The Birth of the Virtual Personal Assistant
————————————————————————————————————
This blog post originally was featured on Mashable on January 31st, 2010.
————————————————————————————————————-
In the near future, anyone who lives a connected lifestyle will be able to delegate their everyday tasks to intelligent virtual assistants that will coordinate, execute and simplify users’ lives.
We will look back on these days and ask ourselves how we ever got by without our trusted assistants, the same way my kids ask in amazement about how we ever got things done before laptops and the Internet.
What Constitutes a Virtual Personal Assistant?
For a long time, Hollywood has been portraying machines that humans can converse with, delegate tasks to, and command. Remember the HAL 9000, KITT the car, COMPUTER from Star Trek, or even the brilliantly conceived and visualized Apple “Knowledge Navigator” from over 20 years ago?
They have symbolized our desire for trusted machine assistants that can help make our lives easier. They have persisted in the creative works of science fiction writers for decades. But have you ever asked yourself why that is? Looking beyond the theatrical and dramatic value of these ideas, the reality is simple — we have always desired more help, less hassle, and higher productivity in our lives.
What about search engines? Aren’t they the modern day version of this? No, at least not the search engines of today.
Search is a fantastic tool to help you find information on the Internet, but try to ask a search engine to actually do something for you. Try typing “get me a seat on the next flight from Chicago to Seattle” and see what happens. Or ask your favorite search engine to book you a table for three at Gibson’s steakhouse in Chicago for the day after tomorrow. Today’s paradigm of 10 blue links doesn’t cut it, and we need a new tool to help.
We need software that is specifically designed to help you get things done — a “Do Engine” rather than a search engine: A virtual assistant.
Intelligent Cohesion of the Tools We Already Use
Here is the good news: The elements, technology and ecosystem needed to build machines and software that can automate many of the mundane tasks of our lives are here already.
We just need to add a little intelligence. It will take some time, maybe 3-5 years, for the concept to mature. But when it does, it will emerge as the most frequently used and trusted online tool. It will make the most common actions on the web as simple as having a conversation. It will integrate into your life, get to know you, and be proactive.
In some sense your smartphone is starting to work like this already. There are already tens of thousands of services, apps, and sites that help you find and do things on the web and in the world. The problem is that they are all islands unto themselves, typically focused on a limited domain, and don’t often work together. They rarely share data or context with each other, have different user interfaces, and require users to spend a good amount of time to discover them, sign up, and get started. In terms of unified personal services, it’s not ideal.
Virtual assistants will help unify these and get them work together at your command. It would be nice to simply pull out your phone one day and tell it to move your 3 p.m. meeting to 5 p.m. and alert everyone invited of the change. That day is coming sooner than you think.
A New Chapter for the Web
There is a direct relationship between simplicity and user engagement on the web. Less clicks means more users — period. When combined with tools like smartphones, virtual assistants will migrate user interactions towards a far more frictionless e-commerce, consumption and collaboration model.
You will soon pick up your phone and start asking your assistant things like “take me to live CNN news,” “send my dad the latest John Grisham book,” or “tell Adam I am running 20 minutes late,” and you will then watch it all happen. This evolution towards simplicity of interaction will reduce the barrier to almost everything you use your mobile device to do.
Furthermore, the device is always with you. The combination of simplicity, impulse opportunity, context, and preference will create the most explosive market opportunity in ages.
This will be a market in which every player along the line wins. Users will be able to click less, enjoy simpler interactions and receive much-needed help getting things done and managing their day. Participating service providers get simpler discovery, more transactions, and higher consumption rates. This then drives more data dollars to networks, fueling infrastructure expansion.
As proof, witness what a cool device called the iPhone (iPhone) has managed to accomplish through a snappy and simple interface with shiny buttons and creative apps. That one device and the competitive response we are now seeing has created a complete transformation in computing.
The Anatomy of the Virtual Assistant
The OS of virtual assistants will be the Internet itself, as Kevin Kelly postulated years ago. The brains will be AIs that are developed by software companies for both general purpose and targeted domains. The arms and legs will be web APIs (many of your favorite brands and services), and the connective tissue will be authentication protocols like OAuth and Open Social, and trust circles like those of Facebook (Facebook).
The rapid maturation of technologies that enable free-form interaction such as natural language processing and speech recognition have vastly improved, to the point of gaining real adoption in many applications today (e.g. Google Speech, Nuance Dragon Dictation, Ford Sync for cars). Virtual assistants will leverage these inputs and begin to integrate them with conversations for a simpler, more natural way to get things done. This concept was best described by the late pioneer from MIT, Michael Dertouzos, who called it “human-centric computing.”
Over the long term, this paradigm will expand to many (or most) of the online services and tools we use to manage our lives like booking, buying, reserving, reminding, and scheduling. As we build trust in our digital “partner” we will put more and more onto its to-do list.
Trust is Key
The vague promises of contextual awareness, personalization, and other generalizations have rarely materialized in real products on the web. We are wary of what personal information we share online, in search engines, and the the never-ending fear of credit card fraud still looms. But this game is changing with the open web.
Mark Zuckerberg is indeed correct that privacy is dead on the Internet among the digital generation. Hundreds of millions of people spend a great deal of time telling the world all about their personal interests and information that forms their “digital face” on sites like Facebook, LinkedIn (LinkedIn), Twitter (Twitter) and others. This will only expand as the demonstrable benefits of this effort become more apparent.
The paradigm shift we will see with virtual assistants is that providing them with access to your preferences, tastes, accounts and more will be the cornerstone of the simplicity they will enable (within a very secure environment, of course). In other words, where we once feared how long search engines kept our personal information, we will now go out of our way to expend time and effort to specifically provide our trusted assistant detailed information about ourselves.
This will be done both manually and via syncing with existing sources of our personal data such as Facebook profiles, iTunes (iTunes) music lists, and contacts. The point is that you will make your virtual assistant definitively yours.
2010 and Beyond
The experience will be like hiring a new assistant that doesn’t yet know you, but eventually becomes so familiar that you can’t live without him or her. Keep your eyes on this space, try out these products as they emerge, and prepare to make your life a bit simpler over the next few years.
As John Battelle has said: “The future of search is a conversation with someone you trust.” 2010 will be the year in which we start to see real progress towards this vision, on many fronts.
Tom to give keynote at Web 3.0 next week
It’s been a while since our last post on this blog – we’ve been heads-down working hard for the past few months. We look forward to sharing what we’ve been working on; 2010 should be very exciting!
We also wanted to let you know about Tom’s upcoming keynote at the Web 3.0 Conference in Santa Clara next week. It is titled “Big Think, Small Screen: How semantic computing in the cloud will revolutionize the consumer experience on the phone” and it will be a moderated Q&A. You can find the program and session details here.
Tom will discuss some of the ways semantic web technology will be applied to consumer-facing products. The session should be very interesting. Just think about how much the customer experience has evolved in recent years – where consumers now make significant purchases on their mobile devices, recommendations and personalization have become the core focus of e-commerce sites, and on-site reviews are now ubiquitous.
The topic of semantic technology and customer experience is obviously near and dear to all of us at Siri, so if you’re able to attend the conference, I hope you catch Tom’s keynote.
It’s time to get back to work, but keep an eye out for exciting things from Siri very soon.
2010 South by Southwest PanelPicker goes live
South By Southwest today launched the 2010 Interactive PanelPicker. Users get a 30% say in what panels will be selected (there’s over 2K+ proposed topics this year). We submitted two panels. If you like our proposed topics, we’d appreciate if you’d vote for our panels so they will become part of SWSXi this year.
Here’s a brief overview of the panels we thought would be interesting:
Idea 1: “Search: a Conversation with Someone You Trust,”
Idea 2: “Software Agents and the Ethics of Assistant-ness.”
“Search: a Conversation with Someone You Trust” is an introduction to the concept of virtual personal assistants (i.e. VPA’S) and explores how the category will evolve, particularly as a new interaction paradigm for the Web. It also explores the many business opportunities inherent in VPA’S, particularly around Web services and API’s.
“Software Agents and the Ethics of Assistant-ness” takes a deeper dive into cultural issues relating to establishing and maintaining trust when interacting with virtual assistants. As we have said, we believe that in the next 5 years almost everyone who lives a connected lifestyle will delegate the details of their everyday tasks to intelligent assistants. Therein, developing trust between the user and the assistant is critical.
To vote for either of these panels, click on the link to the panel topic above and vote on the SWSX website.
Partnering with True Knowledge
We are excited to announce that we are partnering with True Knowledge, an answer engine which uses natural language to understand and answer questions. This is the first of many partnerships that we’ll announce over the coming weeks and months.
The collaboration is a great fit for both of us. True Knowledge offers the leading semantic answer engine, and as a result, Siri gets a lot smarter. In return, True Knowledge will be a big part of our rollout this year, and we’ll be introducing their service to millions of new users.
How does True Knowledge work?
True Knowledge taps subject matter experts around the globe to build its information repository, bringing together the benefits of machine-driven automation and people-driven intelligence. And similar to Siri’s human interaction paradigm, True Knowledge uses Semantic Web technologies to enable computers to work and think like we do, drawing inferences and conclusions to find the right information.
As a result, Siri will now be able to answer questions you might rely on your assistant to field. For example, you can check the local time in London so you don’t miss an important call or quickly look up trivia while you’re out with colleagues or friends; for example, “Who was the 41st President?”.
We at Siri like to quote John Battelle, who once wrote that “the future of search is a conversation with someone you trust”. This partnership moves us one step closer to that vision.
Joining the Verizon Developer Advisory Board
I’m proud to announce that today, we are joining the inaugural Verizon Developer Advisory Board, alongside a host of established mobile innovators.
The Advisory Board is itself a forum designed to provide feedback about how the Verizon developer community and Verizon applications store will grow and evolve, as well as to “foster a productive and innovative developer ecosystem.” Details of the announcement can be found here.
Additionally, I will be speaking today in San Jose at the Verizon Developer Community (VDC) Conference, which can be viewed via live webcast starting at 9:30 a.m. PT / 12:30 p.m. ET. Visit the Verizon Developer Community to register and log-in.
We’re bullish on Verizon’s plans – they are an established leader with a powerful voice and data network, serving 87 million customers. We’re looking forward to helping them make a successful entry into the mobile applications space, and we look forward to announcing more exciting developments over the coming months.
As we’ve previously announced, Siri will also be available for the iPhone at launch and other platforms next year. Siri is coming along nicely and our private beta is just around the corner. Stay tuned, and thanks for your ongoing support – we really appreciate it.
Siri Keynote at the Semantic Technologies Conference: Slides and Video
I enjoyed the opportunity to speak at the Semantic Technologies conference about virtual personal assistants (VPAs) and where the space is headed. Thank you for the thoughtful questions and great feedback. It was fun speaking at SemTech again, and to see the innovations that have been made in the past year. In addition to the fascinating conversations in the halls, I really enjoyed the panel on semantic search. Seems like semantics is going mainstream.
For people who missed the keynote, here’s a link to my talk, Siri: A Virtual Personal Assistant and video of the event.
See the second public demo of Siri at the Semantic Technology Conference next week
On the heels of last month’s D conference unveiling, our CTO Tom Gruber will be giving a keynote talk at the Semantic Technology Conference on Tuesday morning, June 16th, titled: “The Siri Virtual Assistant: Bringing Intelligence to the Interface”.
The keynote is also a follow-up to the talk Tom gave at last year’s gathering, as Siri first started to come out of hiding and describe the many opportunities inherent in assistant-ness. The slides from last May are available here for download, and CNET’s Dan Farber also did an elegant write-up of the concepts outlined therein.
The team here will be attending Semtech in full force, and we’ll also be adjoining to a separate room after the keynote to offer conference-goers and opportunity experience Siri for themselves.
Finally, we’ll be tweeting often, so be sure to follow @hisiri and track the conference hashtag, #semtech2009.
Hope to see you there!
Reflections on D
We’re now back from the All Things Digital conference (D7) and thought we’d share some of our impressions. Wow, what a line up! The quality of the speakers and attendees was unbelievable. D7 teemed with celebrities, tech visionaries, CEOs of top computing companies, and, inspirational to us, young successful entrepreneurs like Evan Williams and Chad Hurley. We caught a glimpse of Martha Stewart, sat next to Tony Hawk at breakfast, benefited from the wisdom of Balmer, Bartz, and Lazaridis who were all speakers, and had countless discussions in the hallways during breaks.
It was exciting to feel the energy and creativity of the presentations and the D7 Science Fair. Microsoft unveiled Bing, a self-proclaimed “decision engine” that aims to morph our concept of what a search engine can be. Palm showed off the Pre, a potentially worthy competitor to the iPhone, with a cool visual UI (multi-apps as cards), slick hardware that includes a pull-out keyboard, and some novel twists to their interaction paradigm (universal search, interactive notification bar). Canesta demonstrated a can’t-lose-it-in-the-couch remote control that relies on 3D hand gestures (no physical device) to control all TV functions. Plastic Logic showed off the new “Kindle Killer”, an e-book with a very thin profile. And, while sitting in the audience watching Dag unveil Siri, I heard people buzz, “they’d definitely use Siri,” and “if Siri can truly pull off what it is trying to do, it’ll be a game-changing product.”
From the vantage point of the All Things D conference, the digital future looks bright!
Video of Siri in action: http://kara.allthingsd.com/20090709/siri-the-full-d7-demo/
Siri will launch later this summer
The interest we’ve been getting in Siri so far is great. Keep it coming! A few people have asked when we will launch. In short, you’ll be able to download the Siri iPhone app later this summer. If you want us to ping you when we launch or to hear about other Siri happenings, please join our email list.
PRACTICALLY HUMAN
In the coming few months Siri will be born.
We are delighted to introduce a simpler, more intelligent, and more personal paradigm for interacting with the internet. Your virtual “personal assistant.” We believe that in the next 5 years almost everyone who lives a connected lifestyle will delegate the details of their everyday tasks to intelligent assistants to coordinate, execute and simplify the details of their life. I believe we will look back on these days and ask ourselves how we ever got by without our trusted assistant in the same way my kids ask in amazement about how I ever got by without laptops and the Internet.
As we prepare for launch, the headlines will undoubtedly herald the proud legacy and complexity of the technology required to bring the notion of the Virtual Personal Assistant to life, and rightly so. There may be stories or blog posts about how finally, and after many failed attempts, there is a “life-like” computer we can hold a conversation with and can start to manage our life for us. As much as I find this an appealing vision we are not there yet. Let me clearly shine a little daylight on that optimistic notion and say without hesitation… we are still a few years away.
The Siri of today is about simplicity and usefulness. Now. Siri will make it easier to ask for, find and do things while you are on the go via your mobile device. That is our promise today. Siri will be “conversational” which we promise will make it easy to use. Siri will get smarter all the time, will continue to learn new areas (which we call “domains”) and will be one of the most useful mobile applications you have ever used. I can also promise that Siri will get confused occasionally, get things wrong, or have a silly answer. But not most of the time. Be patient. Siri will be learning quickly and you will notice improvements rapidly.
What is our motive for creating Siri? Let me offer a personal perspective. I come from the world of mobile internet, where I led innovation in the US and before that in Scandinavia. Being a newcomer to Silicon Valley, I have been surprised by the seeming inability of the collective consciousness of Silicon Valley to escape the bounds of the great Search paradigm. Every new company is a potential Google-killer. Who has the biggest index? Whose results are more relevant? Who adds more pictures? Who cares.
If you step back and ask yourself a more fundamental question like, “How do you make the internet (especially the mobile version) a far simpler tool to help you get things done?” The answer isn’t 10 blue links. It’s about understanding what someone is trying to do, where they are, who they are. And then helping them do it. Make it easy to communicate, through speech recognition, intent understanding, and a simple conversational interface. How about those tens of thousands of iPhone apps out there? Love ‘em. But I’m now in application exhaustion mode. Which of my ten pages of app icons was that movie shortcut on again? What was my log-in and password for that events site? Hmmm.
Today the internet is a sea of web pages and disconnected data. What if we used this technology bring it all together. To work together, for us. In one place and through simple conversation. Use the impressive APIs of our services partners to integrate the experience of finding and doing. Just by asking. That is Siri today.
So where does this road lead us? If you think about how the drivers of this paradigm are evolving, such as the accelerating pace of available services and APIs, the explosive growth of powerful and (ever cheaper) mobile devices, the rapid maturation of technologies that enable free form interaction such as NLP and speech recognition, and the adoption of standards such as Semantic Web, OpenID and OAuth, you see a promising path. Our goal is to turn these enablers into a simpler way of getting things done and managing your life. An assistant for everyone.
And finally, any new major paradigm that involves multiple relationships requires the alignment of business interests. I’m happy to say that in the coming age of the virtual “Personal Assistant” everyone wins. There is a direct link between the time and effort required to do something and the number of people who actually do it. Fewer clicks, less time = more users, transactions and volume. Period. At Siri we haven’t yet met a partner who wasn’t interested in getting involved in this paradigm shift to simplicity. Everyone wins.
This is Chapter 1 of a long and (we think) transformative shift to simplicity and usefulness. Hope you like it.
Dag
CEO and Co-Founder, Siri
