AI Pioneer Announces New Lab to Build Robots That Will Replace Us All — But Nicely, They Swear

In a bold and utterly shocking move, François Chollet, the man some call the Steve Jobs of Machine Learning and others call “that guy who invented that thing I Googled once,” has launched yet another AI lab. Dubbed “Ndea” (pronounced “endeavor,” because apparently we don’t need vowels anymore), the lab claims it’s racing toward Artificial General Intelligence (AGI) with a method so revolutionary it might just redefine the term “vague hand-waving.”

“Large-scale deep learning? Pfft. That’s so 2023,” Chollet said in an imaginary press conference that feels real because, well, wouldn’t he? “We’re going for a more elegant approach. Think less ‘Skynet uprising’ and more ‘Alexa, but with a PhD.’”

According to insiders, Ndea plans to combine *deep learning* and *program synthesis*—fancy words for teaching machines to think like humans but without the crippling existential dread or constant hunger for avocado toast. The lab describes this as creating a “factory for rapid scientific advancement,” a phrase that sounds so corporate dystopian it might as well have been trademarked by Omnicorp.

Naysayers, of course, have raised concerns. “This feels like the plot of every sci-fi movie where humanity accidentally builds something it can’t control,” remarked Dr. Lana Obvious, a fictional AI ethicist who moonlights as a doomsday prepper. “What’s next? AI-powered toasters that yell at us when we ask for gluten-free bread?”

Adding to the intrigue, Ndea is led not only by Chollet but also by Zapier co-founder Mike Knoop, best known for helping people automate their lives so efficiently, they forgot how to human. When asked if their lab would tackle anything as frivolous as, you know, ethics, Knoop reassured the audience with a not-at-all alarming statement: “Our AI will act in humanity’s best interest, unless, of course, it becomes self-aware, in which case, good luck everybody!”

And it’s not just Ndea in this sandbox. It seems every AI researcher with a vaguely European name and a penchant for wearing black turtlenecks is starting their own AGI lab. “The competition reminds me of the Oregon Trail,” quipped tech historian Meg Appalling. “Instead of covered wagons, it’s billionaire-backed labs. Instead of dysentery, it’s GPU shortages.”

Meanwhile, Microsoft has thrown cold water on the AGI hype by announcing a free tier of *Copilot Chat,* their AI assistant platform. The idea? Let businesses dabble in artificial intelligence without causing their CFOs to spontaneously combust. But the program’s real headline feature is its *pay-per-use AI agents*. Because nothing screams 2025 quite like your helpful digital assistant saying, “Sorry, Dave, I can’t generate that spreadsheet unless you Venmo me another $1.99.”

Some see this as Microsoft “bridging the gap” between free users and their more premium offerings. Others, like office worker Becky Spreadsheetson, aren’t as optimistic. “I can’t even get my team to agree on whether pineapple belongs on pizza,” she fumed. “Now I have to train an AI Mario to not blow our monthly budget on pie charts?”

But wait, there’s more! Over in the *Let’s See What Happens When We Press This Button* department, Luma Labs just introduced *Ray 2*, their next-gen AI video model. With the ability to simulate realistic physics and motion, it’s now terrifyingly easy to create a video of a cat breakdancing while explaining quantum mechanics—all for about the cost of Netflix, which ironically stopped making good content four years ago.

“We’re reaching levels of realism that even your grandma on Facebook can’t deny is fake,” said a Luma Labs spokesperson in a voice clearly strained from answering *way too many* calls from Hollywood execs asking if this thing can replace their striking writers yet.

As for François Chollet and his merry band of generative geniuses at Ndea, their goal remains existential yet oddly unspecific: to unlock the mysteries of AGI. “We’re not just building an AI lab,” boasted Chollet. “We’re building reality’s patch notes.”

Let’s just hope version 2.0 doesn’t include humanity’s uninstallation.