ROBINHOOD HANDS YOUR PORTFOLIO TO AN AI AND WAVES GOODBYE
The company best known for making Wall Street accessible to everyday people has now taken the next logical step: it is getting out of the way entirely.
Robinhood announced this week that it will let artificial intelligence agents autonomously trade stocks on your behalf. You set up a dedicated “agentic trading” account, load it with whatever money you are comfortable handing to a machine, connect an AI to Robinhood’s API, and the robot handles the rest. It reads your portfolio, assesses sector exposure, browses analyst notes, and then actually executes the trades. Not suggests them. Executes them. To sweeten the deal, Robinhood is also launching a virtual credit card specifically for AI agents. Set a monthly spending limit, decide whether you want the AI to ask permission every time or just spend freely up to the cap, and away it goes.
Now, before you either panic or start daydreaming about passive income while you sleep, let us be clear about what this actually is and is not. The feature is launching in beta. It only handles stock trading for now, with options, crypto, and futures coming later. The dedicated account is separate from your main portfolio, so at minimum the robot cannot blow your entire retirement fund in one afternoon unless you specifically enabled that. You also get notifications whenever trades happen and can cut the AI off instantly.
But here is the thing that nobody is talking about: this is not a cautious experiment. This is a brokerage used by tens of millions of regular people, many of whom got into investing through its app, handing them the ability to connect any AI agent they want to their money. Not an institutional investor with a compliance team. Not a hedge fund with risk controls. Your neighbor who watches YouTube finance videos and bought a meme stock in 2021.
The history of retail investing is littered with products that felt like democratization and turned out to be sophisticated ways to lose money faster. Day trading apps, crypto exchanges, zero-commission options platforms. They all made the tools of speculation more accessible without making people any better at speculation. This is the next version of that story, except now the agent doing the speculation never sleeps, never second-guesses itself emotionally, and operates at machine speed.
There is something almost poetic about it. The original premise of Robinhood was that ordinary people deserved the same tools as Wall Street. Now the tool is: let the machine do it. Which is, ironically, exactly what Wall Street already does. Maybe the democratization was always going to end here. Whether that is inspiring or terrifying probably depends on how your portfolio performed last quarter.
Read the full story: Robinhood now lets your AI agents trade stocks (TechCrunch)
META WANTS TO WIRE YOUR SHIRT AND LISTEN TO EVERYTHING YOU SAY
Mark Zuckerberg has a lot of hardware ambitions. Smart glasses. Mixed reality headsets. A division that burns four billion dollars a quarter to fund these ambitions. And now, apparently, a pendant.
Meta is developing an AI-powered pendant, according to a company memo seen by The Information. The device would build on Limitless, an AI startup Meta acquired at the end of last year, which made exactly this kind of gadget: a small clip-on device that attaches to your shirt or hangs from your neck and records your conversations, then turns them into searchable, summarized memory. The company said the acquisition would let it “accelerate work to build AI-enabled wearables.” It was not kidding.
The idea is not new. Humane launched an AI pin that crashed and burned so spectacularly that HP bought its corpse for $116 million just to shut it down. A startup called Friend raised millions to make a pendant that was essentially a very expensive robot that pretended to be your pal. OpenAI and Jony Ive have been working on some kind of AI device for over a year with minimal updates and maximum mystery. And yet everyone keeps trying. Because the underlying idea is genuinely compelling even if no one has gotten the execution right yet.
The phone is the computer we carry everywhere, but you cannot talk to it naturally while it sits in your pocket. A wearable that passively captures your day, remembers what you said in meetings, recalls conversations from months ago, and feeds all of it into an AI that can actually answer questions about your life — that is a product people would buy, if it worked, and if it did not creep them out completely.
The “creep them out” part is where every previous attempt has fallen apart. Nobody wants to walk into a meeting and announce they are wearing a recording device. Nobody wants to tell their spouse that the conversation they are having is being transcribed to a server somewhere. The social friction of always-on recording has not been solved by any previous product, and there is no particular reason to believe Meta has figured it out either.
What Meta has that the previous attempts did not is distribution. Ray-Ban Meta glasses are a genuine hit. Meta has shown it can ship AI hardware that people actually want to wear in public. The memo also mentions plans to expand AI glasses and launch a “Wearables for Work” business subscription. Reality Labs lost $4 billion last quarter alone funding this vision. If anyone can normalize a recording pendant, it is a company that already got millions of people to walk around with a camera on their face. Whether that normalization is a good thing is a conversation worth having before the product ships, not after.
Read the full story: Meta is reportedly developing an AI pendant (TechCrunch)
NVIDIA TAKES OVER YOUR LAPTOP AND JENSEN HUANG CONQUERS ANOTHER MARKET
The company that became the most valuable in the world by powering AI in data centers is now coming for your laptop, and it is bringing Microsoft and Dell along for the ride.
Nvidia is expected to unveil the first Windows computers using its chips as the main processor at Computex in Taiwan and Microsoft’s Build developer conference in San Francisco, both happening next week. Microsoft’s own Surface laptops will be among the first. Dell will follow. The era of the Nvidia PC is apparently here, and it arrived faster than most people expected.
This is a bigger deal than it sounds. Intel and AMD have been the only two options for Windows chips for so long that it feels like a law of nature. Qualcomm tried to change that with its Arm-based processors a couple of years ago, offering excellent battery life and solid AI performance, and got a lukewarm reception. The problem was simple: developers and businesses did not see enough reason to build specifically for a chip that ran a slightly different version of Windows. The ecosystem is stubborn.
Nvidia changes the calculation. When the company that makes the chips inside every serious AI system on the planet says it is bringing that architecture to your desk and your bag, the developer community has a compelling reason to pay attention. If AI workloads run dramatically better on Nvidia hardware locally — and given the last five years of GPU dominance, that seems likely — then Microsoft finally has a real story to tell about what Windows AI can do that its competitors cannot match.
Microsoft’s first AI PC push, the Copilot+ PC with its Recall feature, was a mess. Recall was a privacy disaster waiting to happen, the launch was delayed, and the whole initiative felt like rebranding familiar hardware. The promise was local AI, but the AI running locally was not impressive enough to justify buying new hardware. Nvidia changes that premise. The company will also be partnering with Microsoft on software that makes it easier for AI agents to operate directly on your machine, without sending data to the cloud. That is a privacy pitch. That is a business pitch. That is the AI PC story that might actually land.
Nvidia teased “a new era of PC” on X and pointed to coordinates in Taiwan. Microsoft’s Windows chief vaguephrased about “something new coming for developers.” Jensen Huang is now in your laptop. The man just cannot stop.
Read the full story: Scoop: First Windows PCs powered by Nvidia chips to debut next week (Axios)
ERIC TRUMP JOINS ROBOT ARMY STARTUP, TESTS MACHINES IN UKRAINE, WANTS PENTAGON CONTRACTS
This one has everything. Trump family involvement, battlefield testing in an active war zone, humanoid robots, congressional scrutiny, and a startup nobody heard of six months ago aiming to arm the U.S. military in the next year and a half. Somewhere a cable news producer is weeping with joy.
Foundation Robotics Labs, a company aiming to deploy humanoid robots in the U.S. military, has brought on Eric Trump as its chief strategy advisor. The company’s robot is called the Phantom MK-1. Two units have already been sent to Ukraine for pilot tests focused on logistics in hazardous areas. The tests are backed by the U.S. government. The company wants to have its robots operational in the American military within 12 to 18 months.
Senator Elizabeth Warren has already written to the company raising concerns about Eric Trump’s role, noting the obvious fact that a startup with a Trump family member in leadership seeking Pentagon contracts while a Trump runs the executive branch raises some questions that deserve answers. Questions like: is this contract awarded on merit, or on dinner table conversation? The company has not publicly responded.
Now, setting aside the politics for a moment, because the underlying story matters regardless of who is involved. The military is seriously exploring how to use humanoid robots in combat support roles, and Ukraine has become the proving ground for autonomous systems in ways no previous conflict has. Sending robots into hazardous logistics situations — moving ammunition, crossing terrain that would get a human killed, operating in areas too dangerous for soldiers — is not science fiction. It is a procurement decision happening right now.
The question is not whether the military will use robots. It will. The question is which robots, under what oversight, on what timeline, and who decides. Normally that process involves competitive bids, independent technical evaluation, and accountability to Congress. Normally the answer is not: a startup whose chief strategy advisor is the son of the sitting president, with battlefield tests already underway, pushing a 12 to 18 month deployment window.
To be fair, the company might have genuinely good technology. The Phantom MK-1 might be everything they claim it is. The Ukraine tests might have produced solid results. But the people asking hard questions about this arrangement deserve answers before the robots ship, not after the contracts are signed. That is not a partisan position. It is just how procurement is supposed to work.
Read the full story: This Trump-linked startup plans to put humanoid robots in the military (CNBC)