This Week in AI: A Step Closer to AI Robots at Home

The tech industry has reached another "the future is now" moment with the announcement of a humanoid AI household robot called Neo. The promise is plucked straight out of sci-fi: A mechanical maid that does our chores and takes care of our homes.

It won't be cheap, though. The startup, 1X, is taking orders at a $200 deposit, with the device costing $500 a month or $20,000 to fully own when it arrives sometime next year. It'll also come in a few different shades: cream, light gray, and dark gray.

"My chores: I just leave and come back and they're done," the company says in its promo video, showing the robot folding laundry, dancing to music, and cleaning dishes. 1X says it designed Neo to look and feel safe, with a relatively lightweight body at 66lbs, and purposely slow "low energy motions." And it's built on NVIDIA technology, of course.

Right now, the AI robot can't actually do much. In videos, it appeared to only have solved opening doors and collecting an empty cup out of your hand. But using remote control through VR, the company's engineers showed a future they imagine where Neo is performing all sorts of household chores.

Of course, you might wonder when these features will work on their own, without remote controls. The answer, to some degree, is we don’t know.

"I didn’t see Neo do anything autonomously, although the company did share a video of Neo opening a door on its own," Joanna Stern wrote for The Wall Street Journal, after discussing how the 1X may need to have employees take over and "teleoperate" the robot to perform some tasks. 1X CEO Bernt Børnich told her Neo "will do most of the things in your home autonomously," though it may be sloppy at first.

While 1X is certainly making a splash online, it's just the latest AI company making big promises while pitching less-capable devices today. As tech YouTuber Marques Brownlee pointed out, this has become a common theme throughout the tech industry.

Humane, a startup by former Apple engineers, was perhaps the best example when it released it’s $700 AI pin last year. The device was marketed as the next thing after smartphones, removing the need for a screen and instead allowing you to interact with and control various apps and services with your voice.

The idea is cool in theory, but the slow speed of AI systems combined with the need for a constant internet connection made for a disappointing launch. Brownlee at the time said it was the worst product he’d ever reviewed... "for now." The company was quickly sold to HP for a fraction of its earlier value.

So why release these not-quite-finished products at all?

The consensus broadly is that this is the training AI needs to work. It needs to be tested in the real world, over and over, before it can improve. It has to learn what’s right and what’s wrong behavior, just like you or I do throughout our lives.

From that view, Neo isn’t an as-yet unfulfilled promise of the future, but rather an early opportunity to help make the future happen. Quite literally, your home turns into a training ground that will improve future robots.

NVIDIA hits $5 trillion

Another day, another financial record. NVIDIA continues to be one of the most highly valued companies in the world, now worth more than $5 trillion. With a "T."

For context, that's roughly the entire German economy.

Some of you may be old enough to remember when NVIDIA was worth just $1 trillion, way back in 2023.

Since then, we've seen big moments for the company, including this year's launch of 50-series GPUs and the very recent launch of DGX Spark, a deskside AI supercomputer. 

It’s all pretty surreal, in particular because there are so many questions about what these numbers actually mean. On paper, these eye-watering values suggest the AI boom is strong. But we’ve seen this story play out before. If AI is one of the most popular things to talk about, one of the next popular things to talk about, is whether or not AI is in a bubble. 

Bill Gates, co-founder of Microsoft and probably one of the biggest beneficiaries of previous bubbles, believes that we are indeed in another one. "There are a ton of these investments that will be dead ends," he said to CNBC this week. What matters even more for the rest of us is that so much of the economy is tied up in the stock market, and businesses around the world make decisions often based on that. 

Other tech companies also talked about AI in their quarterly earnings reports. Amazon's growth for its cloud business impressed investors, with its CEO saying his company continues to see "strong demand in AI." Microsoft, too, said that it continues "to increase our investments in AI across both capital and talent to meet the massive opportunity ahead."

CNBC pronounced it a "$380 billion splurge," where Google, Microsoft, Meta and Amazon "all raised their forecasts for capital expenditures." For their part, Meta investors seem to be getting nervous about the billions of dollars the company's spent on AI so far. The company's stock dropped 11% the day after announcing its continued "aggressive" AI spending.

Qualcomm's new AI chips

One of the hottest races in tech is among companies trying to build a real, sustainable competitor to NVIDIA's chip technology, which has become a center of the AI world. Apple is trying with its "Apple Silicon" chips, though it doesn't broadly sell those to other companies. AMD too has developed AI accelerator chips that are starting to get attention, and have even led to deals with companies like OpenAI. But it's not yet on NVIDIA's scale. There are many more.

Now, Qualcomm is jumping further into the mix, announcing a series of new technologies planned for 2026 and 2027. CNBC said Qualcomm's new chips are based on the AI parts in Qualcomm's popular smartphone chips, called Hexagon neural processing units. The AI200 and AI250 chips will eventually run in liquid-cooled server racks. 

Qualcomm said one key element of its system will be that it can "mix and match" with parts from other chipmakers.

OpenAI moving toward IPO

Reuters reports that ChatGPT maker OpenAI is "laying the groundwork" for an initial public offering that could value the company at up to $1 trillion. If it succeeds, OpenAI would become one of the biggest IPOs ever.

OpenAI is reportedly considering regulatory filings "as soon as the second half of 2026," with an aim to raise at least $60 billion from the IPO.

"An IPO is not our focus, so we could not possibly have set a date," an OpenAI spokesperson told Reuters in response to its reporting. "We are building a durable business and advancing our mission so everyone benefits from AGI."

Stanford warns on AI privacy

The modern internet may already feel like a minefield of privacy issues, but now Stanford University warns that AI could be a potentially even bigger concern. A new study from researchers at Stanford's Human-Centered Artificial Intelligence program say that popular AI chatbots are storing and potentially remembering sensitive data including names, locations, and confidential details about your life. And while these AIs are not supposed to retain conversations long term, researchers say they do anyway.

"We’re seeing data such as a resume or photograph that we’ve shared or posted for one purpose being repurposed for training AI systems, often without our knowledge or consent and sometimes with direct civil rights implications," Stanford researchers wrote.

More from MC News

• Hands-on with the NVIDIA DGX Spark 
• How to Build a PC with a Hardline Water-Cooling Loop
• 3D Print a Mac Mini Monitor Mount
• The End Has Come for Windows 10: Four Tips to Make the Most of Windows 11
• Everything You Need to Know About WiFi 7
• Keyboard 101: Intro to Computer Keyboards
• Can Your PC Run OpenAI's New GPT-OSS Large Language Models?
• Fix It Yourself: Talking to iFixit on Why Repairable Tech Matters

Ian Sherr is a widely published journalist who's covered nearly every major tech company from Apple to Netflix, Facebook, Google, Microsoft, and more for CBS News, The Wall Street Journal, Reuters, and CNET.