Few hours after the big reveal this morning, Humane opened its doors to a handful of press. The office is located in a nondescript building in San Francisco’s SoMa neighborhood, and houses the startup’s hardware design teams.
An office next door houses Humane’s product engineers, while the electrical engineering team operates from a third space directly across the street. The company also operates an office in New York, though the lion’s share of its 250-person staff is here in San Francisco.
Today, much of the space is taken up by a series of demo stations (with a strict no-filming policy), where various Ai Pins are laid out in various states of undress, exposing their external machinations. Before attending, however, Humane’s co-founders stand in front of a small group of chairs, flanking a flat screen that lays out the company’s vision.
CEO Bethany Bongiorno gives a brief history of the company, starting with how she met co-founder and president Imran Chaudhri on her first day at Apple. The entire history of the company is linked to their former employer. There they hunted down CTO Patrick Gates, along with about 90 other ex-Applers.
Image Credits: Brian Heater
For his part, Chaudhri frames the company’s history as one of S-curves – 15-year technology cycles that underlie and eventually give way to what follows. “The last era has plateaued,” he tells the room, declaring that the smartphone is “16 years old” — though that, too, appears to be a dig at his former employers, whose first iPhone arrived in 2007 .
He frames Humane’s first product as “a new way of thinking, a new sense of opportunity.” It’s an attempt, he adds, to “produce AI.” The in-person presentation is arguably more grounded than previous videos would have you believe. True, the statements are still grand and sweeping, framing the device worn on the lapel as the next step in a computing journey that began with room-sized mainframes, but the conversation becomes more realistic when the device is in front of us.
The cigarette lighter-sized device has a Snapdragon processor and 32GB of local storage. The camera is a 12-megapixel sensor designed for a smartphone but built into Humane’s unit. There is an accelerometer and a gyroscope and a depth and time-of-flight sensor. Like Apple products, it is designed in California and manufactured primarily in Asia.
Most of the device’s exterior is monopolized by a touchscreen that houses the majority of the built-in components and a battery that should last four or five hours on a charge. Above that, a sort of camera stick houses the aforementioned sensors, along with the laser projection system — the most visually impressive aspect of the whole affair. The camera line slopes downwards. Humane says they tested the pin on several different body types and settled on a design that can accommodate users with larger breasts.
The company also told me it tested the laser projection on a range of different skin tones to ensure it would be visible. While visually arresting, the projections are seen as a secondary feature in what is essentially a voice-first product. If, however, you’re in an environment that’s too loud or too quiet to accommodate the small upward-facing speaker that spans the top of the device, tap the touchpad and the camera goes to work looking for a hand. Once detected, it starts displaying.
Image Credits: Brian Heater
Chaudhri demonstrated the possibility during a TED talk in May. In a minute or two, a gradual call comes from Bongiorno, which the pin displays on his palm in text form. From here, he can tap his palm to accept or decline the call, with the system recognizing the motion and acting accordingly.
However, lasers can display much more. They display text from messages, which you can scroll through with a pinch motion on the same hand. They can even show rudimentary previews of the images you take, since the green laser doesn’t do the best job of highlighting the subtle intricacies of a photo.
The AI Pin has a bit of weight to it, though this is offset somewhat by the “battery booster” that ships in the box, increasing the full battery life to around nine hours. The product also comes with an egg-shaped case that adds another full charge to the product. You can put either the pin or the booster or both in and they will snap into place magnetically. Charging is done via a series of pins on the back of the device. Also in the box is a charging base for home use.
Bongiorno confirmed earlier reports of an initial batch of 100,000 units. “I think like with any hardware startup, we want to make sure we plan conservatively for all scenarios,” he says. “For us at the beginning, we really looked at what was the conservative, right and responsible thing to do in terms of demand and allow us the option if our demand goes over 100,000.”
He adds that, as of yesterday, more than 110,000 people had signed up for the wait — though that number is more an indication of curiosity than actual purchase plans, as no deposit is required. The listing is also global, while the device is only available for pre-order in the US, where it will launch sometime early next year. The initial waiting list group will be given “first access” to purchase the product.
Image Credits: Brian Heater
The heart of the device is AI. It’s one of the first hardware products to ride the current wave of excitement around genetic artificial intelligence, but it certainly won’t be the last. Sam Altman’s name has been closely associated with the startup since the day it was announced. I ask how closely Altman and OpenAI were involved in the creation of the product.
“Sam led our Series A in 2020. Imran was very clear that Sam was the Series A target and really wanted him on board,” says Bongiorno. “I think there was a lot of mutual respect and excitement about what we all thought was the future, in terms of computing. He has been an incredible advocate and supporter for us and picks up the phone whenever we need advice and guidance. We are working with the OpenAI team. Our engineering team worked together and worked closely together.”
According to Chaudhri, GPT is one of several LLMs used by the system. He also confirmed that GPT-4 will be among the cases the system uses. Ultimately, however, the exact AI systems leveraged for any given task are somewhat fuzzy by design. They are accessed on a case-by-case basis, based on the determination of the appropriate course of action by the pin.
This also applies to web-based queries. The system crawls a variety of different search engines and resources such as Wikipedia. Some will be official content partners, some will not. So far, partners are limited. There’s OpenAI and Microsoft, as well as Tidal, which serves as the system’s default music app. An example given during one of our demos was ‘play music produced by Prince’ rather than the more straightforward ‘play Prince’.
“Some of our AI is proprietary. We build our own AIs and then leverage things like GPT and models from OpenAI,” says Bongiorno. “We can add LLM and many services from other people and our goal is to be the platform for everyone and allow access to many different AI experiences and services, so the business model is structured in a way that allows us to do this. And I think we’ll be thinking about different revenue models that we can also add and different revenue streams to the platform.”
Image Credits: Brian Heater
The goal is to make the experience seamless, both in terms of what happens on the back end with LLMs and web searches and updates. The system is designed to continuously push updates and add new features in the background. It also uses additional context, including recent questions and location, using built-in GPS.
Photos are also a big piece of the puzzle. The built-in camera has an ultra-wide angle with a 120-degree field of view. There’s no autofocus in the game — instead it’s a fixed focal length. In SF office lights, at least, the photos looked solid. There’s a good bit of computational photography that happens off-device, including calculating whether the stylus is flat when taking a shot and orienting the final image accordingly.
It all still looks very early here, but it’s clear that a lot of care (and money) went into the product. Demand is probably the biggest question mark here. Has Humane really found a killer app? For smartwatch makers, health has long been that answer. But health monitoring plays a significantly reduced role here.
The product doesn’t actually make direct contact with the user’s skin, so the health metrics it can collect are limited beyond serving as a pedometer — though that feature isn’t currently supported. The biggest health-related feature right now is calorie counting, which tells you specifically how many calories and other nutritional information are in the piece of food you’re holding on camera, using an anonymous third-party food recognition platform.
Image Credits: Brian Heater
The price will definitely be an obstacle for the unproven device – $699 is basically nothing by smartphone standards, but it’s a lot to ask for a first-generation product and a new form factor. The added $24 a month doesn’t help either, though Bongiorno adds, “You get a phone number. You get unlimited talk, text and data. you get as many AI queries as you want, in addition to all our AI services. Today, we’re seeing how much excitement there is around ChatGPT, where people are already paying for access to it.”
However, if you don’t pay this month, the product is essentially a piece of paper until you start the subscription.
Before our session breaks, I ask Chaudhri how the company landed on the lapel, of all places, especially when head-worn displays have been considered the default for some time. Certainly the former employer of Apple is betting on the face with the upcoming Vision Pro.
“Contextual computing has always been thought of as something to wear on your face,” he says. “There are many problems with this. Many people wear glasses that you wear for a really precise reason. It is either to help you see or to protect your eyes. This is a very personal decision — your frame shape, your frame weight. It all goes to something that is as unique as you are. If you look at the power of context, and that’s the barrier to achieving contextual computing, there has to be another way. So we started looking at what is the piece that allows us to be much more personal? We came to the fact that we all wear clothes, so how can we decorate a device that gives us a context for our clothes?”
