1.5B requests would probably be half that if Siri got it right the first time.
the iPad is Maserati-class hardware paired with a 1968 Volkswagen
Now, now, the 1968 Volkswagen is cool.
I think it was a pretty solid WWDC. Some nice features added, and the AI direction looks to be good.
I wish they had added more Shortcuts and Home stuff, though. Maybe the other presentations during the week will highlight changes.
I agree with you, particularly about Shotcuts. On the positive, though, it seems like a lot of the Siri improvements will also flow to Shortcuts. That’s not the same as new system affordances in Shortcuts (which all of us who use Shortcuts would love), but I was encouraged by the multitasking of app intents and the obvious benefits that will accrue to Shortcuts. In fact, from the Platform State of the Union, it looks like these new app intents will work through Shortcuts before Siri is fully up and running.
I think most people who really feel a need for a calculator app on iPad already bought one from a 3rd party developer years ago. Those developers are delivering, and will continue to deliver, features that Apple isn’t/won’t with the stock app. For example, I’m not expecting Apple to support RPN anytime soon; this is one of the main features that drew me to PCalc. Unless maybe you can trick AI into doing it?
RPN is the same reason I was (and am) happy to have PCalc!
I love it when other people run beta software. Almost as much as when I’m speeding on the interstate and someone in front of me is going just a little faster.
Substantially better than the old Siri dictation?
Elon isn’t happy.
If Apple integrates OpenAI at the OS level, then Apple devices will be banned at my companies. That is an unacceptable security violation.
But probably not powerful and feature-rich enough to be able to bulk export your data in future-proof interoperable formats that can be used in other apps, sadly.
That is mainly true, though one can bulk everything as plain text with Exporter or import to DEVONthink. If one uses markdown syntax in the notes, the files can be used in a markdown editor once exported. Far from ideal, but it does work.
It “feels” like it’s better; however, It’s hard to tell yet. I’ll report back on this. The system made transcripts available for all my old recordings and voice memos. some were very accurate, some not so good. The new recordings that I tested, seem to transcribe pretty accurately. Also, note, I have not tried anything involving specialized jargon yet. Like I said, I’ll play around with it and report back here.
I find I get a different reaction when I take notes on a phone in a stand using a bluetooth keyboard. Touch typing on hardware keyboard says you’re working/taking notes. Typing with your thumbs or swipe typing on a soft keyboard says you’re texting with your BFF.
HKSV has been a failure, or at least a very narrow success. IMHO, the severe limitations of HSVK implementation means that only a very small percentage of Apple users are content with it.
For many people, the primary need is outdoor facing cameras and they are willing to trade the theoretical lesser security for the practical huge advantages of just about every other security camera offering.
Apple needs to really up their game here but they haven’t show any willingness to do so.
This trend, by Apple and third party apps for “text summarization” has me wondering if they are creating a monster.
Will we all be soon constantly asking “Did you read my ENTIRE email”, “Do you listen to my ENTIRE voicemail”, “Did you see my ENTIRE text message”?
I fear that nuance and details will be lost for far more conversations and interactions as everyone chases the false efficiency of summarized comms instead of just actually paying attention to what someone is communicating?
As a (former) marketer, I do love how Apple is attempting to take the high-ground by obliterating all their AI competitors through the deal with OpenAI by positioning ChatGPT as “just a model plug-in” which they might do with Google, Microsoft, Meta, and anyone else’s cloud-based huge models too.
Also, I noted a freemium approach - ChatGPT will be free, but you will still “connect your paid account” for access to advanced features.
My thought when they said this was, “Google, Gemini could be getting mentioned right here if you’d been willing to make a deal.”
That is a tremendously important point. We already have a bad habit of skimming because of the avalanche of information we face. Relying on summaries of information we did not take the time to read thoroughly could create significant unintended consequences and inadvertently communicate disrespect to the sender. Relying on summations is like quoting Plato, having never read Plato. The mirror side of this is having AI write our text. I have no issue with using it to correct grammar and typos, but doing your own writing is a matter of integrity.
The window tiling features look like they arrange windows exactly the way I like (with keyboard shortcuts!). The big question for me is how it works with 3+ monitors. Of the existing window management apps, half of them only work well with a single monitor, the other half are reasonable with two monitors. The only thing I’ve found that works well with three is rolling my own with Keyboard Maestro.
The new widescreen Mac Virtual Display is a big improvement. I’m also interested in the new photos features in VisionOS. Making hardware and software to capture immersive video to more creators augurs well for additional non-Apple content. As an F1 fan I’m interested to see what will come out of the Red Bull partnership.
Lock Screen customization will be a nice addition. I rarely use the flashlight and I’ve got the Action Button set to open the camera, so neither default option is all that useful to me.
If Smart Script can actually clean up my horrific handwriting, that alone will make this a major iPadOS year for me.
Finally, during the AirPods Pro noise cancellation demo, that was a really great looking sandwich that the guy she called was making.