The advent of Meta's Muse assistant, along with various other similar assistants like Instinct, will probably open the door to making IoT devices finally useful. This will probably be really bad for companies like Xiaomi (this might mean Apple too) but good for new hardware entrants (partially because the personal superintelligence landscape will be multipolar for the foreseeable future, but mostly because the people who will win the distribution game really suck at making good hardware).

This means it will be an exciting time to build "premium" feeling IoT devices that might historically have felt gimicky or unappealing to most consumers. It will be a challenging to compete directly with the big players on their core product since they can just subsidize the hardware device as a loss leader for their ecosystem, but it will be a really interesting time to build ancillary devices. Here's some ideas I was thinking about that seem interesting.

Consumer

• Smart fridges
  • The idea of a fridge that keeps track of what food you've bought, what is about to expire, and so forth has historically felt silly, but in a world where your agents are making your grocery list (or perhaps shopping on your behalf) it could actually be a strong value-add.
  • I kind of like the idea of a "home vending machine" that gets restocked while you're out at work.
  • I would actually buy a smart fridge for myself today when a year ago I would have thought they were mostly a waste of money.
• Home robots
  • The Matic voice experience is pretty magical. There will probably be another magical experience when it is mostly controlled by an agent like Muse or ChatGPT.
• Multi-agent smart speaker
  • Smart speaker that works with any assistant and lets people bring their own assistants, with some premium feel.
  • I don't know how big of a market there is for this. It would probably be some niche product for privacy-minded Libertarians (historically not a high-spending demographic).

Medical

• EHR-adjacent med tech
  • My wife is currently working for a company called Ambience. It feels like they're competing with all the big labs to be the EHR for the AI era - it will be exciting to see what that looks like
  • Devices that can seemlessly interface with this new EHR, built by people who can effectively sell into the hospital system, will likely have promising futures
• Prosumer health and fitness
  • If smart assistants really do become multi-polar, and health data storage will mostly shift into some intermediate layer, prosumer medical hardware which can easily interface with agents will probably become a lot more valuable.
  • This could potentially take a lot of different forms:
    • Watches / rings that track cardiovascular data
    • Ambient "life trackers" like Plaud, which could track stuff like what meals you've eaten and when
  • Meta and others will definitely make first-party devices in this space, but there's a lot of niches in prosumer health tech

Commerce

• Assuming that a lot more commerce will be done by agents, there's likely opportunities for IoT-esque devices that agents can talk to and buy things when needed
• Potentially a "vending machine-ification" for how a lot of stuff is transacted - autonomous nodes at various points with easy-to-use transaction APIs
• The package drop-off boxes that many apartments have are a good example - these might become a lot more common in a world where agents are doing a lot more autonomous shopping.
• In a similar vein, hotel delivery robots are super, super common in China, but adoption in the United States is still very limited (my understanding is this is mostly because of weird American regulations around IoT for elevators).
  • If you can solve the button pressing problem, agentic commerce might be the thing that drives more widespread demand for these kinds of robots in American apartment buildings and hotels.