’ screens, according to one of the memos, posted by a staff AI research scientist on Tuesday in a channel for the company’s model-building Meta SuperIntelligence Labs team.
The purpose, according to the memo, was to improve the company’s AI models in areas where they struggle to replicate how humans interact with computers, like choosing from dropdown menus and using keyboard shortcuts.
“This is where all Meta employees can help our models get better simply by doing their daily work,” it said.
I must admit that from the C-Suites, automating the old “train your replacement” scam surely looks like a great thing, but it doesn’t seem to be working.
The new assignments were rolled out with the sensitivity we’ve come to expect, per Business Insider:
As 8,000 Meta employees were reading their layoff notices on Wednesday, 7,000 others opened a very different email. They had been selected to join a new AI initiative spun up directly by CEO Mark Zuckerberg — and it would be crucial to accelerating Meta’s position in the AI race.
…
“I got drafted,” one person said in a Discord server where Meta employees discussed the layoffs.
…
Joining the new initiative, it turns out, is non-negotiable — and different recruits have received the news with either relief, dread, or confusion, according to four employees who spoke with Business Insider and several forum posts.
…
Two Meta employees said they expect the new AI task forces to involve data labeling — the grunt work of manually tagging images or correcting chatbot responses.That’s an area where Meta’s AI chief, Alexandr Wang, has extensive experience. Wang is the cofounder and former CEO of Scale AI, one of the world’s biggest data-labeling companies, which relies on armies of contractors to train AI.
Meta has been rolling out other AI initiatives internally. It has launched an internal tool to track employees’ keystrokes and mouse movements, known as Model Capability Initiative
…
In a leaked audio recording from an internal meeting that same month, Zuckerberg said that Wang knew the data-labeling world well, and then offered a blunt addendum: The average Meta employee has a “significantly higher” intelligence than those contractors — so he’d rather “enlist” top employees from across Meta to train its AI instead.
Feeling warm and fuzzy, yet?
Back to Orosz for an explanation of why the new mandatory assignments galled even more:
30-50% of engineers on core teams have been forcefully reassigned to data labeling and RLHF, upsetting folks even more. Also starting in late April, product engineering teams received a mandate from above, whereby 30-50% of engineers were to leave the team and join the ADO org (Agent Data Optimisation).
“Forceful” reassignment is very relevant here because of Meta’s traditional engineering culture. Between its founding in 2004 and until last year, Meta gave engineers autonomy to choose where they work and what they work on.
…
“Data labeling” is more involved work, even though a bit repetitive. There are labeling tasks, where you look at a website and decide if it looks good or not. But then there are more involved AI training tasks, which looks like this:Come up with a task that the AI should do
Then write the tests that confirm the result
Package all of this up into a Docker container, using Harbor framework
Then read the code that the AI writes – often doing this based on feedback from several models, and give it feedback
This work is not easy to do — and you can see why you need good software engineers to do it! — but it gets repetitive real quickly, and is hardly motivating to do day-in, day-out.
But our Mark is an over achiever and he wasn’t happy with just one epic PR disaster this month.
How Did That Spyware Get in Meta’s AR Glasses?
Zuck may be torching company morale to chase the LLM brass ring, but that doesn’t mean he’s given up on older fiascos like augmented reality glasses.
And with Facebook there’s always a twist (“I’ll have mine with a little extra evil zest on the top, a little more, thanks”), this one was scooped by Wired:
Code discreetly added to Meta’s AI app over multiple updates this year shows that the feature, internally called “NameTag,” identifies people captured by the glasses’ camera and, when activated, alerts the wearer when it recognizes someone.
The discovery of NameTag in the live Meta AI app shows that Meta had begun shipping face-recognition code to users’ phones while publicly describing it as something the company was still “thinking through.” In April, Meta said if it were to utilize face recognition, it wouldn’t be rolled out without first taking “a very thoughtful approach.” But WIRED found that as early as January, core components of the system had been integrated into software distributed to millions of people.
Multiple Meta execs carped about Wired’s reporting, but the blowback was predictable, in face, a February NYT report about a memo planning the move had already triggered backlash:
Meta’s plans could change. The Silicon Valley company has been conferring since early last year about how to release a feature that carries “safety and privacy risks,” according to an internal document viewed by The New York Times. The document, from May, described plans to first release Name Tag to attendees of a conference for the blind, which the company did not do last year, before making it available to the general public.
Meta’s internal memo said the political tumult in the United States was good timing for the feature’s release.
“We will launch during a dynamic political environment where many civil society groups that we would expect to attack us would have their resources focused on other concerns,” according to the document from Meta’s Reality Labs, which works on hardware including smart glasses.
Crafty devils, I’m sure they’re surprised that anyone noticed their cleverly concealed move.
The cumulative backlash has gotten bad enough even Zuckerberg has noticed.
The ‘Mistakes Were Made’ Phase
Reuters has Mark’s new memo:
Given the complexity of these changes, we’ve made mistakes and will almost certainly make more,” Zuckerberg said, adding that he is also “focused on providing as much stability as possible” in terms of organization changes going forward.
“I don’t want to overpromise because the world is changing in ways that are out of our control,” he said, reiterating that Meta does not expect more company-wide layoffs this year.
He said Meta will try to find new roles for employees reassigned to train AI models, after the Facebook owner carried out a massive restructuring in May, laying off 10% of its workforce globally and transferring 7,000 employees to new initiatives related to AI workflows.
“By creating important new roles for people, this also allowed us to shrink the size of teams knowing that if we make mistakes in some places, then we could transfer some people back,” Zuckerberg said.
But there’s little reason to fear that Zuck has actually learned any lessons.
The Money Burning Will Continue Until Air Quality Improves
TechCrunch has the bad news for Meta investors:
On Wednesday, Meta’s earnings report showed that its embattled virtual reality business had lost some $19.1 billion in 2025, which is slightly more than it lost in 2024 (that year, the losses hovered around $17.7 billion). In its fourth quarter, the unit posted a loss of $6.2 billion, the report shows.
“For Reality Labs, we are directing most of our investment towards glasses and wearables going forward, while focusing on making Horizon a massive success on Mobile and making VR a profitable ecosystem over the coming years,” Zuckerberg said, during the call. However, the CEO noted that losses were expected to continue. “I expect Reality Labs losses this year to be similar to last year,” Zuckerberg said, while noting that this year would “likely be the peak, as we start to gradually reduce our losses going forward.”
When Meta announced a pivot toward the “metaverse” in 2021, the move was regarded with a certain amount of skepticism and, during its first year of VR efforts, the company faced harsh criticism — even being referred to as an “international laughingstock.” Nearly half a decade later, that skepticism hasn’t exactly subsided. As the VR business continues to lose money and Meta continues an aggressive pivot away from VR and toward AI, it’s unclear what exactly will turn the ailing business around.
And CNBC points out that for all his success, Zuckerberg’s empire is a one-trick pony and that trick is selling ads:
Since Zuckerberg’s company, previously known as Facebook, began selling digital ads almost two decades ago, advertising has been its only real business. In its earnings report last month, Meta said that nearly 98% of its $56.3 billion in first-quarter revenue came from advertising. It’s a remarkably lucrative market, with some of the highest profit margins in the tech industry, and one Meta has long dominated in the U.S., alongside Google.
Meta just recorded its fastest growth rate for any quarter since 2021, showing that the online ad market is currently as robust as ever. But the rapid emergence of AI has raised questions about what happens if and when users turn to new interfaces for information and are no longer spending so much time on screens where they’re exposed to a constant barrage of links.
And when Meta has asked consumers and companies to open their wallets for anything other than ads, the answer has generally been no.
Time will tell if Zuckerberg will be the next tech oligarch to destroy his own empire. Stay tuned.
Related Posts:
By
