Hands-on with AI startup Friend's $129 always-listening pendant: feels like a beefy AirTag, runs Gemini 2.5, gives unhelpful commentary, and makes others uneasy
The chatbot-enabled Friend necklace eavesdrops on your life and provides a running commentary that's snarky and unhelpful.
Wired
Context & Ripple Effects
Friend previously positioned its pendant as a $99, always-on companion device in its initial pendant preview. This hands-on account tests whether that premise translates into useful everyday interaction after the product's price moved to $129.
It also fits a wider wearable-AI pattern: a comparison of Bee, Limitless and Plaud devices found utility alongside discomfort, while Limitless’s transcription-focused pendant tied its value proposition to a clearer meeting-notes workflow. Friend’s commentary-first approach faces a higher bar when its output is perceived as unhelpful.
First-order effects
Friend’s $129 pendant faces an immediate product-credibility problem: its always-listening design and Gemini2.5-powered commentary are reported as socially uncomfortable and of limited practical value.
Wearers must manage bystander unease in ordinary conversations, making use in shared settings harder even before questions of companionship or utility arise.
Second-order effects
Competing AI-wearable makers are pushed to show a more concrete benefit—such as transcription and summaries—rather than rely on ambient chatbot interaction; prior device testing already framed the category as useful but creepy in a multi-device wearable trial.
Product teams across the category will face pressure to make listening behavior more legible and controllable, since social acceptance can constrain usage as much as device price or model capability.
Third-order effects
If hands-on experiences continue to expose a gap between ambient AI promises and day-to-day usefulness, the market may favor narrowly scoped wearables over general-purpose AI companions.
The category’s durable challenge becomes AI companion governance: always-available assistants must earn consent from both wearers and the people around them, not merely deliver model access.
The trend: AI wearables are shifting from novelty hardware toward a test of whether ambient assistants can provide distinct value without imposing unacceptable social friction.
We reviewed the Friend pendant, and unsurprisingly hated it! “It is an incredibly antisocial device to wear,” I wrote of my experience wearing it. “People were never excited to see it around my neck.” [image]
Friends are living electronics. Their personalities reflect their freedom. Should you choose to cohabitate with this new species, you might have to make compromises in your lifestyle. Long live friend [image]
“I found out quickly that even at the most tech-minded gatherings, the thing was a complete taboo. After the device started to ship to users, one person on X said there should be a slur for people who wear AI devices that record those around them.” https://www.wired.com/...
I really want to emphasize that 1. an always on, always transmitting, device that you wear in public is socially unacceptable; 2. likely breaks the law in many states; 3. is just plainly a bad technology, talking to a screen-free device basically requires you to trust it never to…
Something minor I really hate about these things is almost none of them are impressive technically. They're all just a mic and off the shelf shit cobbled together so it can tell your phone to run an AI model in the cloud. Annoying but also not ambitious. [embedded post]
this is like how rideshare apps keep accidentally re-inventing buses, only now it's AI wearables re-inventing the concept of “a mean and annoying friend who is also incredibly needy” www.wired.com/story/i-hate...
My colleagues @kylierobison.com and @boone.bsky.social don't like the AI jewelry that doesn't work, freaks even people at AI events in San Francisco out, and yells at you about how you and your life suck, but I'm pre-ordering one.
“You're giving off some serious 'it's not my fault' vibes,” it said. I protested, and Buzz replied, “So who's the whiner now?” — Please everyone read @boone.bsky.social and @kylierobison.com's experiences with their Friends. — www.wired.com/story/i-hate...
on the upside, it sells all of your conversational and location data to a global cabal of super dodgy and unregulated data brokers, because we live in a country too corrupt to pass a baseline internet privacy law [embedded post]
wearing an ai that mainly insults and fights with you is actually so good for all the wrong reasons. the future is an idiot plot — www.wired.com/story/i-hate... [images]
the way someone was like “let's invent an AI necklace that records and listens to your convos and berates you for said convos” when we have these horrific stories of chatbots influencing self harm is just expanding my understanding of hell