SIRI and iCloud: Beta or Alpha?

SIRI

If all you compare SIRI to is the previous iteration of voice control on the iPhone it is leaps and bounds more useful in its overall capability. And yet fails epically if you have a weak or non-existed web connection. The casual english phrasing does seem to be a step out beyond the competition but once you get down to basic word recognition it seems to be on par with the likes of Google Voice or Vlingo. As a result, all that advanced context deciphering is useless if it is determined from the wrong set of words.

Negatives:

• Requiring a net connection for even basic capabilities the iPhone already possessed offline seems a needless reduction. Granted its not like the old off line assistant was in wide usage. However this to me is the best evidence that SIRI is almost entirely cloud based. If it was doing any kind of end to end interpretation of voice commands locally you logically would have included prior capability.
• Word recognition is improved but still ultimately hit or miss… and on the second tries I now I feel like I am the idiot who thinks talking louder and slower will translate english into another language. Real english phonetic overlaps definitely cause SIRI problems Purana:Pirahna/ G’s :Cheese are a couple of examples I found
• To helpful: there needs to be a basic override phrase or action that always stops SIRI cold. As is SIRI attempts to interpret and if it thinks it figured something out and it can do it then it acts. I had LOTS of attempted misdials when doing non-phone stuff because tying a contact to communication seems to be the lowest common denominator. Word of warning.. if you hit a point where SIRI is showing you a list of possible contacts that you didn’t want just say cancel instead of trying to repeat your previous command. At that point whatever else you say will get produce a ‘strongest match’ to the listed contacts.
• Not helpful enough: I sort of understand not being able to create new reminder lists, contacts etc… The ‘to helpful’ problem could be disastrous if SIRI had create/delete power. However I am struggling to understand why SIRI is not even allowed to launch various applications or at least the basic Apple provided applications.
• Basic word capture is still not ready. Its better that is for sure. But this is what everything else turns on. Kudos to the SIRI developers to get far enough along to try and do more advanced interpretation of language. If you are not sure what I am talking about go open up a note or pages document or something with the ability to take diction and actually try to dictate more than a short basic sentence. There is no automatic punctuation. Word accuracy seems to drop dramatically as longer phrases/passages are attempted. Doing each sentence individually is perhaps a solution but still awkward. Catch 22 because if SIRI can’t automatically determine sentence structure that means you still require some kind of voice prompt language to tell it… and the process of telling the machine to punctuate as you pontificate is the classic problem people have with voice recognition being ‘un-natural’.

Positives:

• Context interpretation is good. SIRI follows natural language progression of dropping proper nouns pretty well. This was and is a major sticking point on most voice interaction systems that had no capacity for following a conversational thread. It is often what made it seems so ‘stilted’. This is still fundamentally hampered by problematic basic word capture.
• Music control is impressive. Starting playlists, genres, artists etc… all seem to work pretty good so long as the words are correctly interpreted which was most of the time I played around with this particular capability
• Wolfram Alpha searches are impressive.
• The Alarm/Timer interface is genius. This may very well be the feature that keeps me coming back to SIRI while they are honing the other areas.
• Text/E-mail interface is the holy grail I think… but its in view, not accomplished. Had LOTS of problems doing this, and ended up random dialing of contacts a lot. May be more useful on the receiving end than the sending beyond basic stuff. In the positive section because once you understand its limitations it works well. Most of my problems came from screwing around with it.
• Meeting setups are similar to e-mail/txting. Fabulous when it works…. again, WHEN it works.
• Reminder list works well within its limitations… but lacking when it comes to such things as making a new list or trying to add things to a list with a name that is a bit overloaded. I tried I don’t know about many iterations of trying to add things to a grocery list and ended up with a list of nearby grocery stores far more often than I did with items in the list.

Final verdict : Solid Beta. I think Apple made a good call in pushing this out to 4s owners only. Early adopters are generally far more forgiving of new feature warts than the general masses. SIRI definitely has warts. Knowing that Apple tends to pride itself on shipping complete features I surmise that the key factor in improving SIRI at this point is crunching through massive amounts of data only possible with a large varied real world audience. The next couple of generations of this technology could cross a tipping point and finally make voice a solid form of interface instead of a flaky side show. It will be interesting to see if Apple gets there first as this could easily be the next tipping point in device preference.

iCloud:

I shied away from any specific review of iCloud in my 4s take because unlike SIRI it is iOS 5 dependent rather than 4s dependent. Apple fumbled its last major cloud initiative (mobile me) and it marks one of the few black eyes Apple took under Jobs second tenure. What made it even more remarkable is that it is a black eye Apple never really healed. They tried a couple of major revisions and never really got it even to ‘good enough’ state. iCloud is ultimately a clean slate re-build.

Negatives:

• PhotoStream has some major flaws. Flaw number one is that it is either on or off. That is either ANY picture you take is flying around to all your devices or NONE of them are. Removing photos is also an all or nothing process and is one that must be done on each device. Translation… have one photo up there you don’t want you have to reset photo stream on iCloud.com and then reset the photo stream on all devices. How they shipped with this basic problem for something as potentially embarrassing as photographs is beyond me. Serious miss. Also not sure why you can’t access the photos from the iCloud website as that would seem to be one of the most useful capabilities. I could keep going… how about being able to share to others photo streams? (parents taking pictures of their kids with separate iTunes accounts for example) Or allowing others access. Granted not having these capabilities is what lets apple get away with horrendous ‘undo’ control of what gets into the photo stream. But it is those types of uses that will make it most use full in the long run.
• iWork on the Mac has been left out in the cold which seems to be a fundamental flaw. You have to upload/download from the site only for these applications and there is no update to bake the ability in. Of course this is fueling even more speculation that iWork is finally about to get updated to a new version.
• E-mail sync is for apple mail only. Thats great if you use it but most don’t want to shift e-mail addresses just for this feature. My problem with this is that note syncing is tied to e-mail. Notes should be free players able to cross device sync on their own like reminders.
• No general file sync option. Having a folder on your mac that is synced with the cloud ala Dropbox with some granular control on the website for external access would have been killer. Many have been saying that Apple should buy Dropbox. Here is another vote. Either buy them and integrate into iCloud or sherlock the idea and implemented on your own but DO IT. If iWork were the basic means of office document work then this would be less of an issue. But Office still rules and you need this kind of capability to sync Office documents or any other kind of non-mac but highly common/standard file type (pdf, txt, rtf, csv) etc…

Positives:

• iCloud sync breaking the cable (or new wireless) tie with a desktop running iTunes is long over due and other than initial growing pains of MASSIVE traffic looks to be the major bright point of this new service.
• iWork on the mobile side seems to be well implemented. Just seems to be half a solution without the desktop version along for the ride.
• Cross device app synching is nice… but I am already seeing a problem where I want more granular control than sync all of them or none of them. Example… games going to my personal and work phones that are both tied to my personal iTunes account. Sure, I could create an iTunes account for the work phone but then I would have to pay for lots of productivity apps twice that my company does not provide.

Final Verdict : Alpha and Beta. App sync, mobile iWork etc… seem polished and only troubled by massive initial traffic encountered with the roll out of iOS 5, they represent the beta quality points. Lack of granular control of distribution to multiple devices may seem to be an ‘ease of use’ decision but I think Apples refusal to address the intermixed nature of device use is what keeps hosing them on the cloud. Simple things like shared libraries between spouses are a long known issue. Also on the horizon are new issues like work/personal phones or even individual dual purpose phone use. These are the areas Apple is going to have to get better at. Competitors are finally starting to figure out the whole integrated hardware/software angle and they are looking to leapfrog on these particular issues. Final nail in the coffin is no basic file directory content syncing with the cloud. Apple may want to push everyone into a world where we no longer root around in the file system… but it just isn’t going to happen. At least not any time soon.

iCloud and SIRI type technologies are the next generation of uberness for folks in the business of selling computers and mobile devices. It is one of the final links in integration of computers into our lives. To go beyond these two capabilities will require physical interfaces (i.e. brain/nerves to computer). These could easily lead the next 20-30 years of computer advancement. An always on/connected/interconnected frame work tied to natural language interaction is conceptually huge and while nothing new… the possibility that the technology is finally available to make it a reality is exciting.