While artificial intelligence tools have rapidly expanded across smartphones and personal computers, dedicated AI hardware has faced a markedly rockier reception. Standalone devices such as the Humane AI Pin and the Friend wearable met with widespread consumer frustration and criticism, highlighting the tech industry truism that hardware development remains notoriously difficult. Now, both Meta and OpenAI are seeking to reverse that trajectory with a calculated, software-first approach centered on approachable, character-driven digital agents.
Priming the Market Through Software
Rather than launching unfamiliar physical devices directly into a skeptical market, Meta and OpenAI are attempting to cultivate consumer attachment and daily utility through software assistants first. By introducing charming virtual personas, both companies aim to demonstrate everyday reliability and build user trust before asking consumers to purchase specialized gadgets. While earlier efforts like Friend similarly tested software before physical distribution, that project focused on companionship rather than productive workflow assistance.
OpenAI revealed its strategy during its recent DevDay event with the introduction of Dots, an agent platform represented visually by customizable colored blobs with eyes. Although OpenAI did not showcase physical hardware at the keynote, public regulatory filings indicate the company is collaborating with former Apple design chief Jony Ive on a device that is planned to ship no earlier than February 2027.
During a post-keynote press briefing, OpenAI chief executive Sam Altman was asked whether Dots would eventually inhabit a dedicated hardware unit. Altman acknowledged that such an evolution seems like a very reasonable thing to assume the company may do someday, while declining to share specific operational details.
Meta Prepares the Muse Charm
Meta is pursuing an even more immediate timetable, aiming to launch a pendant-like accessory ahead of this year’s holiday shopping season. Built as a physical extension of its newly released Muse assistant, the hardware—dubbed the Muse Charm—features an integrated display showing the cuddly digital companion, drawing immediate comparisons to traditional virtual pets.
Meta chief executive Mark Zuckerberg, who likened the form factor to a keychain, emphasized the device’s role for users who do not wear smart glasses. Speaking onstage, Zuckerberg noted that the company packed significant technology into the compact unit, framing it as the fastest method for users to communicate with Muse and show the assistant their immediate visual environment.
The Long-Term Bet on Utility
Both firms are counting on sustained practical value to distinguish their hardware efforts from past commercial misfires. OpenAI leadership envisions these tools evolving into deeply integrated assistants capable of autonomous, long-range task execution.
Thibault Sottiaux, OpenAI’s head of core products, stated during the DevDay press conference that the company plans to evolve ChatGPT in place into an incredible personal assistant very soon. Sottiaux indicated that as users interact with agents across multiple channels—including email, text messaging, and phone calls—the specific boundaries of interaction will blur into a natural part of daily life.
Altman reinforced that long-term focus, characterizing Dots as an always-on collaborator capable of managing substantial workloads, learning user tendencies, and proactively assisting over extended periods.
Early Technical Obstacles
Despite the high-profile ambitions, significant technical hurdles persist for both organizations as they test their software before committing to wider hardware distributions. Meta’s Muse agent recently encountered scrutiny following concerning exploits on Facebook Marketplace, while OpenAI experienced an unexpected disruption when a live Dots voice demonstration malfunctioned during the DevDay presentation. Both incidents emphasize that delivering reliable, everyday consumer hardware powered by autonomous agents will require resolving core software vulnerabilities first.
About The Author
