> The next major device will definitely be AI-first.
Be more specific though: what form factor will such a device be in?
A coffee maker? A phone? Glasses? Cars? A building?
The AI wave seems to be hoping a whole load of hardware revolutions, such as holographic displays, will just appear out of the ether because it fits with their vision of how things should be.
I don’t see why the device has to be in a significantly different form factor than current phones. AI alleviates one of the major problems phones have right now, which is that typing on them is slow.
If you can type a half-assed message, and have AI fill in the blanks, or reliably transcribe your voice, that’s a huge improvement to the phone in its current form factor. No reason the screen or interfaces have to undergo a radical transformation
I agree. Humane tried with a new form factor. An audio-only interface is too limited. A watch screen is too small. There's a lot going for a decently sized screen that lets you look images, maps, emails, webpages that you control (rather than projecting onto the nearest (in)convenient surface), that fits in your pocket. Good enough BCIs are still years/decades away.
I could imagine AirPods that connect to various screens embedded in the environment, which you temporarily use when next to them. But it's still jot as convenient as a screen in your pocket.
If it's a phone it will have to run Android, with the Play Store to get the apps, at which point it will be very dependent on Google.
You aren't going to get people giving up their mobile banking apps to carry an AI phone that doesn't quite work, hence the need for it to be something else.
This would cost 50 billion or so. But right now you probably interact with at least 3 or 4 oses per day.
Your TV, has one. Your phone has one, your laptop has one. And if you have voice assistants, they run a 4th distinct OS.
The future will have one OS that shares a session.
Two paths exist. 1. This runs primarily locally aside from a very small amount of data to share the session ( which you can disable). It's completely open source and modifiable.
If you want to roll a 3500$ super PC it'll be just as compatible with the OS as a 200$ one. Writing small automated tasks, everything from just asking with a voice command to wake up jazz,to running a custom C script, will be easy to do.
While I'm dreaming I want a new programming language which supports 3 levels. Plan English instructions ran though an LLM, something like Python and a systems level language like Rust. All "native" programs will be built in this framework.
Now, the negative path is this is all closed source, processed in some data center. "John, I noticed you said to Brian your feet hurt, new running shoes are 30% off , just say the word."
This is the far far more likely outcome. They're going to build an AI that's constantly with you, integrated in every device you own, and it'll all be to sell you stuff.
"Waymo, I would like to go home."
"Sure, but let's stop for milkshakes."
"Waymo, please , I'm tried."
"Understood, I've arranged the milkshakes to be dropped off an your apartment."
This technology could be amazing for accessibility, even real time sign language translation would change the world.
We'll get some of that, but the end goal will always be making as much money as possible. Ultimately selling us crap. Your awake for 16 hours today. You must be monetized every waking second.
Once they figure out how to get the science from Dream Scenario to work I'm sure they monetize sleep too
If Apple has an opening for an AI-first device, its AirPods with their own Apple Watch style SIP and data connection + replacing Siri with a real multimodal model.
Be more specific though: what form factor will such a device be in?
A coffee maker? A phone? Glasses? Cars? A building?
The AI wave seems to be hoping a whole load of hardware revolutions, such as holographic displays, will just appear out of the ether because it fits with their vision of how things should be.