This year's explosion of the chatbots industry has been exponential, and the impact of virtual assistants can be felt throughout the tech world. But why now? How will the rapid spread of bots shape future AI applications and customer interaction? As part of our ongoing speaker Q&A series, we asked experts in chatbots and virtual assistants for their predictions for the field over the next 5 years, and why conversational AI is currently gaining so much success.
Tania Peitzker-Lingham is an AI bot technologist and Co-founder of velmai (Virtual Empirical Lifeforms with Multifunctional Artificial Intelligence). Tania has been working with and championing the commercial uses of chatbots for 7+ years.
Alberto Chierici is COO and Co-founder of SPIXII, an automated insurance agent. Alberto is a data scientist and has worked for large insurance companies and consulting in actuarial roles, advanced analytics and developed machine learning solutions for road safety and motor insurance.
Kuan Huang is Founder of Poncho, a New York based company aiming to make weather fun again. Kuan has a background in product design and software engineering, following a Masters at NYU where he learnt to build "technology products with soul".
Andrew Magliozzi is Co-Founder of AdmitHub, a chatbot that provides free expert college admissions advice. Andrew has dedicated his career to developing tech to help students succeed, whilst also founding Signet Education and the FinalsClub Foundation.
Ashley Hathaway is Senior Developer Evangelist at IBM Watson, who are renowned for their expertise and advancements in virtual assistant technology. Ashley acts as the voice of the developer within IBM Watson, to help teams create the best product possible.
Arte Merritt is Co-Founder and CEO of Dashbot.io, a bot analytics platform that helps users increase engagement, acquisition, and monetization. Arte built the first version of Motally, and led the company through a successful exit with Nokia acquiring it.
What do you feel are the leading factors enabling recent advancements and uptake of virtual assistants?
Tania, velmai: The American tech giants only opened up use of their APIs to bot developers in 2015. This has suddenly created critical mass as evident in the 11,000 and more chatbots now running on Facebook Messenger. velmai expects FB's rivals Microsoft's Bot Framework and Slack's Bot User Integration feature for developers to soon catch up. For instance our chosen launch platform Kik has already leapt from around 100 "robust bots" in spring to over 1000 now operating on the platform as of summer.
As the Canadian unicorn chat app Kik Interactive says, "first there were websites, then there were apps, now there are bots - 2016 is Year of the Bots". velmai agrees with that market summary because we have had to wait nearly 10 years for this tipping point where the dominant industry players - the tech giants I've named - have provided the enabling technology. Bot developers simply could not afford to scale up on their own using their own servers until now. But a note of caution: though there are now tens of thousands of bots within the space of year, is there a correlating uptake of virtual assistants by users? I think bot developers and the bigger tech platforms who are determined to feed off them are faced with two universal market problems: trust in the cybersecurity behind a particular bot and over-saturation of the chat app space by mediocre chatbots.
Alberto, SPIXII: There are many factors and a short answer is certainly a reduced vision. I'd like to point out some thoughts on social trends. It's funny, virtual assistants have been around from ages but never really made it to the mass adoption. So, more than a technology trend, I would see social trends (of course induced by technology trends) as the forces leading the uptake of VAs. Just a few ones:
1. We have about 20 years of search engine experience now, which led to the habit of querying technology and expecting an answer quickly. 2. Texting - there are hundreds of billions of text messages sent out daily from smartphones. It all started with the email, than the Instant Messaging over internet, then on mobile, unlimited text messages were included in telecom companies' tariff plans (hence it became cheaper than calling), then the 3G/4G networks opened to apps for 'free' and more engaging ways to chat. 3. Speaking to machines - when SIRI appeared on the iPhone for the first time it was mostly used as a game rather than for its real purpose, but usage apparently increased over time. People used to feel weird to be in a crowd and talk to a VA on their phones. But now we see more and more people sending voice messages on WhatsApp for example. It starts being less awkward to talk to technology, either because "others do it" or because of laziness - it takes longer to type.
Kuan, Poncho: 1. Utility is usually the first thing most people look for from a product. If you look back at the early days of mobile apps, social network, web, desktops softwares, products that focus on utility are usually among the first wave of innovation at each technological cycle. The rise of virtual assistants fits that pattern.
2. The commoditization of the core technology. Some NLP, AI, machine learning technologies such as Wit.ai (Facebook), IBM Watson, Google Cloud, etc are available and free for everyone to use now. The technology barrier has been lowered. 3. Messaging platforms are eating the internet. If you look at the amount of time people spend on their mobile apps, a big part of that is in the messaging apps. The platform shift definitely has accelerated the uptake of virtual assistants.
Arte, Dashbot: Conversational UI/UX is the future of human computer interaction. The Amazon Alexa is a great example of this. My co-founder Jesse often tells this story of how we used to see the videos of two year olds swiping on iPhones and iPads, now with his two young kids they're talking to Alexa - "Alexa play the happy song... Alexa play Star Wars..."
All the major players are opening up their platforms - Facebook opened up messenger, Apple opened iMessage, Microsoft opened Skype, Amazon opened Alexa, and Google is building an Alexa competitor. The brands are right there along side them building bots on these platforms. The users are already there as well - Facebook has nearly 1B users on Messenger.
What developments can we expect to see in chatbots and virtual assistants in the next 5 years?
Ashley, IBM Watson: For developers, the ability to pick and add custom bodies of content or corpuses. I.e.- add slang greetings, add common questions, add questions specific to banking, etc. That will happen pretty quickly. For users, more successful interactions, deeper interactions, and way way more personalized experiences. I think we'll be staring at our phones a lot less in 5 years and talking a lot more.
Andrew, AdmitHub: Knowledge work is on the cusp of monumental change akin to when the steam drill replaced John Henry. Virtual assistants are already proficient in many tasks -- communication, Q&A, scheduling, etc. -- but they require lots of supervision. At first, we'll supervise bots, because they need a lot of help. We'll continue this supervision longer than necessary for a wide array of reasons:
- lack of trust in bots who make occasional mistakes
- fear of having to do more complex work
- curiosity, spite and boredom
In the next five years, as virtual assistants fulfill the promise of reducing our burden of repetitive knowledge work, the largest changes will be our own psychology as we learn to accept and ultimately rely on this new technology.
Today, it is nearly impossible to imagine life without smartphones, Google or Wikipedia. Similarly, in five years, we won't be able to imagine how we ever lived without a virtual assistant handling all the minutiae in our daily lives.
Tania, velmai: Over the next few years, millions of chatbots are going to come onto the market to match the increased demand of 100s of millions of users who will want "service tech" primarily. This means VAs or AI bots will enable, ironically, humans to spend more time offline whilst their algorithmic avatars do the drudge work online.
Alberto, SPIXII: Chatbots and VAs are at the beginning of the technology adoption life cycle so gaining the trust of the early adopters is crucial. For example at SPIXII, to gain this trust, we start by selling simple insurance product that people are already used to buy online. Once people become accustomed to it, we will expand to more complex products. Chatbots will gain trust by doing things well in a repeatable way. These things have to be simple at first. If SPIXII can’t fulfill the expectations of a customer, then the customers can always talk to an agent. We will learn from this situation to teach SPIXII.
Our CEO, Renaud, recently tweeted a quote from David Marcus at a conference in Paris: 'Chatbots were over-hyped recently but are under-hyped for the long run' says @davidmarcus @VivaTech - I can't agree more.Many share the view that chatbots will be more and more present in our life and we still lose an incredible amount of time doing things we shouldn’t do. Think about setting up a meeting with someone else. It is very time consuming and there are some bots out there doing it for you. Chatbots will become your helpers saving you time when ordering food, clothes, buying products, even complex ones like insurance policies! Chatbots are a natural expression of the communication between human and technology. Technology is everywhere and chatbots are an elegant way to make it accessible to everyone.
Arte, Dashbot: With conversational interfaces, in the short term, we should see the major platforms launching bot stores, if they don't already have one, adding monetization capabilities, and providing multi-user bot interactions, if they don't already.
The multi-user bot interaction is particularly useful - if you're communicating with friends or co-workers in a messenger platform, it's a lot more useful if you can pull the bot into the conversation, rather than have to bounce out. Imagine searching for a restaurant - instead of bouncing out to Yelp, you can bring Yelp into the conversation. The AI and NLP will get better overtime as well - making the conversations more natural, with better response rates and understanding of the actual intent. We'll most likely see more interoperating amongst bots as well - having bots interact with other bots. Learn more at the next Virtual Assistant Summit taking place in San Francisco on 26-27 January! Speakers include Roberto Pieraccini, Director of Advanced Conversational Technologies at Jibo; Jordi Torras, CEO & Founder of Inbenta; Anjuli Kannan, Software Engineer at Google; Alonso Martinez, Technical Director, at Pixar; and Milie Taing, Founder of Lili.ai.Discounted tickets are available until 7 October, book your pass at an Early Bird rate and view more speakers on the event website here.