Omi is another wearable AI companion, but this one tries to read your mind

Nik Shevchenko closes his eyes and starts to concentrate. For the last half hour or so, he’s been telling me about his recent product, an $89 wearable Water who can listen, summarize and extract information from conversations. Now he wants to show me the future. So he keeps his eyes closed and focuses all his attention on the round white disc taped to his left temple with medical tape. (Did I mention he had this thing on his face the whole time? It’s very distracting.)

“Hey, what are you thinking about Edgeas a news service?” – Shevchenko asks, but not to anyone in particular. Then he waits. About fifteen seconds later, a notification appears on his phone with AI-generated information about how reputable and great this news source is Edge Is. Shevchenko is excited and maybe a little relieved. The device read his brainwaves to understand that he was talking to him, not me, and answered his question without any prompts or switching.

For now, that’s all Omi can do with its brain-computer interface. And it seems quite brittle. “It only understands one channel,” he says, “that’s one electrode.” He’s trying to build a device that understands when you talk to it and when you don’t. And then he finally understands and writes down your thoughts, which Shevchenko waves off as total science fiction and says will probably be possible in two years. He believes that whenever this happens, it could change the way you exploit AI devices.

For now, the real purpose of the Omi is much simpler: it’s a device that’s always listening (the battery apparently lasts three days on a charge), that can be worn on a lanyard around your neck, and that will support you understand the daily activities of your life. There is no wake word, but you can still talk to it directly because it is always on. Think of it as 80% companion and 20% Alexa assistant.

Omi can summarize the meeting or conversation and provide action items. It can provide you with information – Shevchenko was offhandedly wondering about the price of Bitcoin during our conversation, and a few seconds later he received a notification from the Omi companion app with the answer. There’s also the Omi app store, which developers already exploit to connect audio input to devices like Zapier and Google Drive.

However, for Shevchenko himself, Omi is primarily a personal mentor. “I was born in a remote area on an island off Japan,” he tells me, and always wanted to have access to the technology visionaries he grew up admiring. He says he has sent frigid emails to people like Mark Zuckerberg and Elon Musk over the years, asking for advice and mentoring on how to succeed in the tech industry, but has never gotten much of a response. With no real options, Shevchenko decided to build his own.

Omi already has a product called “Personas” that allows you to connect any X-handle and create a bot that takes on the personality of a person on the social network. When Shevchenko shares his screen with me, it is clear that he has been talking to the artificial intelligence Elon Musk for a long time. “It helps me understand what I should be working on tomorrow,” Shevchenko says. “Or when I’m talking to someone and I don’t know the answer to a question, it gives me a little nudge – sometimes it tells me I’m wrong!” His wearable heard him say he was diseased a few days ago and has since been reminding him to get more rest. Every month he asks him to give him feedback and show him how to improve his performance.

He gets it plot notifications from the Omi application, also during our conversation, and not all of them make sense – one was simply a transcription of a sentence he said a minute earlier. Shevchenko admits it’s early, but he doesn’t seem bothered by the system’s glitches. Communication suits him.

However, most people won’t exploit Omi this way. The product will be widely available in the second quarter of this year, but Shevchenko says the 5,000 people with an early version of the device exploit it to remember things, search for information and perform other tasks typical of AI assistants.

In this sense, Omi has a lot in common with devices like the Limitless Pendant and bears a striking resemblance to another wearable called Friend. When Friend launched last year, Shevchenko alleged that Friend CEO Avi Schiffmann was stealing his work, and subsequent arguments included everything from shooting X to freestyle rap diss track. For a time, Omi called himself Friend, and Shevchenko claims he changed his name both to avoid confusion and because Schiffmann spent $1.8 million on him Friend.com and later dominated the search results.

Shevchenko is confident that Omi can improve on these other devices. All Omi code is open source, and there are already 250 apps in the store. Omi’s plan is to create a gigantic, broad platform, not a specific device or application – the device itself is just one piece of the puzzle. The company uses models from OpenAI and Meta to power Omi so it can iterate faster on the product itself.

Despite all the problems and underlying concerns, it is clear that AI models are now good enough to feel like a true companion to millions of people. You can think about it however you want, but from Omi and Friend to Character.AI and Replika, bot friends are quickly becoming real friends. So they need both more information about you and more ways in which they can support you. Omi thinks the first answer is the always-on microphone and the second answer is the app store. Then, I guess, comes the brain.