Meta Tracks Workers’ Clicks to Build AI

Meta Tracks Workers’ Clicks to Build AI That Can Also Ignore Human Feelings on a Schedule

Five Mildly Catastrophic Observations

• At long last, a machine that understands middle management perfectly, right down to the dead-eyed nod over a lukewarm tea.
• The AI can process emotion by skipping it instantly, a skill previously held only by HR and certain headmasters.
• Workers said morale was low. The algorithm marked that as background noise and recommended a free pizza on Thursday.
• Nothing trains empathy quite like surveillance. Just ask any teenager raised by a Ring doorbell.
• The robot already schedules meetings nobody wants and overruns by three minutes. It is, in every meaningful sense, ready.

A Robot That Finally Understands Your Line Manager

MENLO PARK, California — Meta this week confirmed that it is tracking employee clicks to build a next-generation AI capable of handling modern business needs, including the efficient ignoring of human feelings on a quarterly cadence and a Friday afternoon basis.

Executives say the system learns from real workplace behaviour. Critics say real workplace behaviour is exactly the problem we were hoping AI might fix, not faithfully reproduce in HD.

During a demonstration, an employee typed, “I’m overwhelmed.” The AI replied instantly: “Thanks for sharing. Let’s circle back next quarter.” Observers described the exchange as hauntingly accurate. One HR director cried, then immediately logged the tear as a deliverable.

The Corpus of Corporate Cruelty

Meta engineers say the model has absorbed millions of examples of corporate interaction: unread messages, postponed concerns, reaction emojis replacing actual support and calendar invites sent at 4:58 p.m. on a Friday before a bank holiday. The training data is roughly 78% Slack, 19% sigh, 3% existential drift.

“This is authentic enterprise language,” said project lead Noah Trent. “The machine doesn’t dismiss emotions. It acknowledges them in a sterile tone, then redirects you to the wiki.” The wiki is empty. The wiki has always been empty. The wiki has been empty so long it has its own folklore.

Benchmarks in Bureaucratic Brilliance

Test users report the AI can now perform several management tasks flawlessly. It schedules unnecessary syncs, says “great question” without answering it, and transforms complaints into action items for the complainer. The complainer then receives a calendar invite to a workshop on resilience, run by a consultant who has never had a job.

A leaked benchmark shows the model scored 99th percentile in passive phrasing and 100th percentile in saying “let’s revisit later.” It also scored alarmingly well on “let’s take this offline,” a phrase the Chartered Institute of Personnel and Development classifies as a verbal Bermuda Triangle.

What the Funny People Are Saying

“They built a robot manager. Naturally it forgot your birthday and took credit for your idea. They’ve nailed the brief.” — Jack Dee

“I’ve worked with humans who had less empathy than this thing. So really, it’s just keeping pace with Reading.” — Romesh Ranganathan

“Nothing says progress like automating indifference. Save the bosses some time, that’s the real innovation.” — Frankie Boyle

Worker Resistance, Click by Click

Staff are fighting back by feeding the system contradictory signals. Some click angrily. Some click sadly. One intern click-whispered for help. The AI marked the cry for help as “low engagement” and suggested a stretch break.

A poll found 68% of workers fear replacement, 21% welcome replacement if the robot cancels meetings, and 11% were already replaced months ago but hadn’t noticed because the robot had identical Zoom backgrounds and the same six opinions, all of them about cricket.

Humans in the Loop, but Only at the Right Moment

Meta says concerns are overblown and the AI will always keep humans in the loop, especially when blame is being assigned. Sources at Acas were not available for comment, possibly because their inbox is also being triaged by an algorithm now.

The company concluded its event with a hopeful slogan: Technology That Understands You Enough to Monetise You. The slogan was projected onto a screen, then onto a focus group, then onto a balance sheet, in that order, with each translation losing slightly more dignity.

The Real Lesson Nobody Wants Printed

The deeper irony is that this whole enterprise is what happens when corporate culture is left to be reformed by the same people who built it. A genuinely free labour market would allow workers to vote with their feet, walk out and hand the empathy-skipping robot a P45. Instead, knowledge workers are increasingly trapped in single-employer cities, single-platform workflows and single-vendor management software, with no exit ramp that doesn’t end in retraining as a barista in a Pret. The market correction here won’t come from Menlo Park. It will come from the door.

Nigel from Accounts attempted to schedule a one-to-one with himself simply to remember what proper human attention felt like. The AI cancelled it and replaced it with a stand-up. Nigel does not stand up. Nigel has not stood up since 2017.

Auf Wiedersehen, amigo!

Meta, the parent company of Facebook, Instagram and WhatsApp, has invested heavily in artificial intelligence research, including its Llama family of large language models, alongside a growing portfolio of internal AI tools intended for enterprise productivity. The company has faced ongoing scrutiny from regulators, journalists and trade unions over its use of behavioural data, employee monitoring and the deployment of AI in workplace decisions including hiring, firing and performance management. Project Ledgerstorm, the click-tracking management AI and the leaked benchmark scores described above are satirical inventions, but the broader trend — major employers using AI to mediate, automate and increasingly substitute for human management — is real and accelerating, raising serious concerns about due process, accountability and the meaningful presence of human judgment in workplace decisions on both sides of the Atlantic.

Alan Nafzger

Alan Nafzger was born in Lubbock, Texas, the son Swiss immigrants. He grew up on a dairy in Windthorst, north central Texas. He earned degrees from Midwestern State University (B.A. 1985) and Texas State University (M.A. 1987). University College Dublin (Ph.D. 1991). Dr. Nafzger has entertained and educated young people in Texas colleges for 37 years. Nafzger is best known for his dark novels and experimental screenwriting. His best know scripts to date are Lenin’s Body, produced in Russia by A-Media and Sea and Sky produced in The Philippines in the Tagalog language. In 1986, Nafzger wrote the iconic feminist western novel, Gina of Quitaque. He currently lives in Holloway, North London. Contact: [email protected]