The build-up to WWDC – Apple's annual conference for developers – has been an interesting one. The company normally uses the event to release major new software features and, occasionally, hardware, so tech fans and users of a very popular camera known as the iPhone watch it intensely.
Since it's Apple's practice to deliver many of the features within days and for free to existing users, it's usually well worth paying attention to for all Apple users.
This year it's also of particular interest because Apple, despite holding a leading position, is under a lot of fire for being behind the game in the AI space. Photographers are bemoaning the lack of AI tools to help with their on-the-go snaps in comparison to Google Pixel, while Siri is definitely not as smart as recent demos from Open AI.
I'd go so far as to say that I'd be very surprised not to see some kind of AI object removal in Photos.
There have been a lot of rumors about a 'deal' to bring ChatGPT to help Siri out in some way, and this will almost certainly be the moment we get to find out how that will impact just how 'smart' that'll make what for many of us is our most commonly used camera. We might also get to find out if the unnerving (and very unlikely) idea of a fee-paying Siri+ are going to become a reality!
Whether we see that or not, a LLM (Large Language Model) seems very likely, with the result being that Siri will chat more like a human, and be able to remember things from one question to the next.
We're also expecting to see changes to the control centre in iOS and Mac OS, perhaps a new Vision OS for the Vision Pro (we see that 'missing feautres' will be added according to reliable Apple rumor-monger Mark Gruman).
As well as watching the Apple feed here, it'll be possible to watch on an Apple TV – if you've got one. The Keynote, by none other than Apple CEO Tim Cook, will begin at 10AM in Silicon Valley, which is 1PM Eastern or 6PM in the UK.
I see on Twitter that Tim Cook is promising that it's going to be the 'Best Ever' WWDC, while journalists are taking their seats (or sharing selfies with some well-known tech reviewers – perhaps there will be a solid product to talk about!)
In any case I definitely need this event to start because the Apple animation is starting to wear me down ("Is this Eurovision?" a fellow – admittedly somewhat more disinterested – Apple user observed of the graphics and music choice). It is definitely not that impressive. Still, seven mins to go.
Ooooo...... here we go! Kick off sees... er... well the same Apple graphic crystalize. And then we're in a plane preparing for an air drop apparently. Straight after the D-Day landing commemorations they're parachuting onto the donut. Well, I mean probably not really, but it's an impressive bit of CG. And now Tim Cook is on the roof of Apple Park (so we can see the solar panels).
He's reminding us why developers are good (which they are, but that isn't an update...)
AppleTV+ is getting the initial love from TV. Looks like a new season of, well, lots of good TV is coming. And Tim's said we're going to get to "Intelligence" so the enormous elephant in the room has been acknowledged.
OK, now onto VisionOS... Hmm... "It's been just 4 months" and yet somehow there seems to be a slightly 'justifying' tone to my ear.
After that, VisionOS 2 presentation kicks off with photos – machine learning can make spatial photos from 2D originals. (Or 'bring them into the future')
There is an enterprise leaning, and Canon's EOS-R7 spatial lens gets a mention, so a workflow between Canon, Final Cut Pro, and Vimeo as the player app is being trailed. I tried the Canon tech at The Photography and Video show and it was impressive.
Apple Immersive Video content partners – including Redbull – will be bringing content to the TV app. A scripted short film was mentioned. Real challenges for creatives.
Oh, and it's coming to the UK on July 12th.
Now the iPhone; customization will now allow us to frame our home wallpaper picture by placing the app icons where we like them, and choose a 'dark look'. A complementary color can even be selected by iOS.
Ah, and as expected, Control Center is up...
It looks like the Action Button on the iPhone 15 Pro will be accessible to external developers settings features now, if developers create the feature using the new API and the user adds them.
Messakers – Ronak Shah tells us ways we're going to be able to take even longer composing texts with text effects, extra emoji on the tap-back 'quick' responses, and more that add more personality. All sounds great, but I know from experience that all styling options take more time than just typing!
The photos app – "Biggest redesign ever" – and it does look better. Now a single view with photo grid at the top and libraries below.
There's a finction to filter out screenshots to keep clutter free – that'll be bery handy for journalists on photography and tech sites, I can tell you!
The new style has a small white bezel around the albums, very Polaroid-chic!
Thre seem to be more options to put pinned collections near the top (and help dissuade the system from choosing the wrong pics). It does look better... but where's the AI... oh, yeah... Tim said later...
There was more on TVOS but it felt like they were holding back (or didn't have much to say). Finally catch up with other apps it'll tell me about the actors on Apple shows (but if there was real AI it could tell me who the actors were on any show – not just Apple ones, right?).
OK, now WatchOS 11. Exercise stuff here isn't bringing a lot for photography fans I'm afraid and – if the watch is anything like my Series 8 – Apple's software team do seem to forget that you have to take the hardware off and charge it from time to time. Oh, it seems it'll be able to tell if it thinks you've drunk too much?
We do learn that the 'Photos Face' is popular – and in WatchOS 11 machine learning AI will pick aesthetics and smiling faces to help choose the best photo to serve as a watch face.
OK, back to Craig and the iPad OS (and he's able to remind us about the still very new iPads...). Guess what, it's all very similar but bigger when it comes to the Photos app, but there will be a new floating tab bar / side bar in some apps (potentially familiar to Apple TV users).
And – one for the fans – Calculator is coming to iPad (believe it or not, the app didn't come to the iPad until now). But it's going to have a bonus feature – Math Note. You can draw with the Apple Pencil and as you write the equal sign and some AI reads your writing and completes your equation for you.
This actually does look like 'proper AI' (at least in Apple's carefully planned slick presentation!)
Time for the Mac. Craig does a 'superhero run' to another part of the donut.
Sequoia seems to feature a lot of the tools we've already seen, including Math Notes in Notes.
iPhone mirroring might be a star new feature though – you can access your phone in a window on your Mac desktop; use your pointer as a finger. iPhone notifications will be mixed into the MacOS stream.
The iPhone looks locked, or shows standby of that's the mode it's in – so people in whatever room of the house you left your phone in can't see what you're doing, and (at least in Apple apps) you can drag and drop between apps.
Nice! (Or the beginning of the end of desktop apps?)
Footnote there – Safari offers 4 hours more battery than Chrome when streaming video, Didn't say which Apple device, but that's something to remember (or be mindful of if you feel you have to use Chrome).
Ooo... and more AI – article summaries. Didn't say how it was being done though.
Tim's back – we're about to find out the big AI answers...
Starting by laying out the goals – powerful, intuitive, integrated, personal, private.
Here goes...
"Apple Intelligence" – well, it's not as nice a name as Siri. Oh, and back to Craig.
He dismisses chat tools that know things through what they've learned from world models, saying that Apple's approach is to work with personal, private knowledge of the user. Personal context. Potentially very interesting angle, but is it sales-speak?
Writing tools, proofing tools, summarizing, OK. Well, we saw a bit of that.
"Now you can create totally original images" he says, and demonstrates a cartoon image of "someone in your photo library". Sketch, illustration, and animation.
OK, so it can make generic emoji-like images because it knows the names of people in your Photos.
Now we're moving onto 'on device processing' and privacy. It's going to come back to the processing power of the M series chips, and on-device semantic indexes.
Many but not all are on-device, he says; now promoting security of Apple's cloud – 'Private Cloud Compute'.
Now we get to find out what that means. Siri gets 1.5 billion requests a day, apparently (which probably isn't that many if you think about it).
You can now type to Siri, and Siri can now remember context, though she still sounds a lot more robotic than Scarlett Joh—, er, I mean Chat GPT.
Siri can also work with Photo requests i.e. – "Show me photos of Dave in an orange coat" and then "Make this photo pop".
My inner cynic is deeply worried by Apple's integration of Apple Intelligence into Apple Mail (the example being demonstrated) – not because it looks bad, but because of the potential unforeseen results. These days we all have a lot of email because, well, it's easier than print mail was. Now we barely have to read or write it, what's going to happen?!
Apple are also creating a dedicated 'Image Playground' app (and API) for the 'genmoji' for the quirky illustrations the system can create – no generated photos though. Not this time anyway, Apple are not risking that copyright minefield!
Or did I speak too soon?
There is a new 'cleanup tool' in Photos though – as expected – similar to Google's – demoed removing an unwanted stranger in the background of a beach photo.
Search in videos was also mentioned – that has the potential to be interesting as it can be much more time-consuming to hunt through video clips. Apple Intelligence can also create movies with a narrative arc, apparently (though how different that is to the current tool I'm not sure).
Apple Intelligence will be free... BUT...
...ChatGPT 4.0 can be 'borrowed' by Siri – who asks permission to use the system first. This does unlock ChatGPT's image generation and is also going to be free to some extent but paid features will also be available for subscribers – so Apple are going to unlock the door for OpenAI to charge users for their features via Siri after all.
And that's your lot.
It was a big one though, and as Tim Cook wraps up he's not kidding that it's going to be a big week for developers who need to get to grips with all the new features and tools to integrate with AI, for features like the ones demonstrated – asking a photo app to apply the cinematic filter to the photo of Tim I took yesterday.
Given ChatGPT Plus comes in at $20/month and Cook was vague about how long 'other languages' outside American English would take be supported, there is a lot of potential for irritation as well as excitement from this event. It'll be interesting to follow the reactions and I'll keep posting!
We've started full write-ups of some of the stories of WWDC, including news about Apple's plans Blackmagic and Vision Pro content creation (it might still make sense to read the earlier entries in this first) and the new AI tool to create 3D images from 2D.
Interestingly Apple's share price seemed to drop about 2% during the initial WWDC announcement, with a lot of people seeing the potential Trojan horse for Microsoft / OpenAI which Apple have allowed, but it now seems to have rallied and then some as the reaction sinks in.
Also as part of the post-match we've just published a story on how you can access the new iOS features – including Magic Eraser copy – early (and why you shouldn't).
We're seeing a lot of tweets about Apple's decisions and announcements at WWDC, not least from Elon Musk, who seems to be deliberately attempting to start some kind of CEO war. Musk seems to have, at best, misunderstood the way Apple is restricting ChatGPT's access to user's information.
Don’t want it. Either stop this creepy spyware or all Apple devices will be banned from the premises of my companies.June 10, 2024
