CES wrapped this week and, as always, the volume of coverage vastly outpaces the number of things that will actually ship and matter for how anyone works. Filtering through my own notes from the show, here's what I'm actually taking seriously versus what I expect to quietly disappear.
The smart glasses category keeps fragmenting, which is good
Rather than everyone converging on one design philosophy, this year made clear the category is splitting into genuinely different bets: display-forward devices, camera-forward devices, and now a lighter screen-less style aimed at all-day comfort over feature density. That fragmentation is healthy for the category even if it's confusing for buyers right now, because it means real experimentation rather than everyone copying the same approach and hoping marketing carries it.
Computational imaging previews worth remembering
A few sensor and chip announcements this year point toward meaningfully better low-light performance in small-sensor devices, action cameras specifically, over the next product cycle. Nothing ships from this immediately, but it's the kind of announcement worth filing away, because it usually predicts what next year's actual camera releases lean on.
- Smart glasses: fragmenting into distinct design philosophies, worth watching each separately
- Small-sensor low-light gains: a preview of what's coming to action cameras next cycle
- Most home AI and robot demos: still not relevant to how I actually work
What I'm skeptical will hold up
A handful of AI-everything product pitches this year felt like a feature bolted onto a device that didn't need it, and I'd bet a fair number quietly disappear or get scaled back before an actual retail release. CES optimism has a long track record of not surviving contact with actual manufacturing costs and real-world battery constraints, and I don't expect this year to be dramatically different from that pattern.