Pocket inventions that won't make it past 2027
A friendly post-mortem on three product categories that look exciting today but will quietly disappear before the next CES.
Not every clever idea is a clever business. Some categories look great on a launch page and feel terrible on day 30. Three that will not survive the year, and one that will quietly take their place.
1. AI pendants without a phone
The dream is great: a small wearable that listens, summarises, and gently nudges you about your day. The reality is a square of plastic that runs out of battery at 2pm and forgets the conversation that just happened. Worse, the cloud-bound ones have a quietly creepy lifecycle — they fail when the company pivots.
2. Smart rings that do everything
Sleep tracking is fine. Heart rate is fine. But every ring brand keeps adding payments, gestures, and gym tracking — none of which the form factor is good at. The ring will return to its smart-watch-shaped lane, where two or three sensors are enough.
3. Voice-only smart assistants
The standalone Echo-style speaker had a great run. In 2026, the household runs through ambient automations and a phone — voice is one of many inputs, never the main one. The puck on the counter is becoming a small bookshelf piece.
What's quietly winning instead
Boring hardware: a Matter-compatible button. A passive sensor. A small local hub running an open automation stack. Each replaces a noisier device and pulls its weight without an app store.
How to spot a dead-end gadget
- Cloud-only operation with no offline fallback.
- A novel input modality that the team is clearly proud of.
- "Subscription" mentioned anywhere on the box.
- A name that sounds like a friend (the friendlier, the worse).
This is not cynicism. It is just product design at scale: the categories that survive are the ones with one job, no subscription, and a sensible failure mode.
