“If bots take over from humans, we only have ourselves to blame.”

I made this statement in an interview with fellow analyst Brent Leary earlier this year to highlight the fact that “AI in the enterprise,” or “Conversational AI” as it is sometimes called, exists to help make contact center personnel more efficient and customers more product and happy. Talk of “agents training their robotic replacements” or refrigerators robbing us of our privacy in order to send us supposedly personal, but often ill-targeted alerts or suggestions, are headline fodder, but definitely not the ideal end-state for Intelligent Assistance.

That said, when it comes to that “ideal end-state,” we’ll know when we get there, but we’re not there yet. Fast-forward to roughly 2028. Our self-driving car will know who we are (as individuals), understand what we say flawlessly and remember where we left our car keys (oh wait, we won’t need those keys anyway). Flavors of machine learning and basic pattern recognition will be seamlessly integrated into everyday objects, not as an existential threat to mankind, but in order to provide assistance, i.e. “Intelligent Assistance” aka “Intelligence Augmentation” (IA not AI).

It will be the future as rendered in the Star Trek movies co-mingling with that of Minority Report (without the dark, dark implications ingrained in nearly all movies derived from Phillip K. Dick novels) or a benign rendition of Skynet from the Terminator series.

Okay, Pollyanna

That’s how I’ve been addressed when I deliver my typical “vision statement.” This is typically followed by “Just how does one avoid going to the dark (and deeply annoying) side between now and then?”

That’s a great question, and one that I’d phrase this way. “What should brands and business do now – when many of the elements of ‘artificial intelligence’ are underperforming buzzphrases – to ensure that the power of machine learning, natural language processing, knowledge management and biometric ID & verification are melded into trusted solutions that people use freely and confidently every day?” In other words, what will transform core technologies into commodities that are building blocks for trusted, opti-channel digital commerce.

Job 1: Bake in Trust By Design (or How to View GDPR)

In 2028, the year 2018 will be remembered as the year privacy laws got teeth. Europe’s General Data Privacy Regulations, which were formally agreed upon in 2016 and went into effect in late April 2017, will go into full force in May 2018. From that point forward, mishandling personal data becomes a very expensive mistake.

The regulations spell out terms and conditions for businesses to collect, store, share and otherwise use data about the individuals with whom they do business. They return control of personal data to the individual. Such data can only be gathered subject to informed consent and it must be stored “pseudonymously” (i.e. encrypted, and free from association with a specific person without using additional information or a “key”).

User control also means portable. Ideally, everybody will be able to transfer their data (like health records) from one electronic processing system to another. Finally, there is the “right to be forgotten,” which has been replaced by the “right to erasure” enabling individuals to request erasure of personal data from such things as searches on Google.

GDPR related fines -– which can reach 20 million Euro — are onerous enough to give companies financial incentive to establish trusted repositories for data. That’s a major step toward making the services of intelligent assistants, virtual agents and bots more trustworthy from the get-go.

In 2018, this means you must get acquainted with conversational and pseudonymous front ends to Blockchain. It has little or nothing to do with cryptocurrencies or virtual cats and everything to do with turning control of personal data to the individuals from which records and activity streams are originated.

Job 2: Treat Conversations as “Content Co-Creation Opportunities”

Hundreds, perhaps thousands, of Conversational Commerce solutions started out as glorified, conversational frontends to static FAQs. Yet, as we set our sights on 2028, we realize that intelligent assistants –- be they message bots, chatbots agents or virtual agents -– are being called upon to do much more complex tasks by quickly recognizing or predicting the intent of an individual and taking the proper action.

At the Opus Research Intelligent Assistants Conferences and upcoming 2018 Conversational Commerce Conferences, forward-looking companies like USAA, Nationwide Insurance and Ticketbis (a global ticket marketplace) showcased the ability of current solutions platforms have moved rapidly to carry out transactions or provide recommendations based on ingestion and analysis of voluminous amounts of dynamic data input. Customer records from CRM systems, transcripts or prior chats or call recordings, product literature and agent training materials are the raw material for today’s transactional interactions.

The next stage in the conversational continuum should be thought of as “content co-creation.” The new generation of intelligent assistants are constant companions, ever attentive to each individual’s needs and intents. Together they can generate the narrative that shapes experience and execute plans surrounding daily activities. As an early example of what is possible, Google Assistant’s response to “OK Google, Good morning” is to provide updated local weather and calendar reminders along with news headlines. Say no more.

Intelligent Assistants, like Google Assistant, assemble, interpret and infer “what you need to know today” based on past activity, behaviors and indicated preferences.

Ultimately, the role of an intelligent assistant is to make complex tasks simple through a process of constant verbal and non-verbal conversations. From what I can tell, Luka’s new chatbot app called Replika is making the most transparent attempt at enabling co-creation by explicitly walking you through the process of building your own avatar, or 100% AI friend. Time will tell if this is the right path to success for content co-creation.

In 2018 brands are preoccupied with steps they can take to avoid disintermediation, and rightfully so. If Facebook Messenger is the platform of choice for your bot, you well know that Facebook already has access to the activity streams, likes/dislikes, browsing patterns and posted opinions of your customers. Using the tools of Intelligent Assistance to build a conversational relationship with customers and prospects, as companies like Automat.ai (with L’Oreal) and Snaps with Nike’s Jordan brand have done is the way to build trusted bonds with individual customers and prospects.

Don’t Wait: The Future Starts Now

Opus Research has documented the growth of intelligent assistance across a broad spectrum of verticals, company sizes and application areas. We started the year with our Decision Makers’ Guide to Enterprise Intelligent Assistants, in which we documented the fact that the 28 solution providers that we evaluated had brought about over 2,700 deployments of intelligent assistants to a over 1,200 a enterprises or brands around the world. We’ve also provided advice to companies contemplating text-based messaging approaches in reports addressing text-based virtual assistants and chatbots.

Whether they emphasize deep neural networking, big data & analytics, and robotic process automation, or bring value through low-cost, speed-to-deploy and quick results, businesses around the world will be able to find a solution provider that is affordable and fits specific needs. Time to get in position on the path to 2028.

Categories: Intelligent Assistants