🤖 I just watched AI take over my computer.

And it was terrifying. And amazing.

Last night, sitting at home on my own personal laptop, I sat frozen as an AI agent browsed websites, filled forms, and (almost) booked flights—all while I did absolutely nothing. Awestruck and flabbergasted.

I was testing a new AI browser from one of the big AI providers (still in beta). Within minutes, I asked it to:

• Find me a pancake recipe and add the ingredients to my shopping cart at my online retailer.
• Search for cheap flights and email me an itinerary for next year’s winter holiday.
• Build me a website based on web gathered research on the theme of Scratch Programming for Kids.

What happened next? My screen came alive. The AI navigated between sites, entered information, and executed tasks with surgical precision. I was just… watching.

The wake-up call: While there’s still a “human in the loop” feature, I couldn’t stop thinking – …and then what?

My instructions were pretty innocent. Yet, the more autonomous and complex systems become, the more we have to engage in second order thinking – that is to say – think about the consequences of consequences.

What happens when AI programs increasingly trigger other AI programs? When robots not only can replicate, but also improve upon themselves? We risk cascading and self-reinforcing effects we never intended or could have possibly foreseen.

What would have seemed science fiction a few years ago now looks to me to have become real possibilities.

The pace of development is breathtaking. Guardrails may struggle to keep up. Yes, the potential to solve real human problems is enormous—but we need regulators and companies to move just as fast on safety and reliability.

I have seen many ridicule AI for not being able to answer simple questions or even basic math problems obvious to us (often due to generic or poor prompting). But this way of thinking should be inverted, focusing instead what AI can do!

Because it can do more and more. Perhaps it might even do things that you do, better than you and sooner than you think.

While Nobel Prize winner and instrumental AI pioneer Geoffrey Hinton says becoming a plumber is the safest career bet against the AI disruption, I’d rather heed the famous words of Peter Parker’s Uncle Ben: ‘With great power comes great responsibility.

💭 Can you imagine where this technology will take us in five years?

I’m both excited but also a little concerned. What’s your take?