Comment by giancarlostoro

Comment by giancarlostoro 5 days ago

17 replies

View on Hacker News

I think on-device models will be the breaking point for AI. Nobody wants to pay for a trillion dollar cloud bill. We've made consumers think that the only way you're paying for software is if you have to buy hardware that comes with it. If you want AI to truly blow up, make it run on potatoes. It doesnt have to do EVERYTHING, just specific needs.

That said, what is with Android phones and their back cameras? They look silly. I thought Apple adding 3 to theirs for the 12 was a bit silly, but at least they made it look nice. One of those models looks like a Battlestar Galactica villain...

nkrebs13 5 days ago

It's preference. I think the cameras on the non-pro iphones are so ugly -- especially the diagonal design. The pro cameras look ok to me. Can't not see my old college stove when I look at it, but I don't think it's too bad.

I, too, am biased but prefer Pixel's camera layout. Visually, I like the symmetry of the camera bump on the back of the device. Functionally, the symmetrical bump means the device will not rock on a table and it's a nice place to rest your finger and support/handle the device. A design decision that's unique and has some (small) utility.

Tier list:

Good: Pixel line, any phone with no camera bump Ok: iPhone Pro Bad: Samsung's many iterations, iPhone 2 camera vertical layout Horrible: iPhone 2 camera diagonal layout

Reply View 3 replies
  • somat 4 days ago

    The whole idea of needing a "camera bump" is sort of a ridiculous design choice just use the extra two millimeters for more battery. It is almost as goofy as the "notch"

    Reply View 1 reply
    • mallipeddi 2 days ago

      Batteries are heavy. I don’t need a fat ass battery making the phone heavier just to hide the camera bump.

      Reply View 0 replies
  • prmoustache 4 days ago

    > Visually, I like the symmetry of the camera bump on the back of the device. Functionally, the symmetrical bump means the device will not rock on a table and it's a nice place to rest your finger and support/handle the device.

    Is anyone using smartphones without a cover that pretty much negates any camera bump those smartphones have?

    Reply View 0 replies
lbrito 5 days ago

>That said, what is with Android phones and their back cameras? They look silly.

Isn't it a market thing though? Doesn't Apple have a phone with horrendous, trypophobia-inducing camera nests?

Reply View 0 replies
ZeWaka 5 days ago

They have the same camera bump design on the Pixel 9 phones.

I quite like it, it's a natural rest for my phone to sit at an angle (and protect the camera glass), and is great for holding it with a single hand.

Reply View 0 replies
izacus 5 days ago

I (and many other people) think the cameras look great and are a nice change from the repetitive boring Apple designs.

Reply View 0 replies
cbsmith 5 days ago

> Nobody wants to pay for a trillion dollar cloud bill.

Buying dedicated hardware as a way to keep your AI bill down seems like a tough proposition for your average consumer. Unless you're using AI constantly, renting AI capacity when you need it is just going to be cheaper. The win with the on-device model is you don't have to go out to the network in the first place.

Reply View 6 replies
  • giancarlostoro 5 days ago

    You misunderstood what I meant, I mean make models that run on potatoes, nobody wants to pay what chatgpt's subscription model probably SHOULD cost for them to make a profit.

    Reply View 3 replies
    • cbsmith 3 days ago

      So the idea is that it SHOULD cost OpenAI a trillion dollars to do what you can accomplish with a potato?

      Reply View 2 replies
      • giancarlostoro 3 days ago

        No, not even sure how you arrived to that conclusion. The idea is that there are models out there that can run on small amounts of VRAM. If all it costs is charging your phone, as opposed to some subscription to some overvalued AI company, people will choose ‘free’ first. We have models that can google things now. They only need to know so much when online, and a specific subset when offline.

        Reply View 1 reply
        • cbsmith 3 days ago

          I think there are lots of advantages to running a model locally. Saving money is one of them, but that's only if you can keep the thing busy. You wisely put the "free" in quotes for a reason: you paid money for the hardware the model is running on, and you're paying for the electrical bill to power it too. Even if you pay a 100% markup to the cloud, unless you're keeping it busy 50% of the time, it's cheaper to rent.

          Reply View 0 replies
  • theshrike79 4 days ago

    The "dedicated hardware" will be an Apple TV in the Apple ecosystem for example if something centralised is needed.

    Or just your phone or laptop. Fully local, nothing leaves the device.

    Reply View 1 reply
    • cbsmith 3 days ago

      So if your AI compute needs are handled by an Apple TV, I'd be really curious how those same needs served by the cloud work out to a trillion dollars.

      Reply View 0 replies
jayd16 5 days ago

I mean, look at these examples. Is a big LLM really needed to hit most of what people want?

Seems like Android just needs to lean into the voice command hooks API. A local LLM can grease the natural language into the mechanical APIs installed on your device. That's a much simpler task than an omniscient robot with access to all of your data.

Reply View 1 reply
  • prism56 5 days ago

    Smaller specialised and targeted models are cheaper, faster and more accurate.

    Reply View 0 replies