After Layoffs, Meta’s AI Hackathon Set Employees Off

After layoffs, an AI hackathon at Meta carries a very different taste.

First, what is an AI hackathon? It is not hacking into systems. A hackathon is an event where a company brings people together for a few days to build prototypes, internal tools, or small projects around a given theme. Meta’s event is scheduled for July 14 to 16, 2026, and the theme is AI. According to WIRED’s reporting on internal discussions, many employees reacted badly: after layoffs and pressure, some already felt they were barely keeping everyday work running, and now they were being asked to join another extra activity.

This is no longer just a scheduling issue.

What the boss sees is survival

From Zuckerberg’s and the company’s side, there is a logic here.

Meta is no longer asking whether it should use AI. It is asking whether moving too slowly means getting eaten by other giants. For capital markets, AI is a growth story. For companies, it is a cost structure. For bosses, it is the next form of organizational efficiency.

So a boss naturally asks: can the same group of people deliver more? Can the same team become flatter? Can the same coding, operations, ads, recommendations, and content moderation workflows be partly handled by AI agents?

That logic is cold, but not surprising.

A company is not a family, and capital is not charity. If AI can compress parts of white-collar work into processes, tools, and agents, companies will push it.

To bosses, AI is an efficiency tool. To capital, AI is a valuation story. To employees, AI can feel like a countdown clock on their jobs.

Workers are not just afraid of failing to learn

Many people will interpret employee resistance as fear of new technology or unwillingness to learn.

That is too shallow.

For employees, the painful part is not simply whether they can learn to use AI. It is that their code, clicks, meetings, decisions, bug fixes, and daily judgment are turning into training material.

Reuters reported that Meta planned to collect mouse movements, clicks, keystrokes, and screenshots from US employee computers to train AI agents that can handle everyday computer tasks. Meta said the data would not be used for performance evaluation and that sensitive information would be protected.

But workers will naturally think one step further.

If my operations, experience, and judgment can be observed, recorded, and trained on, then one day the company may no longer need me. It may need a digital worker trained by me.

That is the deeper fear.

It is not: “I cannot use AI, so I will be eliminated.”
It is: “The more I cooperate, the more parts of my job may be distilled away.”

The irony is brutal: the company is not only asking you to become more efficient, but also asking you to enthusiastically improve something that may reduce your own value.

Why the hackathon lit the fire

If Meta had simply announced another AI tools training session, it may not have felt this sharp.

But this is a hackathon.

The word used to carry a bit of engineering romance: voluntary building, quick prototypes, creative ideas. But when it comes after layoffs, team changes, data collection, and AI metrics, the taste changes.

Business Insider reported that some Meta teams had specific AI usage goals. For example, one Creation organization goal for the first half of 2026 was for 65 percent of engineers to use AI to write more than 75 percent of committed code. Other goals reportedly included 55 percent of code changes being agent-assisted and 80 percent AI tool adoption among mid-level and senior engineers. Meta said rewards are based on the impact of AI usage, not simple usage frequency.

From the boss’s side, that sounds reasonable.
From the employee’s side, it becomes pressure.

If you do not use AI, how do you prove you can keep up?
If you use AI and productivity rises, will the team need fewer people?
If you help fill in process details for AI, are you making your own replacement more complete?

That is the sting of this conflict.

The AI trend may be unavoidable, but employees are not obligated to package the process of being replaced as enthusiasm.

This will spread beyond Meta

Meta is only showing the contradiction earlier than others.

Many companies will likely roll out AI in this order: first encourage it, then make it the default, then push outcomes into performance evaluation, and finally recalculate the organization. Each step can sound advanced, reasonable, and market-driven on its own.

But together, workers see another line:
work is recorded, experience is absorbed, jobs are split apart, and value is compressed.

So this is not just an internal Meta drama.

It is a reminder for ordinary technical workers: when a company says it is going fully AI-native, do not only ask whether the tool is good. Ask three harder questions. Will my data be used for training? Who owns responsibility when AI output goes wrong? When efficiency improves, who keeps the gains?

If the gains go to the company, the risks stay with individuals, the experience goes into models, and the job is repriced afterward, then employees hating AI is not surprising at all.

After layoffs, an AI hackathon looks bad not because of the event itself.

It looks bad because the company is turning people into raw material while still expecting them to applaud the machine.

I’m Easton Hua, also known as Lao Hua. Follow me for more on AI, tools, and information gaps.