This is good news. When I read the news that Apple is going to invest into OpenAI, I was a bit sad because Apple always tries to avoid to jump into the hype trains like blockchain, NFTs, metaverse etc.
Yes, there is Apple Intelligence but it is mostly writing tools, Xcode code completion, an emoji generator and so on. Nothing like something that is going to replace jobs, AGI and so on how OpenAI is marketing their products.
The knee-jerk replies you're getting are hilarious. The reality distortion field is as strong as ver. You see, because it's Apple, it can't be a bet in virtual reality with a bad connotation (metaverse). To some, Apple can do no wrong nor do they make mistakes. If they invest (and fail) in VR it was from a good angle, not purely for profit (just like Meta).
But seriously now, you can bet your left kidney that if Apple ever manages to make AR/VR work they will try to milk every penny and dime from the gullible while pretending that "it's good for you" like they do with their broken walled gardens. /rant
Apple has been talking more about AR and discounting VR (like Meta’s grocery shopping and other “metaverse”) for many years now. Past keynote addresses as well as interviews with Tim Cook show that Apple didn’t consider VR alone as a game changer.
So far, it looks like Meta changed its view from VR-only and got into AR (with the new glasses). Apple is still pushing AR (with some VR).
Will Apple change its mind? Perhaps. That’s a good thing for companies to do in the face of new evidence/reception from the market.
Meta was in AR R&D before Apple. Not sure where you’re getting the impression they pivoted. They’ve always wanted to do AR but the tech isn’t there (that’s why they announced the Orion internal prototype but still don’t have a product).
Metaverse means the Ready Player One / Horizon Worlds social experience where everyone is an avatar in a shared virtual world all able to interact with each other.
Apple hasn't just not implemented this they've made it impossible for developers to as well. Because only Apple can access the sensors needed to create the 3D avatar and only Apple is allowed access to this avatar i.e. it is restricted to SharePlay with your friends.
The metaverse , in most common definitions, refers to a virtual world and ecosystem as an alternate for the real one. See Ready Player One.
Second life is a metaverse platform in essence. But isn’t VR.
Fortnite is a metaverse but isn’t VR. Their programming language is even called Verse.
The Quest itself isn’t a metaverse product BUT Horizon Worlds (not The confusingly named HorizonOS) is. You can participate in Horizon without a headset.
Apple has no metaverse product. There might be metaverse products on Apple devices but that’s no different than an iPhone or Mac playing the things I mentioned above.
There is almost no definition of metaverse that includes augmented reality.
>Second life is a metaverse platform in essence. But isn’t VR.
Second Life can and has been used with a VR headset, so yeah, it is in essence VR. Whether or not it's a good experience doesn't matter, Second Life certainly can meet the requirement for "virtual reality", depending on your definition of "VR".
There are VR mods for many games but nobody considers them VR games. But if we’re being pedantic, then “Second Life doesn’t require VR”. It predates mainstream VR headsets by a decade.
“Killer app” (as far as there is one) is focus. Being able to surround yourself with a large Mac screen & several Safari & iOS-type app screens, in a beautiful natural environment, and get real work done.
Those features are often assumed, but are NOT present in the Vision Pro.
There is only one display for apps. It’s not large. Rather tiny compared to your field of view. The features are restricted around what an iPhone can do: View photos, view movies, view your browser one tab at a time.
Immersively, yes, but anything actually productive is excluded from the Vision Pro. There is no killer app in the Vision Pro.
My 32” screen can display more windows than the Vision Pro.
Where are you getting this from? It's not correct. You can absolutely have multiple browser windows open with different pages in them and put them all around you
You can't get multiple windows from your mac though. My main complaint is that you're still subject to a single mac os x window. Although I've heard visionOS 2 is improving this haven't had a chance to try it out yet
Are you using the Vision Pro, or are you relying on 3rd party information? I use the Vision Pro every day and I have several apps open at the same constantly. That said, it is currently only one macOS screen and macOS apps don't bridge. This isn't (to me!) a huge limitation as the applications I might put on a second monitor I simply run natively. So macOS has Xcode full-screened, and I have the relevant documentation website up in a native Safari screen (to my right), YouTube (Juno app) floating above the macOS screen, and the Music app on my left.
They haven't marketed it as "metaverse" to try to please investors, it's something they've been working on for years and have now got a working initial version they've released.
Their stock crashed hard after they pivoted to focus on the metaverse. It has since more then recovered - but saying that Meta marketed their VR to please investors is kinda silly considering how Mark Zuckerberg keeps doubling down on this investment to date - despite how massively the stock crashed.
I suspect Apple’s approach will depend on whether they see LLM integration as critical or not.
If not (which I think is the way to go, a functional purpose-built accessible UI is always better than a chatbot, and generative fluff is best left to apps) then they will go for swappable LLM backends, bypassing any legal/geopolitical restrictions. Keeps eggs diversified in multiple baskets and offers negotiation leverage (and that if Apple doesn’t find a no-frills way to pass system-integrated LLM use costs onto end users).
If yes, they will probably go full in-house. They can afford it.
I see the two interlinked, similar to how modern day smartphones such as the iPhone with full web page support weren't called Web 2.0 devices, yet were integral in the proliferation and growth of Web 2.0 tech to what it is today. In the same vain, even if the highest end VR/AR device on the market isn't marketed with a "Metaverse" focus, it will be integral in whether such a thing succeeds or fails.
In my opinion, Apple Vision Pro is a device that will eventually replace iPhone, iPad and Mac. See Meta’s Orion glasses. That’s the goal of the Vision Pro, not metaverse. It just needs some time, like 5-10 years.
I hope so. But their marketing for the Pro focused a lot on trying to replace your Mac (neat) while at a cafe (a bit dystopian, so I'm ambivalent), and on the same kinds of conference calls Meta tried to push. I guess we'll see.
i can’t imagine being a a trillion dollar company doing any serious business with a company where the entire executive suite has vacated except the ceo that’s clearly doing things.
at this phase, it would be disastrous for apple to bank their reputation on sam.
Funny, though let's not forget, Google and Microsoft are willing to pay merely for default placement on Apples devices. Here, users don't even get the choice to use an alternative.
They claim that they're going to offer alternatives in the future. iOS used to have a native "share via Twitter" flow until they introduced share extensions. I imagine this will be similar.
While building lix change control I figured out that collaboration was/is the driver behind local to cloud. A secondary reason is compute.
If a system exists that enables (distributed) collaboration on files, the appeal of cloud-based software vanishes. Remote compute becomes possible as well.
I think a short-term driver towards local models is tooling. As difficult as reliably getting useful output can be, and as fragile as the resulting product can be, you want assurances of stability that don't exist with a model in someone else's cloud.
Yeah but you have to acknowledge the distance between local compute and a gazillion dollar data center, and even past that the hardware for a local setup is premium tier so it's not gonna quite catch up considering the cost.
As this is certainly not a money issue it may be in relation to privacy issues found by the Apple due diligence team at some point. Fixable - let's do it. Not fixable - let's move on to something else.
I was part of a few due diligence teams and the things we found sometimes broke the potential investment in less than a day in...
Nobody's perfect - but I think it's safe to say Apple makes a significant effort when it comes to privacy.
Apple realized that 1/OpenAI is no longer the ONLY game in town, and 2/Models != Platform. Maybe, maybe Apple can achieve what they need (Apple Intelligence), while keeping the doors open for multiple models.
No one from the list (also add Anthropic to that list) is SOTA right now, but maybe Apple wants to keep its options open as the gap closes between OAI and the others.
Last time I checked, o1 did much better in PlanBench that models from other providers indicating better perf in that area [1][2]. Probably that lead will rapidly decay as other players implement similar approaches to "reasoning"/LRMs.
o1, which you seem to be referring to, is at best slightly better at specific tasks than Sonnet 3.5 while being completely unviable economically for production use, despite the amount of subsidizing OpenAI has been doing to make them lose a cool $5 billion this year. GPT has barely improved over the last 1.5 years [1].
Yeah, I know. I didn't say it was incredibly better in every bench, but there are reasoning and planning benchmarks it does much better (not just slightly) than Sonnet 3.5 and the rest right now, and that still that makes their offering SOTA vs the others, even if its not in terms of cost.
I put sources in another sibling comment, and as I said in both comments, I think that gap will close. I don't expect OAI to keep their moat much longer, especially with the people they've lost.
PS: I say all of this as a constant critic of OAI: both how they behave and their offerings. But credit where credit is due, I don't see out-of-the-box alternatives to OAI's rlhf'd CoT at any price point right now.
Just a guess but I think Apple was just scrambling to have a foot in the door and is likely working on gradually integrating their own Apple Intelligence services as replacements.
In Apple’s presentation they specifically say that the OpenAI integration is akin to a plugin and they may have other providers, especially for domain specific knowledge in the future.
The third party providers is smart imho. Apple take on little risk, and it allows them to avoid being reliant on any one provider.
Apple is the most brand-aware IT company in the world. It is likely they simply recognized the synergy of partnering with the creators of “ChatGPT” that holds the mind space equivalently to “googling” but for AI among their customers.
That's only to weirdos like us. the general public has _maybe_ heard of openai but otherwise "ChatGPT" is 100% of their knowledge on AI and the industry
Nothing. More integrations are coming in the future and the user will be able to pick which one to use. Apple have already reportedly had talks with Google and Anthropic. They went with OpenAI first because that’s currently the best.
Making your competitor solely responsible for an important component in your ecosystem is bad news, and ClosedAI is basically an arm of Microsoft only separate to shield them from legal concerns around copyright. (And if the component is not that important, which is possibly the case with LLMs, then why spend so much money and undermine user privacy?)
My hunch is that OAI's models do not not really offer visible advantages over other open models for Apple's use cases. Even if OAI's models are more advanced, Apple can resort to cheaper post-processing to fill the gap. In Apple's scale, Apple should just host its own models to lower the long-term cost, as well as to have full control to iteratively improve the quality of the AI offerings.
Apple is on track to own everything in the vertical except chip manufacturing. It would make sense for them to invest in their products and services over something with more varied interests.
Maybe Apple just realized that they would have bought the top of a massive pump and dump scheme with the huge risk of the researchers leaving the company whilst the training and inference costs suffocates OpenAI.
Sometimes, FOMOing into a pump and dump is not worth it and they were just too late.
I don’t think Apple was too late at all. In all the hype, nobody asks “how does this help $family_member in their day to day”.
I’ve used generative AI to help with writing documents and code. I think it’s fun for creating and modifying images. As someone with no creative talent, Photoshop has been quite fun to use.
But frankly, AI is completely useless in almost every way I see it implemented. I cannot see my grandparents, my mom, or my brother use it for anything practical. Everyone’s rushing to flesh out an investor slide deck without taking a moment to think: “but how can this actually be used in someone’s day-to-day”
I’ve enjoyed Apple’s extremely conservative approach so far — they’ve kept it mostly practical.
For example, let’s take the photos app. Just being able to search for “(dogs name) lake” and having images of my dog near a lake is day-to-day useful.
Another example is their introduction of writing tools to proofread blocks of text. My parents already use Spellcheck and Grammar check, this feels like the next step up.
Compare this to Samsung who took the SV Venture-Fund-appeal-style route and blasted “AI” everywhere they can call a function or put a button. Or Microsoft, who’s Copilot story is focused on imprinting keys on keyboards without answering “how does this help the end user”
This hype around AI is gaudy. Nearly every implementation feels cheap, half-baked, and obnoxious. If Apple’s skepticism of implementing AI so far is “late”, late is not a bad thing at all as an end user.
(All that said, Genmoji looks incredibly stupid and I’m wondering what they were thinking with this…)
> This hype around AI is gaudy. Nearly every implementation feels cheap, half-baked, and obnoxious.
is true but does not imply:
> AI is completely useless in almost every way I see it implemented
which is not true. It's just harder to get real value out of it than junk. And it's a completely new "skill" to get value out of LLMs which most of us aren't learning.
But why bother learning this as a skill? The time value simply does not exist despite the tech-bubble’s hype.
Let’s look at real-world usage. Two of the most grueling tasks most people encounter in a year are filing taxes and understanding contracts. LLMs can ingest a lot of data to help in both scenarios, but whatever it advises is useless until a company accepts liability for it’s output.
I suspect we don't want it in our day to day, removing human interaction? But it is being used in day to regardless. Eg picking the next song to play.
It seems to excel at helping people do their jobs, double checking their work, etc. It's maddening since it makes constant mistakes. But Apple should be utilizing this to make HomeKit as magical as promised -- rather than automating email, which is basically the lowest hanging copy and paste able fruit available
You don’t see how an always available generative agent that could help monitor for cognitive decline and offer relief from loneliness could be helpful for the elderly?
Relief from loneliness? Maybe in the short term, but it sure seems like relying on a chat bot as a “friend” is likely to lead to even more isolation and loneliness long term, not more. Just take a look at the folks interacting with the “AI girlfriends”. It’s sad, really.
I am disappointment in the AI advancements in the recent years. It turns out progress is made not for the sake of the technology and the good for humankind, but for money. Just another pump and dump...
Is this the newest form of protest after Satya tried to burn the company to the ground? Throwing the toy away and saying "I don't want to play with this anymore, it's! It's a pump and dump"?
I don’t disagree about a bubble. A bubble for promising technology is probably inevitable. But what is this about going back to “business as usual”? Now that GenAI exists it has fundamentally changed what tasks a computer is good at and which use cases are expensive vs. cheap. So there’s no going back. Even if all models stopped improving tomorrow, we’ll be adapting our products for 10 years to take advantage of this fundamental change in capabilities.
Now all that is a long way away from creating an AI god which is what the peak hypers are selling.
A few years ago saying that there will be soon AI that perfectly understands human natural language was viewed as unrealistic science fiction. Today the possibility of an AI god is regarded as unrealistic science fiction.
Just to clarify a few things. First, there is no AI, only LLMs. Second, no LLM understands anything at all, never mind human language. Third, no LLM operates anywhere near to perfection at any task. It is still unrealistic science fiction.
Just to clarify a few things. First, there is no intelligence just evolved primates triggering neurons with chemicals. Second, no human understands anything at all they're just dumb machines reacting to stimulus. Third, no human operates anywhere near perfection at any task.
I keep watching old Star Trek TNG episodes and the way Data and the ship’s computer is portrayed seems to have been largely matched or even surpassed. And that was supposed to be 24th century tech that we caught up to in 35 years.
Is this actually related to the recent exodus? Apple is paying OpenAI nothing for use of their model on iPhones. Maybe they just realized being able to get away with paying nothing is a sign of how commoditized models will be.
At the time, I read the Apple deal as OA strategically finding an alternative to MS as the sugar-daddy, in terms of free cash flow and datacenter capabilities: it didn't need to be true, it just needed to be a compelling negotiation point with MS. The fact that Apple was outright investing in OA was in line with this view: if OA was just going to be some pluggable commodity, why invest in them? Apple doesn't do random VC for the hell of it.
So Apple withdrawing would seem to be bad news for OA, even if the rationale was something like "the valuation is now too high for our taste" rather than "we are losing faith in them as an organization" or "we think they are completely replaceable".
That isn't useful to Apple though. Apple has a lot of money.
What Apple needs more than anything else is market differentiation. They need products, patents, ideas, etc that set their products apart from their competitors. Buying into OpenAI doesn't get them that (unless they outbid everyone else). Apple will be better off using their capital to invent AI of their own, or buying/investing in an OpenAI alternative.
Apple can use their market leverage to get OpenAI integration. They don't need to spend money.
https://archive.ph/Dc6jS
This is good news. When I read the news that Apple is going to invest into OpenAI, I was a bit sad because Apple always tries to avoid to jump into the hype trains like blockchain, NFTs, metaverse etc.
Yes, there is Apple Intelligence but it is mostly writing tools, Xcode code completion, an emoji generator and so on. Nothing like something that is going to replace jobs, AGI and so on how OpenAI is marketing their products.
> Apple always tries to avoid to jump into the hype trains like blockchain, NFTs, metaverse etc.
What do you mean by "avoid ..metaverse"? They released a premium headset for AR/VR.
The knee-jerk replies you're getting are hilarious. The reality distortion field is as strong as ver. You see, because it's Apple, it can't be a bet in virtual reality with a bad connotation (metaverse). To some, Apple can do no wrong nor do they make mistakes. If they invest (and fail) in VR it was from a good angle, not purely for profit (just like Meta).
But seriously now, you can bet your left kidney that if Apple ever manages to make AR/VR work they will try to milk every penny and dime from the gullible while pretending that "it's good for you" like they do with their broken walled gardens. /rant
Apple has been talking more about AR and discounting VR (like Meta’s grocery shopping and other “metaverse”) for many years now. Past keynote addresses as well as interviews with Tim Cook show that Apple didn’t consider VR alone as a game changer.
So far, it looks like Meta changed its view from VR-only and got into AR (with the new glasses). Apple is still pushing AR (with some VR).
Will Apple change its mind? Perhaps. That’s a good thing for companies to do in the face of new evidence/reception from the market.
Meta was in AR R&D before Apple. Not sure where you’re getting the impression they pivoted. They’ve always wanted to do AR but the tech isn’t there (that’s why they announced the Orion internal prototype but still don’t have a product).
Metaverse means the Ready Player One / Horizon Worlds social experience where everyone is an avatar in a shared virtual world all able to interact with each other.
Apple hasn't just not implemented this they've made it impossible for developers to as well. Because only Apple can access the sensors needed to create the 3D avatar and only Apple is allowed access to this avatar i.e. it is restricted to SharePlay with your friends.
The “metaverse” isn’t tied to AR/VR.
The metaverse , in most common definitions, refers to a virtual world and ecosystem as an alternate for the real one. See Ready Player One.
Second life is a metaverse platform in essence. But isn’t VR.
Fortnite is a metaverse but isn’t VR. Their programming language is even called Verse.
The Quest itself isn’t a metaverse product BUT Horizon Worlds (not The confusingly named HorizonOS) is. You can participate in Horizon without a headset.
Apple has no metaverse product. There might be metaverse products on Apple devices but that’s no different than an iPhone or Mac playing the things I mentioned above.
There is almost no definition of metaverse that includes augmented reality.
>Second life is a metaverse platform in essence. But isn’t VR.
Second Life can and has been used with a VR headset, so yeah, it is in essence VR. Whether or not it's a good experience doesn't matter, Second Life certainly can meet the requirement for "virtual reality", depending on your definition of "VR".
There are VR mods for many games but nobody considers them VR games. But if we’re being pedantic, then “Second Life doesn’t require VR”. It predates mainstream VR headsets by a decade.
should mention VRChat in that list too, imo
That isn’t the metaverse.
No dancing Zucks.
“Killer app” (as far as there is one) is focus. Being able to surround yourself with a large Mac screen & several Safari & iOS-type app screens, in a beautiful natural environment, and get real work done.
Those features are often assumed, but are NOT present in the Vision Pro.
There is only one display for apps. It’s not large. Rather tiny compared to your field of view. The features are restricted around what an iPhone can do: View photos, view movies, view your browser one tab at a time.
Immersively, yes, but anything actually productive is excluded from the Vision Pro. There is no killer app in the Vision Pro.
My 32” screen can display more windows than the Vision Pro.
Where are you getting this from? It's not correct. You can absolutely have multiple browser windows open with different pages in them and put them all around you
You can't get multiple windows from your mac though. My main complaint is that you're still subject to a single mac os x window. Although I've heard visionOS 2 is improving this haven't had a chance to try it out yet
> Those features are often assumed, but are NOT present in the Vision Pro.
Yes they are.
Source: I develop for the device.
Ok. Well that’s possibly under NDA, because in the Apple Store demo they say it’s impossible.
But in any case: Several browsers isn’t several apps. It’s really not like having several computers screens around you.
Are you using the Vision Pro, or are you relying on 3rd party information? I use the Vision Pro every day and I have several apps open at the same constantly. That said, it is currently only one macOS screen and macOS apps don't bridge. This isn't (to me!) a huge limitation as the applications I might put on a second monitor I simply run natively. So macOS has Xcode full-screened, and I have the relevant documentation website up in a native Safari screen (to my right), YouTube (Juno app) floating above the macOS screen, and the Music app on my left.
They haven't marketed it as "metaverse" to try to please investors, it's something they've been working on for years and have now got a working initial version they've released.
For the record: neither has Meta/Facebook.
Their stock crashed hard after they pivoted to focus on the metaverse. It has since more then recovered - but saying that Meta marketed their VR to please investors is kinda silly considering how Mark Zuckerberg keeps doubling down on this investment to date - despite how massively the stock crashed.
Metaverse isn't VR. Meta in particular just happens to be integrating them since they make both.
I suspect Apple’s approach will depend on whether they see LLM integration as critical or not.
If not (which I think is the way to go, a functional purpose-built accessible UI is always better than a chatbot, and generative fluff is best left to apps) then they will go for swappable LLM backends, bypassing any legal/geopolitical restrictions. Keeps eggs diversified in multiple baskets and offers negotiation leverage (and that if Apple doesn’t find a no-frills way to pass system-integrated LLM use costs onto end users).
If yes, they will probably go full in-house. They can afford it.
> metaverse
What about the Apple Vision Pro?
What about it? VR != Metaverse.
(FWIW I think both are way overhyped)
I see the two interlinked, similar to how modern day smartphones such as the iPhone with full web page support weren't called Web 2.0 devices, yet were integral in the proliferation and growth of Web 2.0 tech to what it is today. In the same vain, even if the highest end VR/AR device on the market isn't marketed with a "Metaverse" focus, it will be integral in whether such a thing succeeds or fails.
In my opinion, Apple Vision Pro is a device that will eventually replace iPhone, iPad and Mac. See Meta’s Orion glasses. That’s the goal of the Vision Pro, not metaverse. It just needs some time, like 5-10 years.
I hope so. But their marketing for the Pro focused a lot on trying to replace your Mac (neat) while at a cafe (a bit dystopian, so I'm ambivalent), and on the same kinds of conference calls Meta tried to push. I guess we'll see.
i can’t imagine being a a trillion dollar company doing any serious business with a company where the entire executive suite has vacated except the ceo that’s clearly doing things.
at this phase, it would be disastrous for apple to bank their reputation on sam.
> Apple reportedly believes that the exposure the iOS 18 integration is giving OpenAI is “of equal or greater value” than cash.
So it's not just gig workers who get offered exposure to pay the bills.
Funny, though let's not forget, Google and Microsoft are willing to pay merely for default placement on Apples devices. Here, users don't even get the choice to use an alternative.
Alternative for what? I don’t use default Microsoft for anything on Apple, or recall seeing anything Microsoft as a default choice.
Google was default search engine in Safari, but I changed that to DDG.
Alternative for the AI service provider.
They claim that they're going to offer alternatives in the future. iOS used to have a native "share via Twitter" flow until they introduced share extensions. I imagine this will be similar.
I think you mean artists.
Gig workers are paid in peanuts and dreams.
The future will be local models. Let these investors burn the money.
I wish, but why do you think so? Last decade the flow was from local compute to the server, not the other way around.
While building lix change control I figured out that collaboration was/is the driver behind local to cloud. A secondary reason is compute.
If a system exists that enables (distributed) collaboration on files, the appeal of cloud-based software vanishes. Remote compute becomes possible as well.
At least, that’s the bet we are taking with lix https://lix.opral.com/
I think a short-term driver towards local models is tooling. As difficult as reliably getting useful output can be, and as fragile as the resulting product can be, you want assurances of stability that don't exist with a model in someone else's cloud.
I'm not an expert on anything, though.
Agree. What’s your reasoning?
Apple is already doing this:
https://www.apple.com/uk/apple-intelligence/
Yeah but you have to acknowledge the distance between local compute and a gazillion dollar data center, and even past that the hardware for a local setup is premium tier so it's not gonna quite catch up considering the cost.
As this is certainly not a money issue it may be in relation to privacy issues found by the Apple due diligence team at some point. Fixable - let's do it. Not fixable - let's move on to something else.
I was part of a few due diligence teams and the things we found sometimes broke the potential investment in less than a day in...
Nobody's perfect - but I think it's safe to say Apple makes a significant effort when it comes to privacy.
Apple realized that 1/OpenAI is no longer the ONLY game in town, and 2/Models != Platform. Maybe, maybe Apple can achieve what they need (Apple Intelligence), while keeping the doors open for multiple models.
> Apple realized that 1/OpenAI is no longer the ONLY game in town
Who is the other game in town for them?... Meta? Google?
Anthropic as well
No one from the list (also add Anthropic to that list) is SOTA right now, but maybe Apple wants to keep its options open as the gap closes between OAI and the others.
> No one from the list (also add Anthropic to that list) is SOTA right now
I don't use OpenAI's products. Is there an area where they are clearly ahead? I thought it was at least a tie in all categories.
Last time I checked, o1 did much better in PlanBench that models from other providers indicating better perf in that area [1][2]. Probably that lead will rapidly decay as other players implement similar approaches to "reasoning"/LRMs.
[1] https://www.arxiv.org/abs/2409.13373
[2] https://x.com/omarsar0/status/1838353480672563581
o1, which you seem to be referring to, is at best slightly better at specific tasks than Sonnet 3.5 while being completely unviable economically for production use, despite the amount of subsidizing OpenAI has been doing to make them lose a cool $5 billion this year. GPT has barely improved over the last 1.5 years [1].
[1] https://aider.chat/assets/models-over-time.svg
Yeah, I know. I didn't say it was incredibly better in every bench, but there are reasoning and planning benchmarks it does much better (not just slightly) than Sonnet 3.5 and the rest right now, and that still that makes their offering SOTA vs the others, even if its not in terms of cost.
I put sources in another sibling comment, and as I said in both comments, I think that gap will close. I don't expect OAI to keep their moat much longer, especially with the people they've lost.
PS: I say all of this as a constant critic of OAI: both how they behave and their offerings. But credit where credit is due, I don't see out-of-the-box alternatives to OAI's rlhf'd CoT at any price point right now.
What makes openai special in comparison to its competitors that apple wants to integrate their service?
Just a guess but I think Apple was just scrambling to have a foot in the door and is likely working on gradually integrating their own Apple Intelligence services as replacements.
In Apple’s presentation they specifically say that the OpenAI integration is akin to a plugin and they may have other providers, especially for domain specific knowledge in the future.
The third party providers is smart imho. Apple take on little risk, and it allows them to avoid being reliant on any one provider.
Apple is the most brand-aware IT company in the world. It is likely they simply recognized the synergy of partnering with the creators of “ChatGPT” that holds the mind space equivalently to “googling” but for AI among their customers.
Also being brand aware would probably put it off the recent goings on at Open/ClosedAI.
That's only to weirdos like us. the general public has _maybe_ heard of openai but otherwise "ChatGPT" is 100% of their knowledge on AI and the industry
Nothing. More integrations are coming in the future and the user will be able to pick which one to use. Apple have already reportedly had talks with Google and Anthropic. They went with OpenAI first because that’s currently the best.
Google is a major investor in Anthropic.
Musk is too much of a loose cannon for Apple so Grok is out.
Mistral is in the EU which doesn't have a stable regulatory regime.
Leaving OpenAI which at least has the full support of Microsoft i.e. it will always have access to the best compute resources and endless money.
> Mistral is in the EU which doesn't have a stable regulatory regime.
I guess the US's "no regulations ever" can be considered a stable regulatory regime.
I think "predictable" would have been a better word than "stable" here.
The word they were really looking for was "favorable".
Yeah, I guess "stable" works too, though.
Making your competitor solely responsible for an important component in your ecosystem is bad news, and ClosedAI is basically an arm of Microsoft only separate to shield them from legal concerns around copyright. (And if the component is not that important, which is possibly the case with LLMs, then why spend so much money and undermine user privacy?)
> Google is a major investor in Anthropic.
And Microsoft is an even bigger investor in OpenAI. Why couldn't Apple join Google in this?
There's also Meta, whose money is as endless as Microsoft's.
They have a deep hatred for Meta, so Llama is out
Almost certainly because of the sheer popularity of the ChatGPT brand and majority of user/mind share.
My hunch is that OAI's models do not not really offer visible advantages over other open models for Apple's use cases. Even if OAI's models are more advanced, Apple can resort to cheaper post-processing to fill the gap. In Apple's scale, Apple should just host its own models to lower the long-term cost, as well as to have full control to iteratively improve the quality of the AI offerings.
Apple is on track to own everything in the vertical except chip manufacturing. It would make sense for them to invest in their products and services over something with more varied interests.
They aren’t doing that either. Over 80Bn in cash, they’re doing their best Smaug impression.
Maybe Apple just realized that they would have bought the top of a massive pump and dump scheme with the huge risk of the researchers leaving the company whilst the training and inference costs suffocates OpenAI.
Sometimes, FOMOing into a pump and dump is not worth it and they were just too late.
I don’t think Apple was too late at all. In all the hype, nobody asks “how does this help $family_member in their day to day”.
I’ve used generative AI to help with writing documents and code. I think it’s fun for creating and modifying images. As someone with no creative talent, Photoshop has been quite fun to use.
But frankly, AI is completely useless in almost every way I see it implemented. I cannot see my grandparents, my mom, or my brother use it for anything practical. Everyone’s rushing to flesh out an investor slide deck without taking a moment to think: “but how can this actually be used in someone’s day-to-day”
I’ve enjoyed Apple’s extremely conservative approach so far — they’ve kept it mostly practical.
For example, let’s take the photos app. Just being able to search for “(dogs name) lake” and having images of my dog near a lake is day-to-day useful.
Another example is their introduction of writing tools to proofread blocks of text. My parents already use Spellcheck and Grammar check, this feels like the next step up.
Compare this to Samsung who took the SV Venture-Fund-appeal-style route and blasted “AI” everywhere they can call a function or put a button. Or Microsoft, who’s Copilot story is focused on imprinting keys on keyboards without answering “how does this help the end user”
This hype around AI is gaudy. Nearly every implementation feels cheap, half-baked, and obnoxious. If Apple’s skepticism of implementing AI so far is “late”, late is not a bad thing at all as an end user.
(All that said, Genmoji looks incredibly stupid and I’m wondering what they were thinking with this…)
From my perspective,
> This hype around AI is gaudy. Nearly every implementation feels cheap, half-baked, and obnoxious.
is true but does not imply:
> AI is completely useless in almost every way I see it implemented
which is not true. It's just harder to get real value out of it than junk. And it's a completely new "skill" to get value out of LLMs which most of us aren't learning.
But why bother learning this as a skill? The time value simply does not exist despite the tech-bubble’s hype.
Let’s look at real-world usage. Two of the most grueling tasks most people encounter in a year are filing taxes and understanding contracts. LLMs can ingest a lot of data to help in both scenarios, but whatever it advises is useless until a company accepts liability for it’s output.
I suspect we don't want it in our day to day, removing human interaction? But it is being used in day to regardless. Eg picking the next song to play.
It seems to excel at helping people do their jobs, double checking their work, etc. It's maddening since it makes constant mistakes. But Apple should be utilizing this to make HomeKit as magical as promised -- rather than automating email, which is basically the lowest hanging copy and paste able fruit available
You don’t see how an always available generative agent that could help monitor for cognitive decline and offer relief from loneliness could be helpful for the elderly?
Relief from loneliness? Maybe in the short term, but it sure seems like relying on a chat bot as a “friend” is likely to lead to even more isolation and loneliness long term, not more. Just take a look at the folks interacting with the “AI girlfriends”. It’s sad, really.
Can you say more about pump and dump?
I am disappointment in the AI advancements in the recent years. It turns out progress is made not for the sake of the technology and the good for humankind, but for money. Just another pump and dump...
Is this the newest form of protest after Satya tried to burn the company to the ground? Throwing the toy away and saying "I don't want to play with this anymore, it's! It's a pump and dump"?
"It turns out?" "In recent years?" It has forever been so.
AI in its current state is a bubble. It's going to pop soon, and then we'll get back to business as usual.
I don’t disagree about a bubble. A bubble for promising technology is probably inevitable. But what is this about going back to “business as usual”? Now that GenAI exists it has fundamentally changed what tasks a computer is good at and which use cases are expensive vs. cheap. So there’s no going back. Even if all models stopped improving tomorrow, we’ll be adapting our products for 10 years to take advantage of this fundamental change in capabilities.
Now all that is a long way away from creating an AI god which is what the peak hypers are selling.
A few years ago saying that there will be soon AI that perfectly understands human natural language was viewed as unrealistic science fiction. Today the possibility of an AI god is regarded as unrealistic science fiction.
Just to clarify a few things. First, there is no AI, only LLMs. Second, no LLM understands anything at all, never mind human language. Third, no LLM operates anywhere near to perfection at any task. It is still unrealistic science fiction.
Just to clarify a few things. First, there is no intelligence just evolved primates triggering neurons with chemicals. Second, no human understands anything at all they're just dumb machines reacting to stimulus. Third, no human operates anywhere near perfection at any task.
Wishful thinking.
I keep watching old Star Trek TNG episodes and the way Data and the ship’s computer is portrayed seems to have been largely matched or even surpassed. And that was supposed to be 24th century tech that we caught up to in 35 years.
Todays models don't hold a candle to data, they are better at reading emotions yes but everything else they are horrible at in comparison.
Sci fi writers just assumed that emotions would be hard for computers but apparently it isn't that hard at all compared to rational thinking.
We're nowhere near Data, or ship's computer, level of AI.
Data has the ability to be accurate and store and index all the information, not present a lossy copy where he hallucinates.
Is this a joke?
I take it you mean that like the dotcom bubble? As in The Internet didn't go away, but the investors returned to a little more sensible thinking.
LLMs will continue to be used by increasingly more developers
[dead]
I'd agree with you except that their product is insanely good compared to the rest of the world, and they're leading into a new space as well.
Sure the bubble will pop with the hundreds of nonsense startups going on at the moment, but OpenAI will be fine.
Soon is what timeline? It's still getting better; as long as that happens, no one is stopping.
What does this even mean, really?
Seed rounds for Ai startups with nothing more than a napkin saying "do Ai stuff" will have valuations less than a billion.
It means that the return of investment for fancy email autocomplete is not that big
Please keep the cheeky reductionist takes to yourself
Is this actually related to the recent exodus? Apple is paying OpenAI nothing for use of their model on iPhones. Maybe they just realized being able to get away with paying nothing is a sign of how commoditized models will be.
At the time, I read the Apple deal as OA strategically finding an alternative to MS as the sugar-daddy, in terms of free cash flow and datacenter capabilities: it didn't need to be true, it just needed to be a compelling negotiation point with MS. The fact that Apple was outright investing in OA was in line with this view: if OA was just going to be some pluggable commodity, why invest in them? Apple doesn't do random VC for the hell of it.
So Apple withdrawing would seem to be bad news for OA, even if the rationale was something like "the valuation is now too high for our taste" rather than "we are losing faith in them as an organization" or "we think they are completely replaceable".
If AMD can do it, they can as well: https://news.ycombinator.com/item?id=41674382
You have to feel that apple is missing out on buying into this investment as it will clearly become very valuable.
That isn't useful to Apple though. Apple has a lot of money.
What Apple needs more than anything else is market differentiation. They need products, patents, ideas, etc that set their products apart from their competitors. Buying into OpenAI doesn't get them that (unless they outbid everyone else). Apple will be better off using their capital to invent AI of their own, or buying/investing in an OpenAI alternative.
Apple can use their market leverage to get OpenAI integration. They don't need to spend money.
The alternative is to invest the money into their own AI which will probably also become valuable.
I think they have enough money to do both
If there is no moat on this stuff, what’s the point of buying it? Just build your own. They’re better off investing in people.
[flagged]
So Nvidia and Microsoft are not big tech companies?
Don't mistake Apple's actions as some industry sentiment. They've always done things different. Time will tell who was right.