Bay Area engineers work to incorporate AI into smart hardware gadgets.

In a Fort Mason workshop with sweeping views of the Golden Gate Bridge, most tinkerers inside rarely peek out to take in the majestic surroundings.

Instead, on the second floor of tech incubator Founders Inc., more than a dozen young inventors sit splicing lines of code or wires into circuit boards in hopes of creating the next great electronic gadget powered by artificial intelligence.

Scattered around the wire-filled lab, industrial rigs and battery components are devices in various stages of creation or repair, ranging from smart glasses to a handheld device for blind wearers to understand the world. is that which will help humans navigate the conditions of the Day of Judgment.

Beyond the world of software chatbots that interact like humans, answer questions or create images out of thin air, these experiments are an attempt to drag AI into hardware and create smart gizmos that interact with the world around them. interact with in ways never before possible.

Founders Inc., which spots and pledges funds for innovation and startups, invited more than a dozen tinkerers from around the world to spend six weeks in its airy Fort Mason hardware lab starting in mid-April. The best ideas stand to receive six-figure seed capital investment.

It's no wonder entrepreneurs and financiers are chasing the opportunity. Market analyst Precedence Research estimates that the market for AI-enabled electronics has already reached more than $50 billion this year — including uses such as computer chips, storage and networking, across all end markets — and will grow to about could reach $500 billion.

Some early attempts at AI-powered consumer devices, such as the Ai Pin from San Francisco-based Human Inc., haven't taken off — because they were brought to market too early. Founders Head of Capital Safwan Khan is looking for innovations that last.

“It's not just about building hardware,” he said. Residency is urging builders to learn from the smartest people around them to accelerate problem solving.

One of the inventors still working out the kinks with his machine was Shobha Mittal, who wore a built-in camera on the bridge of his nose above a broad smile with easy-to-choose dark glasses.

Paired with AI programs like OpenAI's ChatGPT and Anthropic's Claude, Mittal's camera can scan images and text and narrate them into a computerized voice in about three seconds. “The idea is to help blind people become more independent,” using glasses to read menus and describe their surroundings, Mittal said.

But getting complex AI software to play nice with evolving hardware can be extremely difficult.

Case in point: When Mittal demonstrated his latest software update to an assembled congregation recently — a Friday ritual to highlight the week's progress — the results were mixed.

Asked to identify the pink stuffed elephant in front of him through the glass, Mittal got no response at first. Eventually the program denied the existence of the elephant, instead describing a forest scene that was completely unrelated to the hardware lab.

A slightly embarrassed Mittal said that shaking his head created a blurry image that confused the program. Later, the glasses described the lab with extraordinary accuracy, identifying it in detail as a research or industrial area.

Founders Lab and Mittal aren't the only ones working on smart glasses. Engineers at the SRI International Research Center in Palo Alto have poured millions in a government grant into head-mounted AI technology that could one day help untrained soldiers heal a battlefield wound or repair a knocked-out engine. Is. The goal is to create a headset that uses AI to communicate with users using visual displays and spoken instructions.

Those plans are one way. For now it can help you make lunch.

Wearing a Microsoft HoloLens headset connected to a bank of computers, SRI researcher Bob Price recently demonstrated how the device can teach someone to roll a tortilla.

The headset requires the user to press virtual buttons in the air that are presented on the heads-up display. As a robotic voice guides the user through the steps, built-in cameras can reliably detect when a tortilla has been smeared with peanut butter and jelly, rolled and cut.

The hardest part is training the system to recognize errors and respond appropriately, said project supervisor Charles Ortiz. Even a user's hand getting in the way can confuse the machine.

Another project is taking place a few miles away in a lab at Stanford University. A team led by Gordon Weitzstein, an associate professor of electrical engineering who runs the school's computational imaging lab, is trying to perfect glasses that display 3D imagery on glasses that are lighter and sleeker than HoloLens.

Wetzstein used algorithms and laser etching to model how light from a side-mounted source bounces around the lens and into the user's eye, creating a 3D hologram effect.

Rather than headsets projecting flat images, Weitzstein hopes future AI glasses will look like his lightweight device, which uses tricks of light to create augmented reality.

“The convergence of AI and hardware really enables new things,” he said.

While glasses and headsets are an obvious early application for AI hardware, tinkerers back at Founders Lab are pursuing even more esoteric opportunities.

Engineer Adam Cohen Hillel is perfecting a product he calls the Arc, a handheld device intended to help users survive Armageddon.

On the assumption that networks and communications will be destroyed, Ark is preloaded with maps and uses Meta's Llama 3 AI engine to tell users where to find water or shelter. or how to give first aid or start a fire.

Right now, it's a screen on a circuit board with a stylus attached to it that Cohen Hillel jokes resembles a time bomb.

During his demonstration to the group, the map moved smoothly, but the voice chat function went down. Cohen Hillel hopes to lose the stylus in the future so the device can operate entirely on voice commands and add solar panels to power the arc when the grid goes out. He is also planning to add a special case so that it can withstand an electromagnetic pulse from a nuclear bomb.

“Crushing that much data on a small board is the hardest part,” he said. So it's working pretty fast.

But when they work, devices like Mittal and Cohen Hillel offer a glimpse of the future.

Hubert Thiblot, general partner at Founders Inc., said reaching a “magical moment” where doing something is the easy part. It's very difficult to get something to work every time.

Reach Chase DiFeliciantonio: chase.difeliciantonio@sfchronicle.com; Twitter: ChaseDiFelice