Tag Archives: Siri

Tick-tock – Apple goes Intel & other thoughts on the iPhone 4S launch

Just a few thoughts about the launch of the iPhone 4S, in no particular order:

Tick-Tock

Five iterations of a product are not a lot to base this on, but it seems that Apple has adopted for its design something like the ‘tick-tock’ approach that Intel uses for its processor designs: alternating between a complete change of the product and an efficiency/smaller upgrade.

This is disappointing for all the industry pundits who were getting bored of the old iPhone, and for those Apple fanboys eager for an immediately distinctive upgrade to prove their cool, but to me it shows how Apple want to position the iPhone: as a product that is complete enough, and has enough value to the customer, that they don’t need to devalue the older models each year with a complete change of the lineup. For them there is no reason to enter the race of fast upgrade cycles and the technical spec wars, as they position the iPhone as something outside of the feature-sheet comparison buying decisions.

A good example for that type of comparison can be found at engadget . In the end, Apple count on the majority of users not caring what the precise battery capacity is, but whether a charge gets them through a day, and not whether the front camera has more than VGA resolution, but that they can actually make a video call using it. How utterly out of hand the spec comparison can get, with the necessity to have a category ‘winner’ highlighted in green there’s a chart at gsmarena, where the old iPhone wins over the now one because it is 3 grams lighter. 3 grams are an academic difference, not noticeable to anybody who doesn’t own a precision scale and bothers to actually compare them.

Hardly surprising

On the hardware side the biggest surprise was that there were no real surprises. Faster processor, better camera, more memory – all logical upgrades. The improved antenna was really a necessity. ‘Antennagate’ didn’t really influence the sales of the iPhone 4, but to let the problem remain in the next iteration would have been shameful. The ‘world phone’ aspect is interesting in that it has CDMA and GSM, but of more relevance to Apple in that they can now have a single production model than to the consumer. It won’t make switching carriers in the US, the market where this technical change would matter most, any easier.

Still a phone camera

“The best stills camera on a phone.” is something that has yet to be proven in independent comparisons (yes, that’s the N8 owner in me). My guess is that the increase to 8 Megapixels is Apple’s concession to the megapixel race, and that they would have preferred to just increase the sensor sensitivity. In actual use better low-light performance is worth more to users than higher pixel count. Still, I have no doubt that the new camera is an improvement, and that this is another nail in the coffin for standalone consumer digital cameras below the high end.

Fairy dust

Apple continues to integrate concepts and services that they see as proven on other platforms, and apply the magic of Apple branding speak to them. This time round it’s Google Latitude (and all the other lesser known ones that came before), which has the fairy dust of ‘Friend Finder’ applied to it.

Postcards of children

The marketing fairy dust seems to have failed for the snail-mail greeting card service. In addition to apparently lacking any snappy branding, is did seem confusing for young phone bloggers why this is something Apple would add. Really it’s just a sign of what the iPhone has become: an everybody phone, not something targeted at geeks and early adopters. Among the everybodies, the iPhone users with children who can now send a great photo to the grandparents will be glad. A relatively limited use case? Sure, but it is a service that makes perfect sense for Apple. It’s a simple concept that is self-contained,  feature-complete, easy to use, builds on existing, proven infrastructure (online photo services), and is something that will prove indispensable to a subset of iPhone owners who use it. Plus there’s a very respectable profit margin there.

Computer – execute

My first thought about Siri was just: another attempt at voice search. Then I wondered about the branding. It’s not an ‘i’ service, it isn’t descriptive, using one or more common English words (see the above ‘Friend Finder’). ‘Siri’ is an anthropomorphization of the service. It really is intended to be an assistant, not just a piece of software, and, most importantly, not just voice search and simple voice commands for the phone. In view of this the naming does make sense. You traditionally talk to people, so talking to your phone should involve at least some feeling of this being similar. And if you expect some kind of intelligence in the service’s reactions to your requests, then casting it as something quasi-human is probably going to happen automatically anyway. Just witness the people who ascribe a personality to their voice-guiding satnav units.

Regarding the functionality I’m sure that Apple integrating this means that they’re confident it works well enough for a large number of use cases. In principle there is a lot of value in this. Setting an alarm via voice has a real speed advantage, and things like conditional notifications (the ‘remind me after work’ example) are greater help. We’ll see what else Apple offer in terms of functionality at the moment, and how discoverable what the service can and can’t do is going to be.

The general problems of voice operation remain:

  • The adaptation to additional languages is a lengthy and expensive process. This goes first of all for the speech recognition side of things, but also for the background semantic parsing of the input. A lot of this is sure to rely on sheer volume of previous data to compare to, of human-curated or human-accepted responses, and there’s no easy transitioning this to another language – and cultural context. This is entirely a problem of resources and roll-out speed, but some languages for which localization of the interface has been commercially viable will remain uncatered for.
  • The problem of language variation extends within what is formally seen as a single language.  Germany, for example,  is in fact not really a single market you can address entirely with one adaptation. I studied in Regensburg and students from other parts of Germany (native Germans most of them) often didn’t understand the locals. So people there might equally find that their iPhone’s don’t understand them.  Granted, Germany has a particular high variance in its dialects, but I’d expect there to be equal problems in parts of Great Britain, and in other places around the world.
  • Voice control needs situations where clear recording of audio is possible. Construction sites, factory floors, loud bars, the tube or windy streets are all highly problematic environments. This alone means that voice control can never be the sole input.
  • Voice control needs situations where it is possible to speak to the phone. There are two aspects to this: privacy and social acceptability.
    • I may not want others around me to know the content of my interaction with my phone, or the fact that I’m interacting with it at all (e.g. during a meeting).
    • In a lot of situations it is not socially acceptable to be talking much, it at all: open plan offices (without much telephone activity), at the library, or even in most public situations (think Japan). Further at the moment it might be acceptable to talk to another person, but talking to a machine would be seen as weird.

Taken together this means that the utility of something like Siri will vary hugely depending on the user’s everyday contexts. Somebody who drives to work and back alone in his car and has an office to himself is at the high end of the scale, while somebody using loud public transport and working in a shared, possibly noisy space, will be at the low end. As @disruptivedean tweeted: ‘Siri: designed for countries where people drive a lot & need to talk to devices while at the wheel. Not places with crowded public transport.’

So it remains to be seen how well this works in practice, and for how many people. Personally, I’d really like something like this – but with the addition of a text input box that I can type in in all situations where I can’t talk to the phone. Oh, and a desktop version, too. Neither is problematic in principle, and I guess both will come eventually.

Nothing new on the player front

The iPod touch hasn’t been updated. Together with the introduction a the 64GB iPhone model, this means that the biggest internal iOS storage is no longer exclusive to the iPod touch. An update to 128GB would have been easy technically, and wouldn’t have cut too much into the financial margins with the usual $ 100 price increase for the doubling of memory. Maybe Apple no longer sees local storage as that important?

Ranging from the expensive…

Apple now has a range of iPhones, from the 3GS via the 4/8GB to the 4S with different storage capacities. Does this mean they have the all-conquering Nano, the half-price iPhone? Seen from the US perspective that Apple announcements invariably adopt, they had that before, with the 4/3GS paring: $ 199 and $ 99 on contract. Now they have something unbeatable: an iPhone for $ 0! The iPhone 3GS directly competes with low-mid-tier smartphones like e.g. the Samsung Galaxy Ace or the HTC Wildfire, and it’s going to kill most, if not all of the competition there.

In the rest of the world, where operator subsidies are lower, or not the norm, things are different. Looking at the German pricing (which, as usual, translates the US dollar prices to euros 1:1, meaning a hefty markup even once the sales tax in the US is considered), the 3GS is listed at 369 €, the 4 at 519 € and the 4S as starting at 629 €. (Curiously this puts the 4 with 8BG above the current retail price for the 4 16GB – so go figure. I’ve never understood Apple’s German pricing anyway.)

So if you look at it within Apple’s price range, then the 3GS could be seen as the Nano. It’s more than half price, but it is significantly cheaper than the current model. If the prospective buyer is dead-set on getting an iPhone, then this significantly lowers the entry barrier. The problem is that the widening of the smartphone market is driven in large parts by people who don’t have the money to buy even the entry level iPhone. Plus it doesn’t do anything to convince people not yet committed to iOS, who shop based on such prosaic aspects as screen size and resolution, processor speed, camera quality, the ability to play back any video file they throw at it, or need more integrated memory.

Advertisements

Leave a comment

Filed under mobile