Meta Connect's AI Live Demonstration Went Amusingly Wrong for Everyone Involved
"I think the Wi-Fi might be messed up..."
Published Sept. 18 2025, 6:15 p.m. ET

Presenting a new product can be hard for any major technology company. Since the programming involved is new, there are plenty of problems that can arise when a product is taken into the field for the first time.
That's why most companies prefer to record their presentations, giving developers the opportunity to control what the public gets to see. A live presentation is always the riskier choice.
Nevertheless, Mark Zuckerberg wanted to impress the world with Meta Connect. The glasses powered by artificial intelligence are meant to change the game for users, giving people the opportunity to see the world in an entirely different way.
What went wrong during the Meta Connect presentation? Here's what we know about the situation that forced Mark to awkwardly chuckle during a major technology event.

What happened at the Meta Connect presentation?
Meta wants to ensure that the public embraces artificial intelligence and the changes it brings for their products. However, the company never expected the Meta Connect presentation to go south.
Mark attempted to show those in attendance just how reliable the glasses can be. A live video feed showed someone attempting to use the technology to follow a cooking recipe. The goal was to follow the instructions to mix a sauce.
The person hosting the live feed asked the artificial intelligence model to provide the first instructions to make the sauce. ZDNET shared the TikTok video in which the Meta team can be seen trying to make the technology work in real time.
The system failed to give the instructions for making the sauce, with the artificial intelligence model not even understanding what the presenter was looking to achieve.
Mark leads a technology empire that is worth millions of dollars. The entrepreneur knows that he has to do everything in his power to keep the brand's reputation. There was only one way to walk out of the situation without further consequences.
Mark charismatically brushed the incident off, telling the amused audience that situations like these are why the team didn't recommend doing a live presentation for Meta Connect before the presenter blamed the wi-fi connection.
Is artificial intelligence reliable?
Countless people from different backgrounds embrace artificial intelligence as part of their lives as time goes by. Through Meta products, Chat GPT, or other similar language models, kids are comfortable with letting artificial intelligence do their homework for them.
The technology grows, but due to how recently its explosion in the market took place, artificial intelligence systems constantly present flaws.

Time Magazine reports that the constant use of Chat GPT and similar language simulators causes damage to the brain, negatively affecting the user's cognitive abilities. The study, performed by the Massachussetts Institute of Technology, arrived to the conclusion that this technology may erode critical thinking skills.
Chat GPT is normalized at an alarming rate when it comes to users whose brains are still developing.
The system's failure during the Meta Connect presentation can be linked to the study about how the use of artificial intelligence affects the human brain. If the technology itself isn't ready to be used responsibly, companies can't ensure its efficiency to customers.