What to Expect in the Next Two Years of Conversational AI
More than any time in our industrial history, people today expect always available access to research, advice, commerce and support from any location, device, or platform they prefer. Conversational AI provides companies the technological means for meeting that expectation in a way never before possible. It delivers interactive, knowledgeable, micro-personalized and virtual sales and service across an increasing array of connections.
Conversational AI is a category of emerging smart technologies which finally enable an innately human form of interaction between people and machines – natural language. More sophisticated than hard-coded, text only chatbots, the AI technologies in full featured Conversational solutions include automated speech recognition, natural language processing, machine learning and increasingly image recognition, semantic and sentiment analysis and smart analytics.
In 2017, our Conversational AI practice at Cognizant saw our clients began experimenting with these new technologies and exploring possible use cases. Already this year, many are moving past this exploratory stage and now view Conversational AI as an imperative for brand differentiation. As these innovation leaders move beyond theoretical evaluations of the technology to outcome-based projects delivering real business value, they are also seeing the need for an entirely new business plan to guide their thinking. We call this a Conversational Strategic Plan.
Typically led by digital commerce leaders, chief marketing officers and CIOs, the Conversational Strategic Plan defines the goals, approaches and tactics for leveraging not only conversational technologies, but also the many other AI enabled innovations now coming into focus. With such a nascent and fast evolving space this will be a first-time exercise for most and a steep learning curve for anyone just getting introduced to AI. To help these leaders know what to anticipate over the next couple years, our team compiled the following ten predictions for Conversational AI.
Ten Predictions to Guide Your Conversational Strategic Plan
1. The need for ‘academic’ experimentation is fading. 2018 is the beginning of the commercialization phase for conversational technologies and cognitive agents. Lessons learned were not lost on the fast movers and innovators, who are now doing what they do.
2. Conversational technologies are the launching pad the next gen user interfaces – augmented reality, gesture controls, biometrics, visualization and intelligence. Expect much more immersive experiences being realized using conversational AI technology. This will include proliferation of vision recognition as an AI service, which will lead to integration of vision and voice domains, as well as increased interest in augmented reality for virtual sales, training and support.
3. Rapid improvement in contextualization and self-learning. Currently even the leading AI services platforms expect contextualization to be implemented as a customization feature, however we expect these technologies will evolve to handle context innately and self-learn faster and more effectively than they currently do.
4. Accelerating IOT voice integration in devices including automotive, wearables, home automation, personal health monitoring, drones and consumer electronics. This will increase functionality, drive replacement and upgrade cycles and of course, produce rivers of data to feed analytics, machine learning, and business intelligence.
5. The business reality of $15 per hour labor makes the ROI of “conversational AI automation” impossible to ignore. Early lessons learned are already revealing both cost savings and revenue opportunities are there for those with determination, creativity.
6. Better language coverage. Currently there are big differences in language coverage between the tier one cognitive agent technologies. For markets outside US / UK&I (especially Europe and APAC) non-English language coverage is key for success. Expect an increased number of supported languages and dialects and that the proficiency of individual non-English languages will improve drastically.
7. Smoother hand-over between cognitive agent, live chat agent, call center agents and other channels. The user experience across channels should be consistent and aligned, whereas today the technology is not (yet) capable of handling all conversations on all channels. Ultimately, cognitive agent programs will not be implemented stand-alone but rather as a part of an overall interaction strategy that allows interlinked technology and change management across all platforms and channels.
8. Easier and faster configuration. Robust and high quality Conversational AI implementations require considerable effort to design, test, deploy and maintain. The leading providers know that to further lower business case barriers and achieve adoption at scale, the technology must become easier and faster to implement. This evolution of the products and frameworks, along with system integrators further leveraging standards, best practices, re-usable components and one-to-many / as-a-service solutions will also increase adoption.
9. Addition of other AI in cognitive agent solutions. Embedding other AI technologies, such as video, sentiment and emotion recognition, improved recommendation engines and user desirable integrations such as social platform, weather, traffic, news, and mapping API’s will make cognitive agents more intelligent and further improve user satisfaction, sales effectiveness and cost efficiency.
10. Start thinking integrated text, speech and images. Initially, conversational implementations have primarily focused around text-based chat. Meanwhile, intelligent speakers are proliferating, and speech recognition capabilities are improving, so speech-based cognitive agents are fast become increasingly important. Dedicated multi-modal devices incorporating text, voice and video are gaining traction and mobile phone / tablet technology already provides a ready platform for this triple play. Soon, integration between the three formats will be an expectation from users and must be accommodated in forward thinking solution and CX design.
There they are, ten predictions to consider while developing your Conversational Strategic Plan. Bottom line, it’s all changing fast, frameworks are morphing, architecture standards and best practices are evolving, new use cases are appearing daily, and we have yet to touch on the elephant in the room… the challenge of adopting your Conversational Brand Strategy to accommodate all the various agent platforms (Alexa, Cortana, Google Assistant, Siri) and their associated devices and products. Again, a subject broad enough to command its own post – stay tuned…
As always, I would love to hear from you. Tell us what we missed, your take on our top ten or share your favorite lesson learned. Thanks for reading!