Mark Zuckerberg often has ambitious New Year’s resolutions.
Learning Mandarin, reading two books a month and writing daily thank you notes have made his list before. But this year, Zuckerberg has committed to what may be his most challenging resolution of all: building an intelligent personal assistant (IPA).
“My personal challenge for 2016 is to build a simple AI to run my home and help me with my work,” he writes in a Facebook post from January 3. “You can think of it kind of like Jarvis in Iron Man.”
Zuckerberg goes on to say that he envisions AI that can keep an eye on his newborn daughter and even control the lights, temperature, and music in his home. He wants AI that can recognize his friends when they ring the doorbell and that can help him visualize data better.
Ambitious? Very. But not impossible. And “personal challenge” though it may be, it wouldn’t be surprising if Zuckerberg applies this AI to Facebook in the future.
This isn’t the first time we’ve seen big companies think about AI. AI has previously shown up in the form of Apple’s Siri, Google Now, Amazon’s Alexa, and Microsoft’s Cortana. However, this dream hasn’t seamlessly become reality.
Despite the hype that surrounded the iPhone 4S with the release of Siri, the IPA had some drawbacks. Siri had limited functionality in that it could only perform tasks it had been programmed to do. Sometimes, it didn’t understand what was asked of it. A response to Siri’s drawbacks has shown up in Viv — created, in fact, by the same people who made Siri. Viv’s creators have a goal to go far beyond Siri’s capabilities. If Viv’s creators are successful, Viv will not only be able to do more for its users, but it will have a memory. That means it can learn from its interactions with people.
Zuckerberg is no stranger to AI. This isn’t the first time Facebook has tried out artificial intelligence: Facebook dipped its toe into the AI pool with M, the personal assistant inside Facebook Messenger that finds information and completes tasks for its users. M was launched in August of 2015.
Though I haven’t yet tested it (it’s not available in my area at this time), M is a sort of hybrid virtual assistant that uses machine learning alongside human training wheels. (They’re referred to as “M Trainers.”) This composite style that includes human intervention is aimed at helping M deal with queries it doesn’t understand or doesn’t know how to handle — up until the point that it does.
Zuckerberg’s goal is a big one, and he may or may not be able to accomplish it in 2016. However, we can see that his goal is one aimed at making the internet more useful, and our lives more efficient.
What Will Facebook’s AI Mean For Law Firms?
Imagine a dialogue like this:
“I dented my car and I need it fixed.”
“Here are several auto body repair shops in your area. I’ve found three that have a five-star listing. Would you prefer to see only these listings?”
“Yes. I need you to contact the listing for Bob’s Car Repair and book me an appointment on Thursday during my next free hour.”
“Okay. I’ve booked you an appointment on Thursday at 1 PM. I see 1 PM is a time you often eat lunch. Would you like me to show you restaurants that are near your appointment?”
Is this a conversation with a thorough assistant, or with your phone’s AI?
It could be both. It seems futuristic, but the attention AI is getting from big companies may be signaling that this is a reality we’ll see in the very near future.
This is where we’re headed when it comes to AI and search. Soon, voice command could be more useful and efficient than typing a query into a search engine. People will be able to ask AI questions in the same way that they might with an actual person, and they may be able to do it from a variety of devices.
Siri is only capable of doing what its programmed to do, but the future of search is machine learning that requires little in the way of human intervention. Rather, the machine learns from its interactions with humans. Let’s take the above dialogue as an example. Perhaps the person in question is a vegetarian. The goal is that AI will remember that fact from a previous interaction and may offer restaurants that serve vegetarian options (and book a table, and give you directions). Or perhaps the car they dented is a BMW, and if the AI remembers this, it will return results that are local car shops specializing in European car model repair.
So how does this apply to you, the attorney?
This is how potential clients may be interacting with your firm in the future. They may ask their AI about what to do after an arrest, or ask what steps they need to take to get a divorce, or any number of questions that may lead to them engaging with your law firm. We’ve already talked about how potential clients use varied terms to describe their situation and problem, and the future of AI may be able to do more in the way of understanding intent. Google’s working on that with their release of RankBrain — machine learning that is starting to understand how people naturally use language — to give search engine users the most relevant results.
Furthermore, this AI may be able to actually book an appointment for them. Looking at the Facebook M testing among employees, one of the most popular uses was this:
“…The service can call your cable company and endure the endless hold times and automated messages to help you set up home wifi or cancel your HBO.”
The Wired piece notes that, right now, a service like that is still courtesy of an actual human doing the holding for you. (A benefit to M’s two-pronged approach in the absence of better technology at this point.) But soon, it may be that machine learning can do it on its own.
Not only will AI be able to make your life more efficient by correctly handling tasks, but it may also try to give you the best possible options based on available web information and what it learns about its users. We may see AI that, like in the above example, gives you choices like:
“I’ve found four bankruptcy attorneys in a two mile radius from your office. Two of them have marks on their record. Would you like me to eliminate these results?”
“…One result of five is highly rated. Would you like to read the reviews for this law firm?”
This is where the future will lead us. But one thing that lawyers can do right now to ensure that their websites are future-ready is to continue creating high-quality content that specifically answers the questions potential clients are asking — and regularly posting it to the web. Google’s bots are always busy scouring the web for new, relevant information, and the more significant information that Google has to work with, the more information it can offer search engine users. Additionally, the more valuable content attorneys post to their website, the more Google will return to your website to find and index something new.
What lies at the intersection of what Google wants, want search engine users want, and even what referral sources want is high-quality content. AI will likely use quality content to its advantage as it finds new ways to be helpful to its users.
As AI improves, AI users will likely become more comfortable asking it questions, having it seek out the answers, and even handling tasks like booking appointments. Attorneys will want to be sure their law firms are providing the most relevant and thorough answers for AI to provide its users. Whether it’s a Google query or asking Siri, significant, unique and educational content will stay an important part of our current information society.