Siri’s importance

Much has been said about Siri – I’ve made the case early on here that Siri is a big deal and it seems I’m not the only one thinking that (here, here and here). But besides its relevance in the near future – Siri is not the end solution of computer interaction, in fact its strength is also its biggest limitation: spoken commands.

Speaking to a computer is not always desired or even wanted. For example, interactions in the public with Siri could become the digital version of 2nd hand smoking. I don’t really want to know about someone’s appointments – it’s already pretty annoying when people talk on the phone in public and with Siri everyone could potential become annoying.
There are, of course, scenarios where you really don’t want to hear anyone (theatres, lectures, etc.) talking into their phones and have their phones talk back to them. Or think of a business meeting. Some things are better done in silence – not just for your neighbor’s sake but also because you probably don’t want to spread news about yourself or your company.
The other problem with Siri is that it’s just not useful for certain tasks. When I’m searching for something I often don’t know what I’m searching for. To paraphrase: “I know it when I see it” – all I do is try to narrow the search by limiting its scope. For example, consider a recent query of mine: “Why does the Nex5 have significant higher lumen sensitivity compared to micro four thirds?” While Siri is perfectly able to interpret what I’m asking, it won’t know how to look for it! That’s because it doesn’t know (yet) how to research questions around image sensors in mirror less cameras. I would expect in a couple of years (or maybe decades) that this is no longer an issue for Siri (but by then I suspect this question will be irrelevant since the image sensor in my toy key chain will be better than anything we have today). As it stands today, I have to enter cryptic Google searches that have the words “NEX5 light sensitivity Mft” if I want to find a close answer.
The other problems that Siri will have is in recognizing things – for example, “Siri, get me a picture of a Philips screwdriver”. Siri can only do a Google image search with Philips screwdriver but will have no concept of how to interpret the picture. In other words, it doesn’t know if there is really a screwdriver in the picture. It also means that Siri couldn’t actually describe any findings back to me. This is true for pictures as well a text.
Eric Schmidt is concerned about losing ad dollars to Siri – he shouldn’t be, or at least not in the foreseeable future. So in the context of transaction, I might not be a typical shopper but boy do I make comparisons before I buy which is still the forte of Google and Amazon. Siri as comparison and shopping engine? Not so much – consider this:

“Me: Siri buy me AAA batteries”
“Siri: from where?”
“Me: from Amazon”
“Siri: I found 600 different offerings of AAA batteries”
“Me: look for rechargeable ones”
“Siri: I have found 12 brands in 4 categories. Would you like to hear the brand names and the categories?”
“Me: which ones have the highest capacity?”
“Siri: I’m sorry Dave, I can’t do that”
“Me: pick Energizer”
“Siri: There are 10 products related to AAA rechargeable energizer batteries, would you like me to list them?”
“Me: errrr… no, just buy me 4 AAA batteries”
“Siri: There are packs of 2, 4,8, 16, 4 with charger, and 8 with charger”
“Me: 4 without charger”
“Siri: how would you like to pay for it?”
“Me: Credit card”
“Siri: which one? Amazon knows of 3, do you want me to list the last 4 digits of each card?”
“Me: no, use the Amercian express card”
“Siri: that card has expired, would you like to use another card?”
“Me: Yes, use the VISA”
“Siri: There are two VISA cards, which one?”
“Me: the blue one”
“Siri: I’m sorry Dave, I can’t do that”

My point is that, Humans are still beating Siri in most day-to-day tasks when interacting with search and the Internet in general and we’re still a long way off from Siri taking over these tasks (no doubt it will eventually).

But Siri is exactly what I always wanted for my desktop computer. There are real applications for this today, for example instead of me typing the following cryptic code into a cli:

“find . –name ‘*.jpg’ -size +200kb –print –exec mv {} ~/tmp \;”

I could just as easily say: Siri find all jpegs in my home folder that are greater than 200kb and move them to the temp folder in my home directory”.
We might in fact enter a whole new computing paradigm that makes our interaction with the operating system somewhat complete (more on that later) - for example, organizing one’s information and helping to retrieve it. Instead of me looking through hundreds of documents or searching for particular keywords, I really just want to say (or type!): “Siri open the document that I edited yesterday”. There should be an intelligent dialogue following:

“Me: find me the document I edited yesterday”
“Siri: I have found three documents with the following titles: ‘Attack of the bumblebee zombie’, ‘Perrier for everyone but the doplhins’ and ‘Siri’s useability’.
“Me: Open Siri’s useability and go to the end of the document”
“Siri: done”
“Me: change the paragraph at the end of the document from ‘and that’s sili’s ability’ to ‘and that’s siri’s ability’
“Siri: done”
“Me: Also, can you add the following text at the end: ‘I don’t like spring, it causes me to sneeze all the time’
“Siri: done”

See Ma? no files! Would I even care where the file was stored as long as I knew that it's backed up reasonably and retrievable with simple voice commands? We're entering the HAL 9000 era. If Siri can keep track of where I put my stuff and organize it so that I will always find it, now that would be neat.
But further, the question that is on my mind lately is about operating systems in general. The abstraction between the user interface and the underlying operation system tasks is still pretty crude after so many years. The graphical user interface was revolutionary – 30 years ago! As I pointed this out here, it’s just no longer an applicable concept with the growth in storage and CPU prowess, there has got to be a better metaphor for the user/computer interface.
But Siri is clearly part of this, one way or another, our interaction with computers will change to resemble more of a human-kind characteristics than the machine like underpinnings in today’s UI’s. The future is indeed groovy!