Hello, dear readers!

This is our weekly brief on remarkable AI topics, so you can keep up without drowning in the noise.

Today’s focus — AI and the strange new job market it’s creating. Meta is cutting thousands of roles while pouring billions into AI, but at the same time, tech hiring has been steadily rising since the release of ChatGPT. If AI is supposed to be killing jobs, why are companies still hiring?

Also in this week’s edition:

Tech’s split reality: layoffs surge, hiring climbs

Meta is about to cut roughly 10% of its workforce — around 8,000 people — while ramping up AI spending to $135 billion this year. The message is familiar by now: AI makes workers more productive, so fewer of them are needed.

But zoom out, and the picture gets less straightforward.

After collapsing from its pandemic-era peak, tech hiring has been steadily climbing back — and notably, that recovery has continued throughout the entire post-ChatGPT period. It’s still far below the highs of 2021–2022 — when companies were hiring as if growth would never slow — but it’s no longer in freefall. In fact, hiring is up more than 60% from its lowest point.

So which is it — are AI tools replacing workers, or is the industry still expanding?

The answer seems to be both. Inside large tech companies, productivity gains mean fewer people can do the same amount of work. That translates into layoffs, restructuring, and a push toward leaner teams. Meta’s cuts fit neatly into that pattern.

Outside those companies, the same productivity gains lower the cost of building things. Smaller teams can ship faster. New startups become viable with less capital. More experiments get off the ground — and they still need engineers, designers, and operators.

AI isn’t eliminating tech jobs so much as redistributing them. Big Tech is getting leaner. The rest of the ecosystem is getting busier.

Did AI just cost Tim Cook his job?

Apple’s decision to replace Tim Cook with hardware chief John Ternus looks, on the surface, like a standard leadership transition. In reality, it’s a strategic signal.

For years, Apple’s growth has been driven by turning the iPhone into a platform — layering services, wearables, and ecosystems on top of a single device. Cook executed that playbook exceptionally well. But AI introduces a different kind of competition.

Instead of apps and services, the battleground shifts toward infrastructure: chips, devices, and tightly integrated systems that can run AI efficiently. By choosing a hardware engineer over software or services leadership, Apple is effectively saying where it thinks the next decade will be won.

While other companies spend aggressively on models and cloud infrastructure, Apple’s advantage remains control over its own silicon. The bet is simple: let others build the intelligence, and make sure it runs best on your machines.

Cook will be remembered for scaling the iPhone era. Ternus is stepping in to figure out what comes after it.

When should AI call the police?

After a mass shooting in Canada, OpenAI CEO Sam Altman issued a public apology: the company had previously identified and banned the attacker’s account for discussing violent activity but chose not to alert law enforcement.

Cases like this aren’t new in a broad sense. People have always used technology to plan crimes. But AI changes the texture of that interaction.

A search engine is passive — you type, it returns results. An AI system is different. It engages, responds, adapts. It can sustain a conversation, refine answers, and guide a user over time. That makes it feel less like a tool and more like a participant.

That shift matters because expectations change with it. If a system can detect intent in real time — or appears to — people start asking why it didn’t act. Not just why it didn’t prevent harm, but why it didn’t escalate concerns to someone who could.

For now, companies still position AI as a neutral platform. But as these systems become more interactive, that stance gets harder to defend — at least in the eyes of the public.

The closer AI gets to human-like interaction, the harder it is to avoid human-like responsibility.

Thanks for reading AIport. Until next Monday — by then, AI will almost certainly have replaced another job and created two more.