The Lackluster WWDC 2017 Keynote

June 5, 2017

If you haven’t yet watched the WWDC 2017 Keynote, I can save you nearly two and a half hours of wasted time:


iOS 11 has a lot great features

watchOS 4 has a few good features

macOS High Sierra has a couple good features

iPad Pro has some good features

iMac Pro is ridiculously overpriced and not expandable (doesn’t ship until December)

MacBook Pro still has too little power, too little battery life, a substandard keyboard, an unnecessarily large trackpad, a maximum of 16 GB of memory, and is still overpriced

Mac Mini and Mac Pro weren’t even mentioned, and both haven’t been upgraded in years

HomePod: an overpriced speaker to compete in the shrinking Sonos market (doesn’t ship until December)


No tvOS update

No new Apple Watch

No new Mac Mini

No new Mac Pro

No new AirPort

No new


Apple Watch, diabetes, and security

May 20, 2017

There have a been some rumors lately of Apple testing a monitor for blood sugar. Recent comments by Tim Cook suggest that such a monitor might be non-invasive:

“It’s mentally anguishing to stick yourself many times a day to check your blood sugar,” he said. “There is lots of hope out there that if someone has constant knowledge of what they’re eating, they can instantly know what causes the response… and that they can adjust well before they become diabetic.”

If Apple has developed this technique for accurately measuring blood sugar, they could save lots of lives, and Apple Watch sales could really take off to the point that it becomes a mainstream product like the iPhone.

Now, I’ve long assumed that Apple takes adequate steps to insure that private data remains secure, but I’ve also recently discovered from Apple Support that iCloud backups only store health data for sixty days, so if a customer wants to keep their health data indefinitely, they need to make their own encrypted backups in iTunes. Add that to the fact that Apple encrypts at least some iCloud data with both Apple’s and the user’s keys, and a customer might have cause to worry.

Moreover, Apple should already be anonymizing stored private data (say, a user’s browsing history) so it is separate from trivial data, such as purchase history. But health data should be anonymized yet again. After all, what healthcare insurer wouldn’t want to obtain exercise, blood pressure, weight, and now blood sugar data about Apple’s customers? What might Apple’s customers pay an unscrupulous individual to keep that information confidential?

If the Apple does debut this new feature on June 5th, the company should take steps to ensure that health data never leaves the customer’s device, and if it does, that data should be stored separately in iCloud, using only the customer’s encryption key, and not Apple’s.

The iPad as teacher

March 4, 2010

By now, most everyone has an opinion on the iPad: it’s a new computing paradigm, it’s a waste of money, etc. A few days after the announcement, I finally realized that this isn’t a device for the computer literate, it’s a device for those people who aren’t comfortable with computers or don’t want to learn how to use and maintain them. I’ve often written that the real computer of the future should function like a screwdriver: easy to understand, easy to use, no troubleshooting, and you only have to learn how to use it once.

The iPad might be the embodiment of that philosophy. Not only is pointing intuitive, but the iPad could actually help people learn. I assume it will start simply, and then get more complex as time goes on, perhaps going as far as agency.

Imagine opening an Algebra textbook. After the interactive theory, a student begins the exercises, but becomes stuck at simplifying an equation. The iPad software could wait for a minute or two, and then suggest a solution, going slowly, and going into more detail as the iPad and the student together simplify the equation step by step. If this happens with enough students, the iPad notifies the publisher that the thoery may be too difficult to understand based on the number of students having difficulty, and the textbook could be revised, with the updates pushed to the iPad.

Once developers learn how to help people understand and learn, it might be possible to break much more complex subjects into discrete elements. Imagine teaching Java programming interactively. We’ve all seen exams which test adaptively, but the iPad could instruct adaptively, quickly surmising a given student’s skill level and adapting the coursework accordingly. No more classrooms, no more teaching to the average student. Students could progress at their own rate, with the iPad gauging what they’ve mastered and what to teach next.

At that point, I think we’ve reached agency. If a computer has mastered a subject well enough to teach it adaptively, the computer could perform all sorts of functions, and be better at them, than humans. For example, let’s say you want a custom-designed customer management system. You don’t hire developers, rather you and the iPad spend hours together building the system yourselves. The iPad starts broadly, asking the user to select from several elements the system might possess. From there, the iPad and the user design the interface, with the iPad making suggestions, and the user accepting, rejecting, or modifying them as necessary. In this way, the iPad assists the novice user in the design of the system they need, but the iPad handles all the background programming and development. Now we’ve reached the point where the computer is the agent, not just a tool.

This might seem like science fiction, but I don’t think we’re that far away.

Why no live traffic on the iPhone?

January 6, 2010

With millions of iPhones in use today, Apple has an opportunity to create a real-time traffic condition reporting system. Current traffic information systems provided by XM and other are notoriously inaccurate, despite charging customers each month for the service.

Between the built-in GPS receiver, triangulation from cell towers, and the built-in accelerometer, Apple could have the iPhone upload its location, speed, and direction to a central server at predefined intervals. This information would be completely anonymized, retaining the phone’s IMEI just long enough to verify it as an actual iPhone, but recording no other personally identifiable information.

From these millions of iPhones, Apple could record traffic patterns 24 hours per day, correlating the traffic history of a given freeway with time of day, the day of the week, and even holidays and weather. After a few years, Apple would be able to predict the percentage chance of encountering a delay along a given route. The driver would enter the starting location and destination into the iPhone, and the route could be presented with both real-time traffic and predicted delays, and could be recalculated as circumstances changed.

What would such a service be worth to Apple’s customers? I’ll bet most would be willing to spend at least $5 per month for such an accurate system.

Why no wireless synchronization in iTunes?

January 4, 2010

iTunes only allows synchronization of music, podcasts, and audiobooks via USB. Conventional wisdom suggests this is due to the sheer volume of data being transferred. This is unfortunate, because it means customers are less likely to synchronize, and so the content on these portable devices doesn’t stay up-to-date without more effort from the customer. There are a few ways Apple could improve this system.

Both the iPhone and iPod Touch have wireless 802.11g radios. Apple could add a scheduling feature to iTunes which would download all recent podcasts at a preset time, and then synchronize with an iPhone or iPod Touch at a preset time. The downloads over wireless are likely to be 10% to 20% as fast as a USB connection, but as long as the initial synchronization of the device is completed over USB, updates could probably be done wirelessly without much trouble (most of the information is going to remain from synchronization to synchronization). Customers could plug their iPhones in the charger at night, and as long as their host computer is on, the iPhone would wake up at the predetermined time and begin the synchronization.

This works well for the iPhone, which stays with the customer and is likely charged nightly. But what about the iPod Touch? It could synchronize just as the iPhone does, unless it’s left in a car. If Apple produced an accessory combining a vehicle adapter with a longer-lasting battery, the iPod Touch could also perform its synchronizations wirelessly, presuming the car is within range of the owner’s wireless network. The accessory would charge the battery by day so it could power the iPod Touch at night, and the owner could install an 802.11n wireless access point (e.g AirPort Extreme) if they required more range to reach the iPod Touch in the car. Such an accessory could even include a small Wi-Fi radio (much like the EyeFi SD card), and could then be connected to any current iPod Apple sells.

Using these approaches, Apple could automatically keep the content on customers’ devices current without any effort on the part of the owner beyond the initial configuration.


July 3, 2007

Ah, the hype. Truly, it’s a great device. Too bad it’s marred by such a terrible experience with AT&T. One of our phones took 18 hours to activate, and the other took 56 hours. Yes, we’re new AT&T customers, and so had to be “pre-qualified” in person, but no one bothered to tell us that when we bought them at the Apple Store.

If I knew then what I know now, I probably would have stayed home.