[HN Gopher] Apple Ramps Up R&D Intensity to Pre-iPhone Levels
___________________________________________________________________
Apple Ramps Up R&D Intensity to Pre-iPhone Levels
Author : retskrad
Score : 85 points
Date : 2024-01-19 19:02 UTC (3 hours ago)
(HTM) web link (www.statista.com)
(TXT) w3m dump (www.statista.com)
| seatac76 wrote:
| Wonder what they are cooking ?
|
| - AR Glasses would be one but they still feel like a decade away
| - Whatever happened to the Apple Car? - Apple Watch with more
| sophisticated medical sensors for sure must be in there.
|
| Wonder what else?
| iamwpj wrote:
| component chips maybe:
| https://www.reuters.com/technology/qualcomm-supply-apple-wit...
|
| Also, battery technology might be overcoming a hurdle.
| https://electrek.co/2023/12/20/welion-reportedly-expects-to-...
| isaacisaac wrote:
| I guarantee LLMs are a big piece of it. Not necessarily
| ChatGPT-like interfaces but instead dynamic interfaces/behavior
| that intuit what you want the device to do, especially on
| mobile where the user <> device interaction bandwidth is quite
| constrained.
| mustacheemperor wrote:
| This is pretty much confirmed: https://www.bloomberg.com/news
| /newsletters/2023-10-22/what-i...
|
| >the company built its own large language model called Ajax
| and rolled out an internal chatbot dubbed "Apple GPT" to test
| out the functionality. The critical next step is determining
| if the technology is up to snuff with the competition and how
| Apple will actually apply it to its products.
|
| >Giannandrea is overseeing development of the underlying
| technology for a new AI system, and his team is revamping
| Siri in a way that will deeply implement it. This smarter
| version of Siri could be ready as soon as next year, but
| there are still concerns about the technology and it may take
| longer for Apple's AI features to spread across its product
| line.
|
| I wonder, will they launch this as "Siri, but actually
| useful" or will the Siri brand be replaced altogether.
| jwells89 wrote:
| If this next-gen LLM-based Siri is built to take advantage
| of local documents/data with all processing happening on-
| device, I think it could be more useful than incumbent LLMs
| in some ways even if it's technically not as smart simply
| because it has far more relevant context by default. It
| being local-first could also help assuage privacy concerns
| and make users less reluctant to use it.
| micromacrofoot wrote:
| I'd be happy if they used an LLM interpret the command so
| Siri could better understand what to do without specific
| wording or requiring multiple separate commands.
| mustacheemperor wrote:
| There are unlabeled, very slick looking LIDAR enabled cars
| driving around the Mountain View area - I've been told by a few
| folks on HN and IRL that 'everyone knows' they're Apple
| vehicles.
| ayewo wrote:
| Any photos of these slick vehicles you could point to?
| mustacheemperor wrote:
| Unfortunately no, and I should mention a cursory search
| turns up known driverless startups active in Mountain View
| - so I think unless any other HNster has better evidence,
| treat me as hearsay.
|
| I also shouldn't oversell them, the ones I've seen look
| 'slick' in that they are similar to other well-polished
| LIDAR test vehicles that aren't just a bunch of sensors
| bolted to a roof rack, and they're 'mysterious' in that
| they're unmarked.
| brcmthrowaway wrote:
| Probably Ghost Automation
| AlotOfReading wrote:
| For context, this is what Apple's autonomous vehicles
| look like: https://dmxvlyap9srmn.cloudfront.net/article/p
| ictures/8e1f6d...
|
| They're Lexus' with mattresses on the roof.
|
| Ghost has a weird pattern and their name on the sides of
| the vehicle. Nuro is uses Prius' mostly. Woven
| planet/Toyota tends to have slicker looking vehicles.
| They're the right vehicle in this picture:
|
| https://thelastdriverlicenseholder.files.wordpress.com/20
| 23/...
| pests wrote:
| What is attached to the back wheel in the Apple photo?
| AlotOfReading wrote:
| Just a rotary encoder.
| TaylorAlexander wrote:
| I think AR/MR are good candidates for investment for Apple
| right now. They just launched sales of their Vision Pro today,
| shipping in two weeks. They are launching an entire new
| category of product which necessarily requires a broad spectrum
| of deep engineering investment. The current product is clearly
| too bulky and expensive for mainstream appeal, so they are
| going to need to release repeated generational improvements to
| both cost and size. The ten years from AR glasses is ten years
| of heavy product investment.
|
| And as others said, there is a lot happening in AI too which
| would go well with headsets and phones alike.
| sneak wrote:
| The Apple Silicon project has to have been fantastically
| expensive.
|
| They now use them in all of the watches, all of the phones, all
| of the macs, all of the ipads and ipad pros, all of the
| headphones, and the current and future AR systems (there is
| also a special Apple Silicon AR coprocessor in the Vision Pro,
| separate from the Mx CPU/GPU/SoC).
|
| They are designing and iterating like 4 or 5 different custom
| chips, with many of them incorporating industry-leading GPUs,
| and having them manufactured at scale (which I'm sure is not a
| simple handoff but involves additional engineering to support
| high yield manufacture).
|
| This project alone is larger than most technology companies
| that have ever existed, and a lot of the staff for it must be
| extraordinarily compensated to retain them (as evidenced by the
| fact that a bunch of them recently left to start a new chip
| company (Nuvia) which was then acquired by Qualcomm for $1.4B).
| readyplayernull wrote:
| They've been working on an EV since several years ago, wonder
| what happened to that project:
|
| https://i.redd.it/2npa5ztl1o751.jpg
| andy_ppp wrote:
| Probably secret spending on getting away from Chinese iPhone
| production...
| mensetmanusman wrote:
| GPUs!
| thiago_fm wrote:
| Money likely spent on crap.
|
| Apple is now a really bloated company that barely innovates, and
| now with plenty of governments breathing on its neck.
|
| Just look at how many IPhone models exists nowadays, Steve Jobs
| would be shocked to see Apple in 2024.
| mustacheemperor wrote:
| I don't think he would be shocked to see this comment, though,
| since it is materially identical to the complaints I remember
| reading circa 2010 and is as willfully ignorant of the Vision
| Pro as those complaints were of the iPad, and later the Apple
| Watch and the AirPods (when 'the FBI will destroy Apple's
| privacy edge' was the government conflict du jour).
|
| I don't mean to come across as standoffish here, it's just,
| this complaint reads to me like the "HN is becoming reddit"
| complaints. People have been saying it forever, so, without any
| new evidence, it just reads like the same old prediction that
| hasn't come true yet.
| mattmaroon wrote:
| And if they didn't invest in r&d the same people would say
| they're stodgy dinosaurs cashing in on what they've got ABC
| heading for extinction. Like HN said about Microsoft 15 years
| ago.
| 0x457 wrote:
| To be fair, AirPods that aren't Max and (maybe) Pro are
| complete garbage. Any wireless headphones in the same price
| range and often cheaper is better at every category except
| integration with Apple ecosystem.
| AnthonyMouse wrote:
| > as willfully ignorant of the Vision Pro as those complaints
| were of the iPad, and later the Apple Watch and the AirPods
|
| These are the evidence of the opposite.
|
| Apple's hardware from the 70s and 80s, like the original
| Macintosh, were something special. NeXTSTEP, the Steve Jobs
| thing that got rebranded as Mac OS X when Apple bought them,
| was the first modern OS to be simultaneously good and
| popular. The iPhone changed what a phone _is_.
|
| The iPad is an iPhone, but bigger. The watch is an iPhone,
| but smaller. These are not innovations of the same kind.
| They're the sort of thing you'd expect out of Facebook or
| Amazon or any other large mature bureaucratic corporation.
|
| If the Apple II was the 70s and the Macintosh was the 80s and
| NeXTSTEP was the 90s and the iPhone was the 2000s, name the
| thing of this caliber they did in the 2010s.
| jedberg wrote:
| > name the thing of this caliber they did in the 2010s.
|
| Well, you had to wait for the 2020s, but Vision Pro. It's
| an entirely new computer, meant to be your primary device.
| It's similar to how the Mac came out when PCs already
| existed, but were just different.
|
| Just wait 10 years until the market realizes it.
| AnthonyMouse wrote:
| >Just wait 10 years until the market realizes it.
|
| People said that about Google Glass 10 years ago.
| jedberg wrote:
| Google glass was ahead of its time. And was creepy with
| its recording ability and other features it launched
| with. Also, it was only a display for an attached phone
| for the most part. It couldn't really do anything.
|
| This is a whole computer strapped to your head.
| AnthonyMouse wrote:
| Why would you _want_ a computer strapped to your head?
| Computers generate heat and have weight that you 'd
| rather put in your pocket than have to support with your
| neck. To make it work you're stuck with a trade off
| between performance, weight and battery life when none of
| those is fun to sacrifice in this context.
| amanj41 wrote:
| The Vision Pro is hardly differentiated from the top end
| Meta VR model. There are a few UX enhancements, and I
| think they'll take the lead in the space eventually. But
| I wouldn't say this is a big innovation in the VR space
| from what I've seen.
| jedberg wrote:
| > The Vision Pro is hardly differentiated from the top
| end Meta VR model.
|
| I think their approach is unique. They are branding it as
| a replacement for your computer. The one and only device
| you need to be productive. And don't underestimate the
| value of not needing controllers.
| mjamesaustin wrote:
| The iPhone was hardly different from other touch screen
| phones at the time, except for the UX enhancements. And
| it turns out good UX changes the world.
|
| Pay attention to how people talk about the experience of
| using Vision Pro vs other headsets. The eye tracking
| interface is widely praised and described by third
| parties (not just Apple) as feeling like magic.
|
| I haven't used it yet, but I can imagine what using a
| computer would be like if I didn't have to actually point
| my mouse anywhere and instead it could effectively read
| my mind about where I want to click. IMO this is the
| interface revolution that will become ubiquitous over the
| next 5-10 years, and Apple is once again leading the
| charge.
| JumpCrisscross wrote:
| > _name the thing of this caliber they did in the 2010s_
|
| AirPods. It's a category-defining product, and in pure
| Apple tradition, a refinement of an existing one.
| AnthonyMouse wrote:
| AirPods are earbuds. They're nice earbuds, but it's
| hardly the next iPhone.
| hatsunearu wrote:
| We went from earbuds looking like one thing, to basically
| everyone wearing airpods-like things with great noise
| cancellation and usability. I'd say that's a home run.
| MBCook wrote:
| Why weren't there tons of earbuds that looked like
| AirPods before the AirPods came out? Two independent
| things that weren't linked with a wire that you could put
| in a charging case?
|
| I seem to remember there was one other similar looking
| device at the time, I think from Samsung. But it wasn't
| as small and got terrible battery life.
|
| There was absolutely nothing like them at the time in the
| category.
| AnthonyMouse wrote:
| > Why weren't there tons of earbuds that looked like
| AirPods before the AirPods came out? Two independent
| things that weren't linked with a wire that you could put
| in a charging case?
|
| Because making them that way is more expensive.
|
| > There was absolutely nothing like them at the time in
| the category.
|
| iPhone: There is now a usable internet-connected general-
| purpose computer in your pocket.
|
| AirPods: We got rid of the piece of wire between your
| earbuds.
|
| It's not the same thing.
| JumpCrisscross wrote:
| > _iPhone: There is now a usable internet-connected
| general-purpose computer in your pocket. AirPods: We got
| rid of the piece of wire between your earbuds._
|
| Most people who have AirPods didn't have earbuds with the
| wire. Those who did didn't wear them like AirPods. The
| iPhone is a computer in your pocket you interact with
| through touch. AirPods are computers in your ears you
| interact with through voice. _Especially_ with LLMs, the
| latter is a huge step forward.
|
| Also, the processing power in an AirPod outclasses Apollo
| 11's; if we're going to go full tech reductionist, it's
| still a feat.
| CharlesW wrote:
| They literally were the next iPhone. Apple mainstreamed
| the wireless earbuds product category in 2017 and then
| proceeded to dominate it. (Note that Apple owns Beats as
| well.)
|
| https://www.statista.com/chart/26791/most-popular-
| headphone-...
| thoughtsimple wrote:
| >If the Apple II was the 70s and the Macintosh was the 80s
| and NeXTSTEP was the 90s and the iPhone was the 2000s, name
| the thing of this caliber they did in the 2010s.
|
| Arguably, the iPod was 2000s and the iPhone was 2010s. The
| iPhone really didn't get started until the iPhone 3GS in
| middle-2009 so it started slowly but really defined the
| 2010s for Apple.
| AnthonyMouse wrote:
| That's just as true of the others. NeXTSTEP was the 90s
| but "Mac OS X" wasn't released until 2001. But if you
| asked someone in 2004 or 2014 if Apple had done anything
| interesting recently, you'd have been able to name what
| it was.
| chrisjc wrote:
| > If the Apple II was the 70s and the Macintosh was the 80s
| and NeXTSTEP was the 90s and the iPhone was the 2000s, name
| the thing of this caliber they did in the 2010s.
|
| Apple Silicon?
| AnthonyMouse wrote:
| Apple Silicon is slower on the same process node than
| AMD.
| staplers wrote:
| I don't think he'd be shocked. More like disappointed. Most
| innovators and future thinkers understand the entropy of large
| organizations left to business-types.
| mikestew wrote:
| You forgot to mention that the only reason people buy Apple
| products in the first place is because of marketing. If the
| sheeple ever wake up, Apple is doomed.
|
| Or so I've been hearing for, what, the last decade or more?
| gumby wrote:
| Decade? Famously: "No wireless. Less space than a Nomad.
| Lame."
| danaris wrote:
| People have been saying that about Apple for 40 years now.
| micromacrofoot wrote:
| Apple is one of the most successful business on the planet, and
| has been for years. Jobs would be shocked to see how big of a
| money pool he could swim in.
| master_crab wrote:
| I don't think it's anything as sexy as some world-shattering AGI,
| etc.
|
| Remember they are doing a lot of building and research on things
| other companies already do: cellular modems, microled, etc. This
| is probably where a large portion of these costs go: the
| refinement of production ready devices and the manufacturing
| processes/plant needed is incredibly expensive.
| nerdjon wrote:
| This makes me wonder, sure there is a large investment cost in
| creating many of those things. But is the Maintence/upgrade
| cost (particularly for things like the modems) the same or less
| than the investment to do it in the first place?
|
| If they are going to own as much of the internal hardware as
| they clearly want to, they have to be ready to continue to
| innovate or risk falling behind.
|
| I ask because I legitimately don't know the answer, but its
| something I have been curious about.
| parl_match wrote:
| With how much they pay for modems and licensing, I can
| imagine that they have done forecasts of how much they could
| save by DIY. Even having a slightly inferior modem product in
| one low price sku is a huge bargaining chip that could save
| them billions in the long term.
| CharlesW wrote:
| > _I don 't think it's anything as sexy as some world-
| shattering AGI, etc._
|
| AGI is a moonshot, and Apple doesn't talk about their in-
| development moonshots until very late in the product
| development cycle. There's evidence that ChatGPT lit a fire
| under their butt1 as it did the rest of the tech industry, so
| I'd be surprised if whatever they're working to launch this
| year didn't have tendrils reaching years and decades into the
| future.
|
| 1 https://www.macrumors.com/guide/apple-gpt/
| tambourine_man wrote:
| In absolute values, not percentage. IIRC, Apple was a [?]200b
| dollars company back then. It's a 2 trillion dollars company now.
| pohl wrote:
| TFA: "...R&D spending _as a percentage of sales,_ to levels
| last seen before the iPhone launch... "
| MBCook wrote:
| No, as a percentage.
|
| They've always been spending more than last year. It goes up
| every year.
|
| What's new, as the graph in the article shows, is that as a
| percentage they back to spending the level they used to before
| the iPhone.
| ilrwbwrkhv wrote:
| Nice. The only company out of the FAANG which is worth its salt
| and will survive into the future.
| chx wrote:
| If you go back to Steve Jobs' original iPhone speech, the crux of
| it was: revolutionary user interfaces (mouse, click wheel, multi-
| touch) enabled revolutionary products (Mac, iPod, iPhone).
|
| And while we know the LG Prada shipped with a capacitive
| touchscreen that's a technicality: LG took the UI of a feature
| phone, minimally adjusted it for touch and that was it. Meanwhile
| Apple took their desktop OS and shrunk it to the phone.
|
| This is to say, if Apple is working on something big, we can't
| predict it. To quote David Pogue's first iPhone review
|
| > The iPhone is revolutionary; it's flawed. It's substance; it's
| style. It does things no phone has ever done before; it lacks
| features found even on the most basic phones.
|
| Even if we had an inclination of what they are cooking, figuring
| it out is near impossible. And we don't.
|
| Of course, as a person who lives with T2 diabetes, I'd love an
| Apple Watch with a noninvasive glucose sensor. But while I have
| no doubt it costs dearly to develop and most importantly,
| validate that, I would guess it costs like 9 figures total (which
| is enough for most companies not-Apple not to be able to crack
| it). These numbers here are 11 figures. Yearly. Not even close.
| MBCook wrote:
| I'm really hoping they crack that nut. I'm not diabetic, but I
| do wonder what would happen when someone figures it out.
|
| All of a sudden you could basically put real time glucose
| monitors on a huge chunk of the population with all sorts of
| medical conditions or none at all.
|
| What could we accidentally learn from that that no one was ever
| looking for?
|
| The benefits to diabetics are obvious and life changing unless
| you already have a CGM. And if we never learn anything we
| didn't already know it would still be a fantastic improvement.
| But what if there is more?
| chx wrote:
| Trust me my friend, even if you have a CGM the benefits would
| be gigantic.
|
| I can't speak for T1 people who spend their entire lives like
| this and maybe it gets easier but inserting the CGM sensor,
| even if you use a local anesthetics is unpleasant and
| psychologically taxing. Then it interferes with your daily
| life and more importantly, it interferes with your mental
| health. Do you need a visual, can-be-felt reminder every
| second of your life you have a disability? I thought not.
|
| I have worn one and made my own meal plans while doing so and
| then stopped wearing them and I am fine, I can control my own
| blood sugar fine with just eating well, I am now down to six
| month blood sugar checks just to make sure I do not relapse.
| Still, having a non-invasive check so you can more easily
| test new stuff would be more than welcome.
| sho_hn wrote:
| > Even if we had an inclination of what they are cooking,
| figuring it out is near impossible. And we don't.
|
| This is some sort of strange myth-making. Making a touchscreen
| phone had been the writing on the wall and pretty obvious to
| many. Having a company set up and willing to do it and do it
| well is another matter. But it doesn't mean you can't see
| things coming.
|
| Ideas and execution are not the same.
| chx wrote:
| > Making a touchscreen phone had been the writing on the wall
| and pretty obvious to many.
|
| Really.
|
| This reminds of this discussion https://reddit.com/r/oculus/c
| omments/1wdp6n/what_makes_the_o... I opened a decade ago
| asking "What makes the Oculus different from previous VR?"
| and the winning answer was pointing out how much the 1994 IBM
| Simon Smartphone feature list resembled on the surface the
| iPhone feature list.
|
| In other words, still watching the world through my own eyes,
| I have been a columnist of the largest computer monthly in
| Hungary from 1993 to 2001. Yes, the IBM Simon, the Nokia
| Communicator, the Palm, the Pocket PC, yes those were there
|
| But none of that meant anyone could have foreseen the iPhone.
| Android history articles often mention how Google essentially
| needed to restart because of how revolutionary it was! https:
| //www.theatlantic.com/technology/archive/2013/12/the-d... if
| there was a prophet who could foretell the iPhone their words
| didn't register in Mountain View, that's for sure.
|
| If you think anyone did, the Internet does not forget, link
| it.
| sho_hn wrote:
| I'm not talking about columnists. Journalists are not
| people who build things. I'm talking about people who build
| things and think about what to build next.
|
| The option of doing away with the hardware keyboard and
| going touchscreen-only was always there and always up for
| discussion, but for the longest time not easy due to the
| power and performance envelope at the price point. Lots of
| devices already worked like this outside of phones, for
| example a lot of touchscreen kiosk hardware (e.g. ATMs).
| It's just not a particularly big leap of the imagination to
| do this in a phone form factor. It was also present in
| mobile form in plenty of media and scifi, e.g. _2001_ 's
| tablets.
|
| What matters is whether you think you can pull it off, and
| whether you're willing to try, and whether you are able to
| identify the right moment to try. A lot of companies were
| not willing to try (e.g. out of fear of cannibalizing
| existing product lines or rejection by the customer).
| Others didn't have the tech stack necessary to do so.
|
| What Apple did right is to have the good sense (or luck) to
| try it when the technology supply chain to back it was
| there, and not do it too early. Other phone makers wrote
| their tech stacks for much lower-power devices, and really
| struggled to scale their existing operating systems, dev
| community, etc. up to making a richer device, and didn't
| have the ambition to push for it. Apple started, as OP
| pointed out, by scaling down a desktop-class OS at a point
| where the hardware had just made it possible, and this
| worked out well. What made the iPhone a success wasn't just
| the form factor, it was also e.g. having a proper web
| browser (which Apple didn't even write themselves). This
| pattern replicates in a few other aspects of the device
| (supply chain for capacitive touch, batteries, etc.)
|
| The iPhone is a fantastic success story of a good product
| concept, waiting for the right moment, sensible tech
| choices, solid tech asset management by a software company,
| very good industrialization, and great execution and
| integration. It's not a magic fairy-tale unicorn idea that
| no on else could think of.
|
| It's annoying because it misrepresents and misunderstands
| what it takes to make a good product, and leads to
| internalizing the wrong lessons and values. It's one of the
| subtle reasons why there aren't more Apples.
| astrange wrote:
| The novelty in the iPhone wasn't a touchscreen, it was a
| capacitative multitouch screen and a UI framework that
| could do their app designs but still at 60fps.
|
| Although the last one wasn't purely technical; IIRC Nokia
| got it, decided it was cool "because it was 60fps", tried
| to compete just by doing that and failed.
|
| > having a proper web browser (which Apple didn't even
| write themselves)
|
| Yeah they did. There wasn't that much KHTML in there at
| that point.
| ramesh31 wrote:
| >If you go back to Steve Jobs' original iPhone speech, the crux
| of it was: revolutionary user interfaces (mouse, click wheel,
| multi-touch) enabled revolutionary products (Mac, iPod,
| iPhone).
|
| This was specifically called out in the Vision Pro
| announcement. They are completely ditching controllers for
| VR/AR, and going full hand/finger tracked gesture control. It's
| why I think Vision will be the first device that people
| actually end up using in mass. The display fidelity is one
| thing, but interaction control is the number one reason why VR
| has failed. People (average normal people) really have no idea
| how to use game controllers if they didn't grow up as gamers.
| The platform will never take off until fully intuitive
| controller free interaction is acheived, and Apple knows this.
| Ambient computing is the next step, and keyboard/mouse will be
| a small footnote from the early history of HCI years from now.
| Legion wrote:
| Hope that includes the non-invasive glucose monitor for the Apple
| Watch.
| Someone1234 wrote:
| I think everyone is hoping that. Non-invasive glucose
| monitoring is a major holy-grail piece of medical technology
| that would help people almost immediately.
|
| If they did develop this, they'd likely make as much money
| selling the IP to medical technology partners (e.g. hospital
| equipment) as they would adding it to the Apple Watch.
| paxys wrote:
| This is yet another instance of people simply not understanding
| what R&D spending means in the context of tech. Most employee
| salaries are R&D. Cloud compute is R&D. Any expense related to
| developing any of their existing products or services gets
| categorized under that bucket, and they get tax breaks on it.
|
| R&D does _not_ only mean new research.
| sho_hn wrote:
| This is correct and should top the thread.
| edgyquant wrote:
| > Most employee salaries are R&D. > and they get tax breaks on
| it.
|
| This isn't true anymore. The tax break for engineers counting
| as R&D ended last year and this is cited as one reason for so
| many layoffs
| keenmaster wrote:
| AI is the greatest threat to Apple's dominance. They might be
| spending billions on creating an AI-centric iOS and iPhone, and
| that would be money well spent. If Android phones put a personal
| secretary in your pocket, fully integrated throughout the device,
| then iPhone users would be very tempted to switch without feature
| parity.
___________________________________________________________________
(page generated 2024-01-19 23:01 UTC)