AI is getting more powerful.

But this morning, the bigger question isn't what AI can do.

It's how fast we should let it do it.

Three stories from the weekend show why that debate is getting harder to ignore.

David Robinson, who worked on safety at OpenAI and helped with safety reports for major model launches, has resigned and publicly criticised the company's approach.

His argument is simple but serious: as AI systems become more powerful, companies can't rely on a “build it, see what happens, fix it later” approach.

OpenAI says it is continuing to strengthen its safety and monitoring work.

This isn't just a story about one employee leaving.

It's about something much bigger.

We're giving AI more ability to act, make decisions and operate on our behalf.

So the question becomes:

Who is making sure the technology stays under control?

The more powerful AI becomes, the less room there is for learning from mistakes after the fact.

OpenAI CEO Sam Altman has argued that the benefits of AI are significant enough that society will have to accept some level of risk as the technology develops.

Altman favours keeping AI broadly accessible and avoiding overly restrictive regulation.

That puts him on one side of an increasingly important debate, with other AI companies taking more cautious positions.

Now we're looking at the real debate behind the headlines:

How much risk is acceptable in exchange for faster AI progress?

There isn't an easy answer.

Too much caution could slow useful technology.

Too little caution could create problems that are very difficult to reverse.

And this isn't only a Silicon Valley discussion anymore.

It affects businesses, workers, consumers and governments.

The AI boom isn't only about clever software.

It's creating an enormous demand for data centres, chips, electricity and computing infrastructure.

Reuters reports that global data-centre spending could eventually exceed $30 trillion by 2050, while major AI companies are committing hundreds of billions of dollars to infrastructure.

The big question is whether AI's productivity and revenue will grow quickly enough to justify that enormous investment.

We've spent a lot of time asking:

“What can AI do?”

Now there's another question we need to ask:

“Can the economics keep up with the hype?”

The AI race is becoming one of the biggest technology investments in history.

And eventually, somebody has to make the numbers work.

Three different stories.

One bigger question:

The technology is moving quickly.

The debate around safety, regulation and money is trying to keep up.

And that's a story worth watching.

☑️ An OpenAI safety employee has resigned and raised concerns about the speed of AI development.

☑️ Sam Altman says society will have to accept some AI risk to capture its benefits.

☑️ The global AI infrastructure race could involve extraordinary levels of spending.

☑️ The next stage of the AI debate isn't only about capability — it's about control, risk and whether the economics work.

That's your OneClick.

See you tomorrow.