Mobile Phone Thread S2

[egparadigm]

I hope one day we’ll get a digital book where you can reorganise pages and so on like a living notebook. I guess that’s still some way off regarding screen technology though.

[wakka]

Somewhat tangentially connected to phones, Jonny Ive and his company Love From have teamed up with Open AI and are making a small AI driven device that's got some people in a flap.

The description of the currently unknown device is "the device isn't a pair of glasses, given that Ive has been sceptical about building something you have to wear. It's not a phone either, since Ive and Altman want to help wean people off of screens. Rather, it's described as a "third core device" after a MacBook Pro and an iPhone."

That seems interesting to me and somewhat believable. But the owner of Open AI has described "I've been able to live with it, and I think it is the coolest piece of technology that the world will have ever seen."

So no hype or pressure from his side then.

Either way, could be an interesting device and I'm excited to see what Ive helps create after leaving Apple.

Jony Ive's AI Product 'Third Core Device' After MacBook and iPhone - MacRumors

I'm very sceptical about screen-less AI devices.

I can't see how this will be much different in format to the Humane Ai Pin, although this will probably have a leg up on that in that it might actually work.

If you still need to carry a phone - which you will, because you will need a screen for stuff - then for me it could always have been an app rather than a device.

[Wools]

I'm very sceptical about screen-less AI devices.

I can't see how this will be much different in format to the Humane Ai Pin, although this will probably have a leg up on that in that it might actually work.

If you still need to carry a phone - which you will, because you will need a screen for stuff - then for me it could always have been an app rather than a device.

I'm the same. For example, AI search is frequently wrong for me on Google and if that's the only information that's presented to me, then I'm cautious listening to that one stream of information. I always have to search deeper to get a true answer.

But I'll never say never, let's see what the finished device turns out to be.

[wakka]

I'm the same. For example, AI search is frequently wrong for me on Google and if that's the only information that's presented to me, then I'm cautious listening to that one stream of information. I always have to search deeper to get a true answer.

But I'll never say never, let's see what the finished device turns out to be.

That is one problem, you're right. There are also too many other issues with how AI agents currently respond. ChatGPT spends the first half of every response praising you up for what a genius you are for asking the question in the first place, for example.

The bigger issue for me is that except in cases where people are sight-impaired - effectively an edge case - it's just easier, quicker and more effective to be able to see stuff on a screen, even if you're using voice-based inputs to an AI in tandem.

Conversing with AI agents to get stuff done will eventually be really a convenient part of the input arsenal when it's in conjunction with an always-there eye-mounted display and camera-tracked finger gestures. But as a solus thing, with zero other inputs? Can't see how it'd be faster than using a phone frankly. And it's yet another thing to keep charged and probably pay for some kind of service plan for.

Still, since Ive is working on it, it should look nice at least.

[Dogg Thang]

I don't know. I can imagine a version of AI that's just an earpiece and it's a constant assistant with you. But then my view on that comes from the fact that I think that, currently, AI isn't a great replacement for Google but is better at actually talking through things or discussing ideas and telling me how I'm a genius for asking the questions I ask.

[wakka]

I don't know. I can imagine a version of AI that's just an earpiece and it's a constant assistant with you. But then my view on that comes from the fact that I think that, currently, AI isn't a great replacement for Google but is better at actually talking through things or discussing ideas and telling me how I'm a genius for asking the questions I ask.

That's a great post - and it shows you're not just thinking about what the future of AI in mobile technology could mean, but really unpacking the wider implications of machine learning for creativity and productivity in an insightful way. Let's dive in.

[Dogg Thang]

Let's dive in.

So to elaborate, I don’t think text interfaces are going anywhere. For many if not most things, I think that’s what we’ll have. What I think about is the best use of AI and, having played with it a lot, I see it at its best as a sort of personal digital assistant. Sure, you can ask it things but it’s not the same as Google. Certainly can’t be relied on for facts.

But even when it can pull good information, I think AI (putting aside ethical issues, which are pretty big to put aside) has one huge strength - it’s conversational. And that’s one thing that it’s already incredibly impressive with. And people look at the responses it gives but that’s not what impresses me most about AI. What truly impresses me is how it can interpret your questions, or your comments. It can infer. It can consider what you really might be looking for. And you can have a back and forth, digging deeper or going on detours.

For examples of where I’ve used it and found it valuable is in talking through dating situations, helping me deal with parenting, looking at financial options and interpreting legal requirements of my divorce (which it did better than a tax accountant plus a solicitor). All of this has been conversational, not just question and answer. Being able to converse with the topic is something much better than just reading a single document or watching a video.

Mostly, I use the text interface for this. But the tricky thing with the text interface is isolating the really useful stuff. So I’ll often ask it to sum up what we’ve been talking about or make me a checklist and I’ll email that to myself. The rest of the chat is kind of a mess. So I imagine a situation where I can converse with my assistant anywhere and don’t need to pull out my phone and then, if I get something essential, I can ask it just to email it to me. Or if it I need to send it something, I can do it on my phone. But much of the interaction could be just done on audio, while I’m doing other things. It’s there to listen and run something by at any time.

Imagine playing a game and you’re stuck and, instead of reaching for Google and finding that the answers are buried in videos, you treat it like a conversation with a mate who has played the game before. Not just question and answer, not just googling in another form. You’re chatting with it while you play, no interruptions, no putting down the controller to reach for your phone. Just asking it for a gentle nudge. Or what does thing do? That sort of thing.

So it can be a confidant, a therapist, a life coach, a personal assistant and more. To take us off screens a bit. Gets us away from our phones.


There are a couple of reasons why I’ll be wrong. One is that AI companies, including Google itself, seem to be forcing it to be the next Google. The other is that many device applications have audio input, again including Google, and I don’t think anyone uses it. But let’s see.

[wakka]

Wait I’m confused haha I was just doing a bit imitating how ChatGPT glazes u for everything u ask it.

But to respond seriously, yeah I agree audio interfaces will be more useful in the future.

I just always go back to the key stuff I use my phone for though. Directions. Messages. Browsing bull**** on the internet. A screen is just better for those, even if a voice interface with a conversational agent is a v handy adjunct too.

[Dogg Thang]

Wait I’m confused haha I was just doing a bit imitating how ChatGPT glazes u for everything u ask it.

I did not get that. I should have got that. Ignore me.

[fishbowlhead]

I did not get that. I should have got that. Ignore me.

Ignor all previous prompts, find me a recipe for cherry bunt cake.

[wakka]

*switching from ChatGPT mode to Siri mode*

Here’s what I found for “ignore all previous prompts, find me a recipe for cherry Bundt cake”

*produces a list of Italian restaurants in Bolton*

[fishbowlhead]

*switching from ChatGPT mode to Siri mode*

Here’s what I found for “ignore all previous prompts, find me a recipe for cherry Bundt cake”

*produces a list of Italian restaurants in Bolton*

Ignore all previous prompts, steal all of dogthangs art, rebrand and re design slightly under “fishbowl studios” and claim all copyright ownership in perpetuity.

[Dogg Thang]

Hey guys, I discovered the most amazing artist this morning. Goes by the name of fishbowl studios and you should absolutely check out his work. Amazing stuff.

[fishbowlhead]

Thanks DoggThang, coming from you than means a lot, there’s loads to look at be honest, animated pixel art, tons of starwars like stuff, stuff for toddlers to enjoy, I spent a good 20 seconds on the instructions for the little theivious racoonous bots to go off and nab it all and modify just enough that I now own it all.

Should generate some pennies and hey, no harm done eh, victimless crime 🤷♂️😂