The MacRumors Show: Where Is Apple in the Generative AI Race?

[AppleFan735]

Most apple services suck, especially outside the US. Siri was released over 11 years ago and shipped pre-installed on billions of Apple devices sold since. Apple had the time and distribution advantage matched by no other company on the planet. The fact that Siri is not running laps around ChatGPT is a dismal failure on Apple side.

Is Eddie Cue still leading services and Siri at Apple? If Steve was around this would be handled like the MobileMe situation.

 

[purplerainpurplerain]

Microsoft is gonna incorporate Ai throughout windows.
The new Bing is much better than Google search and Google will hopefully incorporate Ai into Android OS.
Apple can continue with its Car project.

Hohoho ClippyGPT throughout Windows and Bing and whenever you see it act buggy it will gaslight you and say ‘No sir this was not a bug it is a feature!’

 

[purplerainpurplerain]

Ross Douthat, writing for the New York Times, warns that "aside minor questions like whether rogue A.I. might wipe out the human race", AI carries other risks that include: "this kind of creation would inevitably be perceived as a person by most users, even if it wasn’t one", and "a place where an entire civilization could easily get lost."

So between OpenAI's ChatGPT, Microsoft's Sydney, Google's Bard, and Apple's souped-up Siri or Siri-enhanced Safari browser, AI may one day have the power to:
1. Terminate us like Skynet
2. Control us like the Matrix
3. Delude us with false truths.

I think that last possibility is the most dangerous.

There are dangers and lots of bugs and hallucinations that could be dangerous. In those cases people and orgs can just sue the mother out of AI companies for releasing and maintain software that is so bad it screws society up.

Simply put, social media has gotten away with being a cesspool because they blame users. They can’t do that with AI. If the machine is a cesspool and produces misinformation then the blame is all on the developers.

As for people who keep posting insane infatuated ideas about AI generated garbage, most of those are not programmers. They are just some people who watched some sci fi movies and are so lazy they want a machine to do all the work AND thinking for them. They don’t understand that it isn’t possible and that it will be janky and broken forever.

Every new tech hype cycle brings these people out.

 

[purplerainpurplerain]

DALLE, chatGPT and dream booth are in embryonic stage.

They have been under development for years if anyone remembers playing with things like deep dream a decade ago. The algorithms were understood years ago. They just needed GPUs for scale and that became available more recently.

 

[TechnoMonk]

They have been under development for years if anyone remembers playing with things like deep dream a decade ago. The algorithms were understood years ago. They just needed GPUs for scale and that became available more recently.

AI is not algorithm, and what was understood 3 years ago is history, let alone 10 years ago.

Like anything else, the generative AI went through variations of CNN, GAN, style transfer among other things. Biggest gains started when combining with CLIP. The Latent diffusion models in past couple of years was a big leap. One could argue diffusion technique in general terms in 3-4 decades old.
What Open AI, stable diffusion, and all the cool demos did provide weights for checkpoints. You can train your own models using those checkpoints, or just use Midjourney. It’s an exciting time with more GPU power available for sure.

 

[purplerainpurplerain]

AI is not algorithm, and what was understood 3 years ago is history, let alone 10 years ago.

Again AI fan who just makes stuff up.

It is true that there has been progress, but the last decade isn’t history in the same way all the software you use today is based on decades of incremental upgrades. Nobody with any sense would say ‘ah well

macOS Yosemite is history and has nothing to do with current macOS’.

Same applies to all these machine learning apps. Incremental upgrades and tons of bugs because this stuff is much harder and compute intensive than anything. There will be limits to what it can do and one shouldn’t hope that some kind of magic will suddenly happen and we will have ‘high res 60fps photorealistic generated images’.

We don’t even have that in gaming with pre-made assets and they have been telling us that photorealistic graphics would come in 2005 then 2010 then 2015 then 2020 😂😂🤣🤣😂

The most absurd and brainless thing AI image people say ‘this will help us imagine impossible things and new concepts!!!111’

Like, wtf do these people think Salvador Dali, Stan Lee, Hideo Kojima and George Lucas team were doing? All the fashion designers who come up with bizarre shows twice a year?

They used their brain. It was free. It was limitless. It didn’t need to use 1000 Nvidia GPUs. 😛😛😂

AI isn’t going to make some lazy ass person into a superstar. The people who use their brain will always have the fans and fame.

 

[TVreporter]

I just asked Siri “Are you AI?” twice and got these responses:

“I’m a virtual assistant, not an actual person but you can still talk to me.”

“I’m not a person - or a robot - I’m software, here to help.”

 

[Arran]

Where is Apple in the generative AI race?

Is that even a good race? Isn't it still just monkeys bashing away on typewriters?

 

[topdrawer]

Id settle for a decent spell checker. Cant be that hard right?

 

[madebybela]

Apple was absent from the crypto/nft/metaverse rush and now they're absent from the AI race. Maybe they're not behind, they're just dodging a bullet.

 

[maxoakland]

Generative AI is bad for artists. Apple shouldn’t be participating in any “race” because the AI race is a race to the bottom where profits are put over people

 

[femike]

Apple hasn't even begun tieing their shoelaces yet. Apple is just too busy undertaking the mammoth task of deciding what colours the iPhone 15 and the Apple Watch will come in. Also, what has them baffled is how to remove the USB port from underneath their mouse. There is no time for AI. But on second thoughts, AI might help them solve these resource intensive problems.

 

[DeepIn2U]

I stopped watching at 35:15!

Hey Arn you need to shake that secondary dweeb whom made this mistake “Apple is a hardware company …”

WRONG!

Apple is a Software company that takes its own software very seriously enough to make its own hardware!

Like how many times did Jobs state this before his demise?! Even quoting Alan so many times about this!
Apple makes more software and software services than they do hardware!! Please stop MacRumors staff from continually making this mistake! 😤🤬😡 (proper use for emoji there!)

" held its annual AI summit for employees"

and nothing really leaked did it?

good on apple employees I guess. (but I want some leaks

Right now it seems e emoji is more important. And with John Giannandrea as AI lead don’t expect too much!

Here is where they are 😮:

View attachment 2160198

^ look no further than the lead for Siri:

John Giannandrea is Apple’s senior vice president of Machine Learning and AI Strategy, reporting to CEO Tim Cook. John joined Apple in 2018 and oversees the strategy for artificial intelligence and machine learning across the company and development of Core ML and Siri technologies.

Since 2018 and in less than 6 mths given stock options and all we’ve seen of Siri since then is:

Slightly more natural voices - software team can do this alone.
Ability to send messages across HomePod minis on same network - whoop-do.
1 or 2 more languages,
And a more FANCY UI presentation - yes presentation!

Oh 1 more thing we’ve got and it had NOTHING to do with John, it was purchasing a savvy Developers’ software called Shortcuts and implementing it with a Siri hook.

5yrs and THATs all we’ve got to show from hiring John! The Apple Silicon ML cores have vastly improved dare I say exponentially and Siri still can’t do much more functionality.

Someone at Apple that’s been there longer than Ti, say dev #8, or Eddie Cue to sit John Giannandrea down and make him watch Apple Knowledge Navigator commercial. Make him really watch it force him to take notes and then ask him:

Why can’t Siri do exactly what was shown in that video since 1987!!??? That video literary is what the iPad was envisioned to be and just about almost there.

I’d love for John Giannandrea to prove me wrong that he’s just punching a clock and collecting funds for his retirement from the tech world. But if Spotlight, Siri, and the kernel doesn’t have a major overhaul beyond what Google and Microsoft is doing with GPT4, by 2024 Q1 - kick em out! He’s further life sucking Siri.

Beyond the benefits of Siri for those physically challenged (I hope I have that right and politely saying so) many users don’t know to the extent what Siri can really do - for just iOS users with just 1 iPhone no other Apple product.

Apple has done a very poor job in advertising Siri to the point it’s embarrassed to even mention it at their events! Is there a full page showing all of what she/it can really do, is so why is it not on the main page?! Why is it not part of Apple’s home page search?!

Why do we need to train Siri on new iPhone or Mac setup to learn our voice IF we don’t have an option to drop “Hey Siri”?! Moreover why isn’t the TrueDepth Camera in FaceID not used to recognize “Hey Siri” facial movements without using our voice?!!!? @JohnGiannandrea this would use on device ML cores incredibly well - especially if I’m looking up an account number for my cell provider in their app but on a call with said providers’ customer service rep asking me for this I can mouth “hey siri” the speak “enter numbers 7733345589” and the automated prompt accepts without me switching to the phone app to dial it. Or asking Siri to show me the account number from an app on screen so I can tell the actual person on the phone the account number without leaving the phone app.

So many uses for a real AI on our smartphones which would remove the need to show two apps on the screen at the same time.

Also WTF is going on with video pic in screen it’s still bonkers and finicky from app to app (1st or 3rd party)!

 

[SoldOnApple]

Apple is going to have to spend big to get and outright own an AI with half the buzz of ChatGPT. Apple should have funded OpenAI really to get exclusive virtual assistant rights to ChatGPT, their technology is basically already mature.

 

[DeepIn2U]

DALLE, Stable diffusion have poorly trained Models which need a novel of negative prompts to generate anything worth. Fingers, arms, toes and legs are big problem unless they are trained on custom models.

Why are human extremities needed to be trained on custom models in the first place? Serious non AI student question here. An AI being built to interact and understand humans doesn’t even know our basic structure, then how far can it really help us?

50lbs crate of products 2 handles and need to be balanced : answer any human can carry this.
- but what it a human has only 1 hand? 1 full hand and the other only a thumb?!

200 pound item across 8ft length needing 2 humans to carry 40ft away. Answer 2 humans needed. What is 1 of those humans has 1 leg and 1 early model leg prosthetic for an amputation above the knee?! Comfort may affect the solution here or distance carried or the time needed to carry across that distance.

 

[Abazigal]

Hey Arn you need to shake that secondary dweeb whom made this mistake “Apple is a hardware company …”

WRONG!

Apple is a Software company that takes its own software very seriously enough to make its own hardware!

Like how many times did Jobs state this before his demise?! Even quoting Alan so many times about this!
Apple makes more software and software services than they do hardware!! Please stop MacRumors staff from continually making this mistake! 😤🤬😡 (proper use for emoji there!)

Apple's profits come primarily from hardware, but they do use software to sufficiently differentiate their hardware and sell them at a handsome margin, with services coming in to further sweeten the deal.

I will argue that Apple sells experiences, made possible by their control over hardware, software and services, to the point where it's difficult to tell where one ends and where the other begins. And that's what people pay for. An integrated computing experience that just works out of the box.

Perhaps this is why nobody has been able to disrupt Apple, despite offering products that may seem superior to their hardware or software on paper, because Apple is competing in neither directly. Nobody has been able to replicate the unique Apple experience (much less surpass it), because to do so, they need to be able to reproduce the entire Apple ecosystem in its entirety. It was too much for Microsoft and even Google, and it will be too much for any new entrant in the market today.

Apple was absent from the crypto/nft/metaverse rush and now they're absent from the AI race. Maybe they're not behind, they're just dodging a bullet.

Agreed. I am of the opinion that a lot of this initial enthusiasm regarding AI chatbots deserve to have a great deal of cold water poured on it, not least because I feel that there is a complete lack of design and human thinking regarding language models being used to push information to us. In short, I feel that Google and Microsoft are going about it the wrong way, in that they are making the classic error of putting too much emphasis on the underlying tech and not enough on the use case (which is the same problem facing folding phones and why they have yet to take off).

I can't quite put my finger on it yet, but I think that as more time goes on, we will start to see more holes in this whole AI narrative.

 

[Piggie]

I may be totally wrong, and please tell me if you think I am.
But my feeling is Apple is WAY behind in AI and will stay that way for the foreseeable future.
Why?
Because anything AI involves taking risks and looking bad.
Apple will want to have their system super safe, and super restricted to enable it to actually be what AI needs to be.
A bit like a "Disney super safe for kids" AI system.

I could be wrong, but nothing I've seen or heard from Apple now of in the past leads me to feel they would want to release anything less than perfect, and AI may never be perfect enough for them, unless it's amazingly restricted in what it's allowed to do.

They are going to need to act fast as others are not going to wait for them, and AI could very well explode with exponential growth over the next decade.

 

[obviouslogic]

Hopefully Apple are ignoring the crapfest.

Yeah Apple is not going to enter a market "just because" XXXX does. As they've stated over and over again, they're only interested when they think they can offer their users something better. Or enhance their ecosystem in some way.

Apple uses AI all over the place in their OSes to enhance the user experience in the form of small conveniences/features.

They do not have a search engine that competes across the web, Spotlight is still only used internally on their platforms, so jumping on this bandwagon doesn't make much sense. Although there were once rumors that Apple was working to release Spotlight as web search engine. So, we'll see what comes of that. (Most likely a lot of litigation.)

Although, with the advent of these new fangled AI features, maybe this would be a good time for Apple to release a basic search engine?

 

[obviouslogic]

Apple hasn't even begun tieing their shoelaces yet. Apple is just too busy undertaking the mammoth task of deciding what colours the iPhone 15 and the Apple Watch will come in. Also, what has them baffled is how to remove the USB port from underneath their mouse. There is no time for AI. But on second thoughts, AI might help them solve these resource intensive problems.

God forbid they take away resources from those tasks to, you know, waste time designing CPUs, GPUs, NPUs, ISPs, SoCs, mobile radios, Bluetooth/WiFi radios, microLED displays, etc. Because obviously Apple's only good at and interested in non-technical stuff.

 

[obviouslogic]

I may be totally wrong, and please tell me if you think I am.
But my feeling is (1) Apple is WAY behind in AI and will stay that way for the foreseeable future.
Why?
(2) Because anything AI involves taking risks and looking bad.
Apple will want to have their system super safe, and super restricted to enable it to actually be what AI needs to be.
A bit like a "Disney super safe for kids" AI system.

I could be wrong, but nothing I've seen or heard from Apple now of in the past leads me to feel they would want to release anything less than perfect, and AI may never be perfect enough for them, unless it's amazingly restricted in what it's allowed to do.

(3) They are going to need to act fast as others are not going to wait for them, and AI could very well explode with exponential growth over the next decade.

1. No one knows for sure, because Apple doesn't flaunt technology for its own sake. Apple's m.o. has always been to utilize technology to enhance something, which basically means the core tech stays in the background. AI is used all over Apple's devices in both hardware and software... It's just not in your face saying look at what I can do! It's used to categorize photos, determine which processing unit is best for a task, select handwritten text, offer suggestions on what you might want to do next, etc.

2. Umm, moving the Mac to ASi was an enormous risk for Apple to make.

3. This sentiment has proven to be false over and over again when it comes to Apple. There's nothing wrong with being late into anything as long as you bring something novel to the table.

 

[ponzicoinbro]

Bing Chat is losing its mind and nobody saw it coming lmao

 

[ponzicoinbro]

Yeah Apple is not going to enter a market "just because" XXXX does. As they've stated over and over again, they're only interested when they think they can offer their users something better. Or enhance their ecosystem in some way.

Apple uses AI all over the place in their OSes to enhance the user experience in the form of small conveniences/features.

Of course "AI" is just a buzzword that covers a range of algorithms and not just one thing.

Because it is the current trend everything gets called an AI even stuff we have used for years.

It's like when this stupid crap "metaverse" became a buzzword last years then suddenly the bum stains at CNBC and other outlets started calling all decade old MMOs "metaverse".

Auto type, grammer checker, spell checker and a wide range of tools use algorithms without using buzzwords. I am betting 10 bucks that Photoshop gets re-branded "Photoshop AI" and almost nothing in the app will change, except maybe those useless stupid neural filters.

 

[obviouslogic]

Of course "AI" is just a buzzword that covers a range of algorithms and not just one thing.

Because it is the current trend everything gets called an AI even stuff we have used for years.

It's like when this stupid crap "metaverse" became a buzzword last years then suddenly the bum stains at CNBC and other outlets started calling all decade old MMOs "metaverse".

Auto type, grammer checker, spell checker and a wide range of tools use algorithms without using buzzwords. I am betting 10 bucks that Photoshop gets re-branded "Photoshop AI" and almost nothing in the app will change, except maybe those useless stupid neural filters.

And it's not just that... it's also the willingness of some companies to throw R&D projects into the public spotlight. Google does this all the time. Apple does not, so it will always seem like they're behind or not even working on something.

 

[BEUVK]

AI is nothing new. Youtube has been generating transcript of audio for sometime. The problem is if you really want a correct transcript, you still need someone to do that. Machine will never be able to generate a correct transcript. It depends on the types of job. Most jobs need 100% correct. Some job do not need 100% correct, like entertainment, talk show, things that burn your time or you have too much time to spend and find something to fill the time, things that you do but you can do without or some useless things.

 

[TechnoMonk]

Why are human extremities needed to be trained on custom models in the first place? Serious non AI student question here. An AI being built to interact and understand humans doesn’t even know our basic structure, then how far can it really help us?

50lbs crate of products 2 handles and need to be balanced : answer any human can carry this.
- but what it a human has only 1 hand? 1 full hand and the other only a thumb?!

200 pound item across 8ft length needing 2 humans to carry 40ft away. Answer 2 humans needed. What is 1 of those humans has 1 leg and 1 early model leg prosthetic for an amputation above the knee?! Comfort may affect the solution here or distance carried or the time needed to carry across that distance.

AI generative image models are trained on billions of images, and the weights in the models reflect it. One of the biggest problem is the art datasets, used in training. The human structure in a painting is far from perfect. There are other technical reasons too.
Just like any other evolving technology, it gets better every day. Like I said most of AI is not in infancy, but it will disrupt the way things are done.
AI doesn’t have to be perfect, it’s another tool that can have great benefits as it evolves.