Report: Apple to 'Re-Examine' AI Development

[3xBoom]

Apple should turn being late to this recent AI party to success by taking a long look how chatGPT etc are going to play out over the next 12 months. It is obvious that there will be more and more real and stupid concerns about AI, more issues, elation, and panic, plus the emerging governmental regulations and demands. By observing and learning from these developments, Apple can save themselves some trouble by developing a strong AI strategy that should be built around improving or optimizing their hardware for AI needs and acquiring the relevant AI talent/startups.

 

[CordovaLark]

Perhaps Apple can work to make Siri recognize a Cat in photos in five years, where it took them a decade to get it to identify Dogs in pictures.

 

[Kierkegaarden]

At this point apple would be ten years behind to catch up unless they buy a big company out

You are assuming that they haven’t had engineers working on this for years. Do you think it was a coincidence that Microsoft and Google were announcing their product around the same time? I think there was a sense of that this was going to be a focus for tech companies years ago.

 

[maflynn]

You are assuming that they haven’t had engineers working on this for years.

I think the implication is they've not had people working on it - at least to the level that google and MS. I don't know for sure, but they seem more focused on consumer stuff and this isn't a direct consumer product.

Additionally, if they had engineers on AI, wouldn't we see some fringe benefits of the work, such as an improvement with Siri? Of all of the voice asst.s Siri is the least powerful and most constrained

 

[JimmyHook]

Generative AI is hot garbage. It’s trash software that has the veneer of success. I’m glad Apple is staying out of it. AI is useful elsewhere, but has zero value “generating” content. It’s all so bland and uninspired, not to mention is massively error-prone. If it can ever determine accuracy, while providing for more creativity, then it’ll be useful. But that’s a loooooong ways away.

 

[JimmyHook]

Surely you aren’t telling me that Apple is just now looking at re-examining Siri? If so, I agree with @Macalway that this is the most pathetic story of the year.

At this point, I liken Siri to the Apple Maps situation. Apple will never catch up. They’re just too far behind. The only way forward, as some others have suggested, is to purchase some companies that already have established technology. It’s the only way I can see them moving up to a decent baseline.

Weird analogy. Maps has more than caught up, it has surpassed Google Maps.

 

[JimmyHook]

Siri was never neglected.
Apple doesn’t want AI to behave like a human being.
That’s Apple’s vision on it.

Good. Because Generative AI sucks right now anyway

 

[JimmyHook]

Apple's really late to the game in this area.
Even for AI image generation, the native stable diffusion implementation (coreML) is significantly limited if compared to the windows Implementation.

Why? Those things are not actually useful at all

 

[wilhoitm]

I think the implication is they've not had people working on it - at least to the level that google and MS. I don't know for sure, but they seem more focused on consumer stuff and this isn't a direct consumer product.

Additionally, if they had engineers on AI, wouldn't we see some fringe benefits of the work, such as an improvement with Siri? Of all of the voice asst.s Siri is the least powerful and most constrained

Apple has been working on AI for years! Don't forget about the elusive self driving Apple Car! I believe that is why Apple Maps has improved significantly also!

 

[Serqetry]

At this point, I liken Siri to the Apple Maps situation. Apple will never catch up. They’re just too far behind. The only way forward, as some others have suggested, is to purchase some companies that already have established technology. It’s the only way I can see them moving up to a decent baseline.

What are you talking about? Apple Maps is great. In my opinion they not only caught up, but surpassed Google Maps.

They can do the same with AI and Siri if they want to.

 

[karranz]

Just stay out of AI, Apple. Or else ChaptGPT re-evaluates who’s going to be the new CEO of Apple. Powered by AI.

View attachment 2170321

Scott Forstall Reportedly Forced Out of Apple

Earlier this evening, Apple announced a major change up in their executive management. Most notably, Apple Senior Vice President of iOS Software...

www.macrumors.com

Scott and Jony are part of the original team that brought the most amazing things apple has seen. They would be obviously welcomed, apple needs fresh air and bold decisions, they have gotten boring and expensive.

 

[klasma]

Buried the lede. When did Microsoft ever have a lead in AI?

Last I heard, they bought their way into integrating ChatGPT with Bing.

Also GitHub Copilot. It doesn’t matter if something is merely bought, Microsoft is not a coherent entity and consists of a bunch of infighting factions anyway.

Microsoft org chart meme:

 

[User 6502]

Siri was never neglected.
Apple doesn’t want AI to behave like a human being.
That’s Apple’s vision on it.

No, apple simply doesn’t have a vision on the matter, or perhaps it’s incapable of implementing whatever vision it has. Siri is a disgrace, and it’s so embarrassing apple should retire it and allow using alternative third party assistants as system defaults.

 

[Serqetry]

Siri is a disgrace, and it’s so embarrassing apple should retire it and allow using alternative third party assistants as system defaults.

That's ridiculous. Siri may be embarrassingly limited now but it wouldn't be hard for a company like Apple to update her with a real AI brain and more abilities if they really wanted to. Hopefully they do because I don't want 3rd party junk that can't interface with the Apple ecosystem properly.

 

[farmboy]

Wake me up when *DigiTimes* is considered an accurate source for internal, top secret activities and developments at Apple, Google, Amazon, Meta, etc.

 

[ian87w]

Siri isn't the worst AI assistant (that dubious honor probably goes to Samsung's Bixby), but it's close. Which is a little extra embarrassing considering it was the first major AI assistant to be put on a phone. 😅

Honestly, some serious improvements in Siri have been long overdue; I'd settle for it being as good as the regular Google Assistant at this point.

Bixby is bad at the general assistant stuff, but if you have a Samsung phone, it surprisingly has access to many more hardware controls than Google Assistant, or even Siri on the iPhone.

 

[JippaLippa]

Why? Those things are not actually useful at all

I'm all against the blatant theft of art (I'm backing the proposal of some eu regulations) but you cannot say it's useless.

If you avoid any specific plagiarism, it's immensely useful to brainstorm some ideas and generate some content that of course you'll need to elaborate into something new (Like textures in my case).

It comes with many controversial aspects but you cannot say it's useless.

 

[JimmyHook]

I'm all against the blatant theft of art (I'm backing the proposal of some eu regulations) but you cannot say it's useless.

If you avoid any specific plagiarism, it's immensely useful to brainstorm some ideas and generate some content that of course you'll need to elaborate into something new (Like textures in my case).

It comes with many controversial aspects but you cannot say it's useless.

My brainstorming is faaaar more powerful than AI. We have brains that are orders of magnitude more complex. Why would I let an inferior tool do my work?

 

[JimmyHook]

Apple should turn being late to this recent AI party to success by taking a long look how chatGPT etc are going to play out over the next 12 months. It is obvious that there will be more and more real and stupid concerns about AI, more issues, elation, and panic, plus the emerging governmental regulations and demands. By observing and learning from these developments, Apple can save themselves some trouble by developing a strong AI strategy that should be built around improving or optimizing their hardware for AI needs and acquiring the relevant AI talent/startups.

Or they can realize that Generative AI produces bland and uninspired copy. If everyone uses it, the entire world’s output will slowly start looking the same. It’s a garbage tool

 

[Serqetry]

Or they can realize that Generative AI produces bland and uninspired copy. If everyone uses it, the entire world’s output will slowly start looking the same. It’s a garbage tool

Everyone's going to use it, so you might as well get over it now.

Also, AI output doesn't all look the same if you know how to use it. It's very easy to make chatGPT output pass the AI detector tests.

I know Skynet is going to kill us all and there's nothing I can do about it... so in the meantime I'll enjoy the hell out of AI tools.

 

[ian87w]

First thing come to mind is to use AI to broaden Siri’s ability to understand commands and intent. I want Siri to be more open ended, like being able to converse first to understand a command more deeply and properly. Right now, I have to specify the whole command in one sentence correctly to get Siri to do anything. That means I already have to think for the complete command and intent before invoking Siri.

Basically at this point, I have to be a robot first, spouting direct whole command in one go for Siri to do it. That’s not an assistant. I want an assistant to be like the computer in Star Trek or Jarvis in Iron Man, where you can just start by conversing, forming the whole task and intent on the fly as we go.

 

[nt5672]

Honestly, that's not really true anymore. So much research is public and so many of the crucial developments are open-sourced from day one, that it's relatively easy to get up-to-scratch quite quickly, with a good team. Apple can hire the best, so they could accelerate a strategy very quickly, I'd imagine. And actually, it was an Apple paper that led to the development of RWKV-LM—which we could be the basis for some significant movement in LLMs this year—so we know they're very capable of good work. I mean, I may turn out to be wrong, but there are good reasons to switch from the original Google/OpenAI approach to RWKV-LM (particularly if you're starting from scratch)... not the least of which is the potential offered by the way-cool SpikeGPT... 😎 Also, Apple has a sheeaaat-ton of $$$, so they'd feel no pain throwing a buh-zilleeon A100s at a problem... which, let's be honest, is one of the major hurdles... just simply raw compute power... which is just dollars... sad but true.

Apple has often benefitted from coming late to the party, so it could certainly happen again.

The problem is that the best don't fit into Apple's management structure or environment. Why do you think Apple chased off the original Siri leadership it purchased.

 

[TracesOfArsenic]

Just stay out of AI, Apple. Or else ChaptGPT re-evaluates who’s going to be the new CEO of Apple. Powered by AI.

View attachment 2170321

Scott Forstall Reportedly Forced Out of Apple

Earlier this evening, Apple announced a major change up in their executive management. Most notably, Apple Senior Vice President of iOS Software...

www.macrumors.com

In all seriousness, I'd love to see Apple run by Forstall. Never going to happen with how he was removed from the company but it's a great thought experiment.

 

[DeepIn2U]

An interesting story.

In other words Apple does NOT have their own generative AI technology to speak of. Nice going Cook & Co. lol.

Yet again what the hell is Apple's SVP of Machine Learning and AI Strategy John Giannandrea been doing the last 5yrs?!!!????

Grrrrrrr! Wasted stock options wasted seating space and moved him to front a Car development that is not going anywhere beyond a UI for existing cars later this year! And Siri his original 2yrs tenure is still a laughing stock amongst its peers!!

 

[jarman92]

Honestly, that's not really true anymore. So much research is public and so many of the crucial developments are open-sourced from day one, that it's relatively easy to get up-to-scratch quite quickly, with a good team. Apple can hire the best, so they could accelerate a strategy very quickly, I'd imagine. And actually, it was an Apple paper that led to the development of RWKV-LM—which we could be the basis for some significant movement in LLMs this year—so we know they're very capable of good work. I mean, I may turn out to be wrong, but there are good reasons to switch from the original Google/OpenAI approach to RWKV-LM (particularly if you're starting from scratch)... not the least of which is the potential offered by the way-cool SpikeGPT... 😎 Also, Apple has a sheeaaat-ton of $$$, so they'd feel no pain throwing a buh-zilleeon A100s at a problem... which, let's be honest, is one of the major hurdles... just simply raw compute power... which is just dollars... sad but true.

Apple has often benefitted from coming late to the party, so it could certainly happen again.

I think the issue is less to do with the actual tech and more with perceptions. Every Apple user “knows” that Siri is garbage, and it’s been shown to be garbage again and again and again for more than a decade. If Apple comes out with a new Siri tomorrow that blows Google and Alexa out of the water, it will still take years to convince people to bother trying it again. Apple let Siri die a long time ago and I’m not sure they care enough to revive it.