First Windsurf and now this. OpenAI is spending billions like there's nothing else to use this money for while being seemingly cash strapped for model training since they already signaled more investment rounds would be needed to remain competitive. They're trying to become too big to fail before they have a moat which won't work well.
They've already claimed that there will be no "GPT-5" LLM, and that instead what they want to call "GPT-5" is a fusion of their various models like 4o, dalle, their video model, etc. That in and of itself is a move that makes it quite clear to me they've hit a wall on the intelligence side.
Add these purchases, and it seems like they are extremely desperate.
Models are getting smaller, faster, cheaper to make, reflecting on their own output, adding modes and running in more places. But they’re not getting much smarter because they can only be as smart as us and each other, because that’s where their training comes from. OpenAI is strongest in a world where models cost billions to train. A world filled with cheap open source models is their worst nightmare. This is what’s happening. So they have to pivot into being a product company and away from being a model company.
> But they’re not getting much smarter because they can only be as smart as us and each other,
That doesn't look to be true in general. AlphaGoZero didn't learn off smarter humans or smarter AI's (at all - it only trained against itself), yet it became better at playing some games than any existing AI or human.
To me it looks like the same thing has happened for LLM's in the one area they are truly good at: natural language processing. Admittedly they only learned to mimic human language by begin fed lots of human language, but they look at least as good at parsing and writing as any human now, and much, much faster at it. And admittedly they have plateaued at natural language processing. But that's not because of any inherent limitation in the level of intelligence an AI can achieve. It's because unlike playing Go there is a natural limit on good how you can get at mimicking anything, which is "indistinguishable".
The other things LLM's seem to be good at a lossy compression of all the text they have been trained on. I was floored when I ran a 16GB locally, and it could tell me things about my childhood town (pop: under 1000, miles away from anywhere). It didn't know a lot, but there isn't a lot out there about it on the internet, and it still astounds me it could compress the major points of everything it read on the internet down to 16GB. The information it regurgitated was sometimes wrong of course, but then you only expect to get a overview of a scene from a highly compressed JPEG. The details will be blurry or downright misleading.
What they are attempting to tack onto that is connecting the facts the LLM knows into a chain of thought. LLM aren't very good at that, and the improvements over the past few years look to be marginal, yet that is what is being hyped with the current models.
None of that detracts from your main point, which I think boils down to the rapid advancements in proprietary models have stalled. Their open source competitors aren't far behind, and if they have really stalled open source will catch up.
But that's only true for the natural language processing side. The shear compute required to keep a model up to date with the latest information in the internet means the model with the most resources behind it will regurgitate the most accurate information about what's on the internet today. Open source will always lose that race.
Are you aware, what hardware is currently needed to run GPT4?
Something bigger than a smartphone usually.
So small mobile optimized LLMs will come, or are rather already there - but if they would manage to make the big GPT4 modell run on an iPhone, that would be a pretty big thing in itself, way larger than GPT5.
But llms are relatively rarely used, and on the other hand, perf/latency is important to ux, and perf is variable(simple question, complex question, visual work).
Userbase and customer relationships are valuable. If someone else creates GPT5, but doesn't have a large user base, then OpenAI the company could buy that invention. Or, as we saw with deepseek in January, fast-follow with a comparable model within a reasonable amount of time.
Brands have value. If someone has logged into ChatGPT for two years daily, they have built a habit. That habit certainly can be disrupted, but there's a level of inertia and barrier -- something else has to be 10x better and not just 2x better.
When DeepSeek came out, I tried it out but didn't fundamentally switch my habit. OpenAI + Claude + Gemini instead caught up.
They're not acquiring Jony or Jony's design firm. They're acquiring the remaining portion of a joint venture. You could even say that LoveFrom is divesting from the joint venture.
Following that logic, they’ll have to keep spending quite a bit to get to the user base of the current hyperscalers, some of which are already ahead of OpenAI in terms of LLM performance.
OpenAI would not be able to as every other company and governments even will make bids and OpenAI is not well loved to get favor to tilt the scale back in their direction.
There is a space to make a suite of products that synergize entirely. Glasses, watches, buttons, clothes (yes, clothes), and home devices/computers/tvs. The reason they are in a spearhead position is because unlike like Google and Apple, they don't need to maintain a legacy paradigm. They don't have to introduce new tech and make it work with old tech, while also maintaining usability familiarity (e.g You can't just change iOS and Android).
They take zero risk while attacking user fatigue (people just get bored of stuff). The current leaders take all the risk following OpenAI because everyone will complain about the changes no matter what they do, and just come up with a reason to switch. This is a human phenomenon that is truly fucked up, the same as when a partner in a relationship is ready to move on no matter what you do.
More like they see the future as more multi-modal, and they're probably right to think that is the best value approach vs. throwing more money at large language models.
I'm not so sure it's desperation. As an alternative hypothesis, we might simply view it as an attempt from a temporary position of strength to secure their tremendous lead as the primary consumer access point to intelligence. I don't think it's much of an exaggeration to suggest that this is one of the most important open questions at the moment -- one which will likely be relatively winner-takes-all (in contrast to the more commoditized B2B/API side) and where the winner likely won't be decided based on the intelligence side alone. The questions also aren't entirely separate since the winner, here, will have such incomparably valuable usage data...
Unlike most successful startups, OpenAI is not faced with the possibility that the giants (Apple, Google, Microsoft) decide to look their way, but the reality that these are their real competitors and that the stakes are existential for many of them (trends indicating a shift away from search etc). The most likely outcome remains that one if not all of the giants eventually manage to produce a halfway-decent product experience that reduces OpenAI to a B2B player.
> I don't think it's much of an exaggeration to suggest that this is one of the most important open questions at the moment -- one which will likely be relatively winner-takes-all
That makes the presumption that we are currently in a `winner-takes-all` scenario, and I'm not convinced that that is the case.
I'm not sure what the criteria is for a winner-takes-all scenario, but it is not at all evident to me that there is one now, or ever will be.
There is, as everyone says, no actual moat here: Google search had a moat, Windows Desktop had a moat, Apple phones had (and still have) a moat. LLM output currently has no moat, not even performance (both speed and accuracy) because the productivity difference between no-LLM and poor-LLM is about 100x the difference between poor-LLM and good-LLM.
My prediction is that the price of LLM usage will slowly but consistently climb until it reaches the floor on LLM cost-to-suppliers. Right now we are all (myself included) being subsidised by VC money. When the supplier has to actually turn a profit, there's no moat that they can use to keep out newcomers, because the newcomers need only a fraction of the money spent by (for example OpenAI) in order to compete.
Maybe Google has a moat, in that they have everything in-house, from the user-facing product to the tensor-processing hardware? That's as close to a moat that I can think off.
I don't think your conclusion of "hitting the wall on intelligence" is warranted.
It makes more sense to believe that scaling has hit the wall on available text data to train on, and that to continue scaling, along with whatever emergent properties arise they need much more data than exists as text.
There are orders of magnitude more data as video, audio, and images and this is what they intend to use to continue scaling.
It kind of is, when they were given $500B and told to make a return in 10-ish years. They have to put the capital in play where it has the largest ROI potential. They are gambling that Jony has another iPhone in him.
I don't know enough about any of this to weigh in on it, but when you take investor money, you aren't supposed to sit on it or do slow burn (at least not VC money), its meant to be gasoline, and you moonshot with it.
> They are gambling that Jony has another iPhone in him.
I seriously doubt it.
If anything because Apple let him go exactly when they were looking for a new hit product like the iPhone.
But also because how he handled the Mac the years before he was fired. All his big decisions were just bad. The butterfly keyboard, touchbar, USB-C only ports, etc. Heck even the 2013 Mac Pro (the trashcan) was an engineering failure. They could never upgrade it because, according to Craig Federighi, they got themselves into a thermal corner caused by the design of it[1].
He transformed the MacBook Pro into its pure essence, the ultimate form, meant to be used on a slab of polished granite in a HEPA filtered room, with only a precisely aligned array of dongles to offset the clean vision.
The fact the you took your laptop out in the field or to a couch in some barn like a filthy animal, corrupting perfection with dust and grease, rendering the keyboard useless is on you. It is a reflection of your own animal nature.
Every Macbook user I've seen seems to have an ultra smudged fingerprint screen as if they were using an iPad with so many layers of oil it shimmers as a rainbow, with sprinkles with various unique flakes of dandruff, and dirt encrusted on their trackpad and keyboard. It stirs up conniptions.
This gave me an idea: a Slack AI that will give me an analogy to support my point, whatever it may be.
“Hey Analogai, help me out here.”
“Ah I see what Chip Frumpkins, Director of Looking Relevant is saying. It’s basically that we need to throw a lot of paint at the wall to see what sticks. And if we fail, at least we’ve got a Jackson Pollock.”
That analogy breaks down almost immediately. I get your point that when you go out and try to do things sometimes you will fail, but the problem is that many of his design failures were seen _even at the time_ to be failures.
I don't necessarily think Ive is going to succeed, but if you're going to make a lot of bets, taking one bet on someone who succeeded before seems pretty reasonable. He wouldn't be the first person to rise to great heights, fall, and rise again, even in the Apple world.
I absolutely agree right up until we start talking about price. Obviously this deal was all in stock from someone who has a history creative corporate control structures, but nevertheless the on paper cost of was $6.4 billion. That's a hell of a bet.
since we're torturing the analogy... you don't measure a baseball team's success by the # of HR's one player records in a season, you talk championships over time. Sometimes they're related, but less frequently than you'd think.
Baseball is the hardest sport, but it’s a zero sum game. The .300 batting average is against equally elite pitching. Engineering or design is about adding value.
Taking the raw engineering of the components and interfaces that defined the iPhone and making a system of it is design at its peak and almost art.
Taking a proven form factor like a laptop, not talking to users and making it worse is just a misstep. It wasn’t a complete disaster only because the bar is so low, the defective Apple laptop is still the best laptop in the market.
You know, when we upgraded to USB-C I thought they were mostly nifty. Reversible, quite universal, fully embraced by everyone.
But over 7 years of using them, I've come to resent some of their differences with past USB connectors. Very small, insecure friction grip, reversible, more delicate.
Also it seems that device designers think that a newer generation of USB needs fewer ports? My Lenovo ThinkPad had 2x USB-A and 2x USB-C in 2018. Now I've got a Pixel with 1x USB-C and a Chromebook with 1x USB-A and 2x USB-C; on each of those devices you need one port for charging. So if USB is more versatile and compatible than ever, why am I not allowed to plug in all my stuff at once?
The trash can was a bet on external GPU enclosures, which are technically feasible but just never took off in the marketplace. It was great engineering for a use case that just didn’t pan out.
As one of the fervent 500 million daily GPT users it’s a no brainer for Open AI to create a personal mobile AI device or an AI phone with GPT accessible right from your Lock Screen.
It could…
- interface with AI Agents( businesses, friends & family’s agents, etc) to get things done & used as a knowledgeable.
Once u pick up the device it’s like a FaceTime call with your AI agent/friend in which u can skin to look however u want (a deceased loved one ..tho that might be too out there).
- It visually surfs the Web with you.. making u not open a web browser as much
- take the best selfies of u…gets you to the best lighting.
Overall excited to see their vision and leave/drop Apple’s now boring iPhone for a GPT phone or personal mobile AI device. I think a phone form factor would be best, but we’ll see.
How did that work for the Facebook phone? And all their billions of fervent Facebook users?
Google own this space - pixel phones already do pretty much all of this, and they have the best models and the most users too. No built in agentic capabilities yet, but I am sure that is just a month or two away (see project mariner).
If you've not tried the pixel photo ai features already, you may be surprised. Things like changing lighting, removing people from the shot, auto-stitching people into a group photo, composites group photos so you get one photo where everyone is smiling and looking at the camera at the same time even if that never happened etc, text-editing photos etc. Gemini live is like a facetime call without the 3d avatar but we've seen they can do it with Veo3 already if they wanted.
This is all reality today already in the hands of billions of Google users, so OpenAI have a bit of a hill to climb: OpenAI would need to not only catch up with Google (both in AI space which they seem incapable of doing right now but also in product too) and surpass them.
Google are totally integrated in this space - the device, the software, the AI models, the infrastructure, the data, the sites/apps people use (search, Gmail, maps, YouTube, docs, ...) and also the billions of users.
I doubt OpenAI can really make a dent here. I suspect any OpenAI-Phone will be quietly discontinued like the Facebook phone
What I was describing is basically H.E.R. the movie on your GPT phone or personal mobile AI device (a hologram.. maybe they are going for). To me it's a new paradigm driven by AI as your friend/assistant unlike Facebook making a social network phone. As well to me GPT feels like how Twitter and Google felt when they started to slowly change the world in their own ways. Do you use GPT frequently throughout your day?
You pick up your GPT mobile device and the UX is a FaceTime call with a real life looking AI person who does everything for you(you can skin it to look like anyone including a deceased relative.. they live on & help you thru life). You rarely will need to go to the web ... your AI friend / agent / assistant could bring up the web right within the FaceTime call yet visual the data you seek. They can take the best selfies of you ... direct to the best lighting within your living space at the time. As noted your AI friend will interface with AI Agents of businesses, your friends & family to schedule things and used as a knowledge base (want to know your cousin birthday ask your agent and if you cousin shares that with family members your agent will tell you via their agent).
You are saying Google just announced a H.E.R. phone or personal AI mobile device where the AI is the focus (apps and the web take a backseat)? As Im describing above?
They stopped the glasses with tiny fov that weren't useful, but they have all (FB, Apple, Google, multiple startups like North, Xreal) been working on more subtle glasses continuously since then, it's just been hard and have needed display/power breakthroughs. Google announced new glasses again yesterday. Looks promising but live demo sputtered out at the end, still not gonna be ready for a while
Meta Ray Bans for sunglasses wearer who takes pics & videos using their phone are very handy; no need to take out or even have your phone to do either. Can also ask it for the time without needing your phone too.
You & just a dude here with ideas that in time go nowhere or maybe somewhere. Altman noted different demographics use GPT differently with 18 to 20 somethings not making decisions without consulting GPT (could be marketing speak but with some truth).
The first iPhone was a hardware engineering marvel, it was leaps and bounds more premium than any phone in that generation. It took other companies years to catch up.
I’m unsure what he’s trying to say either. The gibberish and out-of-context replies ITT are making me think HN, like many other sites, is laden with bots now.
How much credit does he really deserve for the iPhone? Jobs and Fadell were obviously both involved in the iPhone too and Nest has some pretty appealing design without Jony being involved at all.
Source? They did a $40B funding round for which Softbank is on the hook for most of it, and they are going into debt cause they don't have the cash either [0]. IMO, these acquisitions are due to the fact they know just selling the model isn't where the huge margins are, selling the verticals is.
Ive is a good designer sure but "essentially created the iPhone" is absurd. It took thousands of engineers and product visionaries to bring that device together, and OpenAI isn't getting any of that. You aren't going to replicate its success by hiring the guy whose major contribution was insisisting that all Apple products be a few millimeters thinner in every iteration.
Completely agree. He is a good designer, but graphical UX went downhill when he was given more control at Apple and he became increasingly militant about hardware design to the point that the MacBook Pro was kinda bad because it was unreasonably thermal limited and had a terrible keyboard.
Ive is basically the best in the business if your needs are to get a large amount of cutting-edge technology into a ridiculously constrained form factor and have it look good, feel solid, and be manufacturable at scale.
That is what he is world-class at. Not designing comprehensive product experiences or ideating new greenfield products (and definitely not designing app icons).
If IO or OpenAI also has a product visionary of the caliber to fully utilize Ive's singular industrial design talent, they'll rule the world. Otherwise, they're sinking billions into the next Humane Pin.
I don’t think there’s any evidence that Ive has the expertise you claim. He was lead designer for Apple when they did the iPhone, but it is Apple who has the extensive deep expertise in hardware design and engineering.
You're 100% right and in my opinion there is a much higher probability this is a total waste and nothing of similar value will be created. But if you're OpenAI and you have this option I also see why you may take it.
Lots of companies were making smart phones with similar specs. Under Johnny Ive's management, the Apple version made a much bigger splash than any of them and defined the category.
At the time of the public debut of the first generation iPhone (January 2007), the statement "lots of companies were making smart phones with similar specs" is objectively false. Further, there were zero companies making comparable large touchscreen, large cpu phones outside of Apple at the time.
Arguably HP/Palm's WebOS devices were ahead on every mark - easier to use, more featureful, smarter, better physical design than any iPhone of similar manufacturing date.
The difference was management choosing to stick with a platform for long enough for network effects to kick in.
If Apple has any advantages compared to other big tech, it's an ability to look past next quarter's financials.
Palm offerings in 2007, such as the Treo 755p or the Centro, could not compete hardware-wise with the original iPhone. The claim that these Palm phones were "easier to use" is hilarious to me, and probably hilarious to many others.
I explicitly mentioned WebOS, meaning the devices released around 2009, which competed with 1st gen iPhone old stock, and directly against iPhone 3G - the second generation.
The first gen iPhone is not a smartphone by today's standards. No multitasking, no copy/paste, no centralized instant messaging, all things WebOS devices had on release.
Even the second generation of iPhones felt half baked by comparison.
Which just goes to illustrate my point, that they weren't technologically superior, just more committed.
The race was already over by the time webOS showed up. Even Microsoft, with a superior product and many billions spent pushing it, couldn't overcome the network effects of iOS and Android. No one else had a chance.
Comcast and AT&T still exist. Kraft still turns out war rations by the warehouse. Tasteless grocery store tomatoes are still the most widely available.
LuneOS - the direct continuation of WebOS development had a release in Feb 2024 and is still under active development: https://webos-ports.org/wiki/Main_Page
Pretty good for something that supposedly failed 12 years ago.
> Jony Ive, a chief architect of the iPhone, and his design firm are taking over creative and design control at OpenAI, where they will develop consumer devices and other projects that will shape the future look and feel of AI.
> Ive won’t be joining OpenAI, and his design firm, LoveFrom, will continue to be independent, but they will “take over design for all of OpenAI, including its software,” in a deal valued at nearly $6.5 billion
> OpenAI already owns a 23% stake in io as part of a deal from last year, meaning it needs about $5 billion for the acquisition, the Times reported, citing unnamed people familiar with the matter.
“As part of the deal, OpenAI is paying $5 billion in equity for io. The balance of the nearly $6.5 billion stems from a partnership reached in the fourth quarter of last year that involved OpenAI acquiring a 23% stake in io.”
Honestly I think it's a great move if you know you have a hyped up valuation, to exchange that paper valuation for actual company acquisitions. Not every company has that ability.
I have a gut feeling alot of this is going to go negative for OpenAI. I simply don't see what they're going to produce in a reasonable amount of time that justifies hardware, for example.
I'm open to being wrong, very open, but I need to see evidence. Hard evidence.
> You don't hear anyone touting AGI anymore do we?
Apart from, y'know, DeepMind - remember those guys? The ones with the SOTA models at the top of the leaderboards? The ones who just launched Veo3 and blew everyone away?
It feels like OpenAI has kinda jumped-the-shark at this stage. They don't seem to be especially competitive any more, and all the news coming out of them is tinkering at the edges or acquisitions that no-one really cares about.
When are they going to start competing on actual AI again?
After a lot of the drama and a ton of talent leaving all they seem to have left now is a pile of cash that they can spend eliminating competition. Meanwhile like others have rightly pointed out, talent at Google and even Mistral have been crushing it.
I feel like history is repeating itself yes, but actually I was thinking more a out Google.
Everyone was saying "oh man - Google had all this tech and they sat on it and just couldn't move forwards, then they blew their lead and OpenAI came a long and smoked them!"... Now it feels like it is OpenAI who are repeating that story, blowing their lead they got with the original ChatGPT while that upstart Google schools them in model development and vertical integration.
Interesting times. Very interesting times. C'mon OpenAI, move the SOTA forward!
Just to stem pointless debates before they flame up - both these acquisitions appear to be primarily if not exclusively for stock.
Sure, if you want to get into theoretical finance, OpenAI could have sold these new shares for cash, so technically there's no difference, but OpenAI is only spending opportunity cost cash, rather than fiat.
OpenAI's fiat likely still goes to the things you'd expect, like training models and paying for inference.
The AI hype seems driven more by stock valuations than genuine productivity gains.
Developers now spend excessive time crafting prompts and managing AI generated pull requests :-) tasks that a simple email to a junior coder could have handled efficiently. We need a study that shows the lost productivity.
When CEOs aggressively promote such tech solutions, it signals we're deep into bubble territory:
“If you go forward 24 months from now, or some amount of time — I can’t exactly predict where it is — it’s possible that most developers are not coding.”
- Matt Garman – CEO of Amazon Web Services (AWS) - June 2024
"There will be no programmers in five years"
- Stability AI CEO Emad Mostaque - 2023
“I’d say maybe 20%, 30% of the code that is inside of our repos today and some of our projects are probably all written by software.”
- Satya Nadella – CEO of Microsoft - April 2025
“Coding is dead.”
- Jensen Huang CEO, NVIDIA - Feb 2024
"This is the year (2025) that AI becomes better than humans at programming for ever..."
- OpenAI's CPO Kevin Weil - March 2025
“Probably in 2025, we at Meta are going to have an AI that can effectively function as a mid-level engineer that can write code."
- Mark Zuckerberg - Jan 2025
"90% of code will be written by AI in the next 3 months"
The loss of productivity is,as many things are, not directly measurable. Mediocre code making it into the codebase and hindering future development and increasing maintenance time, or even being the cause for some ideas to never be discovered, how are we going to measure that? How are we going to measure engineers no longer properly knowing the codebases?
Businesses will wake up when it is too late and the damage to the engineering side of their products is already done. Or perhaps won't wake up at all, and somehow (to their management levels) inexplicably fail.
> When CEOs aggressively promote such tech solutions, it signals we're deep into bubble territory:
Correct. This is how most bubbles are kept up as they are all exposed in the hype cycle.
You will not hear about the mistakes [0] [1] [2] it makes when AI gets it wrong or hallucinates and all the AI boosters can do will be "it only gets better" and promise that we will soon be operating airplanes without humans. [3]
Surely you would feel safer if you or your family boarded a plane that was fully operated by ChatGPT, because it is somewhat closer to "AGI"?
US shareholder capitalism is increasingly dependent on pushing a fantasy where X Company will eventually operate as an unregulated monopoly with armies of machines run by a small group of contract labour with no benefits and no wage bargaining power.
Sure, copywriting seems to be pretty much irrelevant now. Same with image generation wherever you can get away with it. The quality may be reduced in many cases but the cost is absolutely a fraction of what it used to be.
You fire the copywriter, but you still have to pay for the advertising (google clicks, tv spots etc).
So if you have $100,000 ad campaign and you use AI instead of paying $10,000 to the copywriter, you probably have a higher chance of wasting the $90,0000 ad spend.
So it comes down to, the AI probably makes a skilled copywriter better. But you can't get rid of him
The models will not be a moat, but the products can be. More specifically "sticky" products / killer apps like ChatGPT, and whatever forthcoming products this acquisition of Jony Ive's company may lead to.
Windsurf acquisition may be explained in part by the same logic of owning a strong and sticky product, as well as a good source of data for training.
It doesn't look like they're cash-strapped, more like they want to raise stakes.
To play in the same league as Google and Microsoft you have to be big. So they need to increase enterprise value to be taken seriously.
That's what investors expect them to do.
The only other option is to close it down, as OpenAI would quickly become obsolete if they can no longer produce frontier models.
As for the moat, it's not something you can just conjure, right? Perhaps the whole point of these acquisition is to create a moat, but only time will tell if that worked.
No piece of software inherently makes money on its own... You have to monetize Google search by running ads, Windows by adding subscriptions, etc. Thats not tied specifically to the software that actually drives the value for a consumer
