AI that never leaves your phone

On-device AI is the practice of running an artificial intelligence model on the hardware in your hand rather than in a cloud data center. When your phone transcribes a voice memo, filters a photo, or autocompletes a message without a network connection, that is on-device AI at work. The model, the data, and the answer all stay on the device, and that one fact changes the economics of computing.

George Gilder has built his latest pitch from the George Gilder Report around this shift, calling it “Ambient AI.” The promo argues that AI is about to move out of the data center and into the devices themselves, and it teases an “early stake in a $6 stock” ahead of a “$1 trillion wealth explosion.” It is the same on-device trend the rest of the industry has been pursuing for years, given a fresh label.

Why the device is winning

Three forces are pulling AI onto devices. Latency is the first: a split-second response cannot wait on a network round trip. Privacy is the second: data that never leaves the device cannot be intercepted or resold, which matters for consumers and for defense customers alike. Cost is the third: every inference run locally is one a company does not pay a cloud provider to compute.

These are the reasons Qualcomm has spent years building neural processing units into its mobile chips and Apple ships a dedicated neural engine in every recent iPhone. The shift is not speculative; it is already the default for a large share of everyday AI. We map the investable side in our edge AI stocks explainer.

The company behind the pitch

The stock the Ambient AI promo points to is QuickLogic (QUIK), a fabless semiconductor firm in San Jose with about 51 employees and a market value near $192 million. QuickLogic’s technology is the embedded FPGA, or eFPGA: instead of selling a standalone programmable chip, it licenses the intellectual property so a customer can integrate a small, reconfigurable logic fabric directly into their own system-on-chip. That approach uses less power and less board space, exactly what a battery-powered device wants.

QuickLogic has been public since 1999 and a Gilder recommendation since December 2019, and its customer base spans aerospace and defense, industrial infrastructure, and edge computing, consistent with the promo’s mention of a U.S. military deal for Ambient AI chips in next-generation weapons. We go deeper on the business in our QuickLogic Corporation explainer.

The fine print

The headline numbers need the same skeptical pass the rest of this promo deserves. “40,000 times smaller than NVIDIA” is a size contrast, not a valuation argument: at a $192 million market cap it would put NVIDIA near $7.7 trillion. The “$1 trillion wealth explosion” is the entire on-device AI market, not QuickLogic’s slice of it. And the “$6 stock” framing is stale, since the shares were around $8 to $9 when the ad was re-teased in March 2026 and closed at $10.61 on September 2, 2026.

The execution risk is concentration. QuickLogic’s licensing model means its fortunes hinge on a relatively small number of design wins against competitors like AMD and Lattice Semiconductor, which field far larger engineering teams. Gilder’s own letter shows the range: Cloudflare, up more than 1,000%, beside Inseego, down about 98%. We looked at an earlier Gilder pitch in our Trillion Dollar Triangle teardown.

What the big platforms are already doing

The on-device shift is visible in what the platform giants ship. Qualcomm has spent years building neural processing units into its mobile chips and marketing them around on-device AI. Apple ships a dedicated neural engine in every recent iPhone, and its software runs models locally for everything from photo search to text prediction. The point for investors is that on-device AI is already a shipped, mainstream feature rather than a speculative future. That is what makes the underlying theme credible, and it is also why the small silicon supplier attached to it has to be judged against giants already competing for the same sockets.

The honest read

On-device AI is a genuine, already-arrived shift, and QuickLogic is a real, if tiny, participant in the low-power silicon that enables it. The theme is sound, but the specific stock is the risky part, and the “$6” entry point the promo dangles is gone. Evaluate the company on its fundamentals and its competitive position, not on the strength of a trillion-dollar headline.

In short, on-device AI is not a prediction about the future; it is a description of the present. The open question is only which company captures the most value from it.

NewsletterVetter is an independent publication. We receive compensation from some of the services we review through affiliate links. Nothing on this site is investment advice. Always do your own research.