AI in 2019 will be all about bots and pre-trained models

AI assistants will become a fixture in the enterprise

This past year was big for voice assistants. With significant product releases from Amazon, Apple, Microsoft, Google, Samsung, Baidu and others, vendors are flooding the market with these products so that people can become more comfortable with conversational modes of interaction. While mostly focused on the consumer audience, it's clear that vendors are shifting their attention to the enterprise.

Just as 2018 was the year of consumer voice assistant overload, 2019 will be the year that enterprises will see widespread adoption and implementation of voice assistants. Already, companies are realizing the benefits of AI-based conversational technologies as extensions of their business to support a wide range of tasks.

In addition to relatively trivial workplace tasks, such as scheduling, basic information searching and assisted conference calls, human workers are turning to their AI assistants for help with more complex operational tasks, such as handling email, processing expense reports, and providing Augmented intelligence capabilities and other deep conversational features.

What will turn these novelty devices into more useful enterprise assistants is an increase in their intelligence. Research firm Cognilytica released a benchmark in 2018 showing that these voice assistants lack the critical intelligence and common sense reasoning needed to assist with most critical enterprise tasks.

However, we are already starting to see movement from voice assistant vendors to increase the intelligence in their products. In this light, we predict that 2019 will be the year that these devices increase their overall knowledge and intelligence and add more value as augmented intelligence tools.

For AI in 2019, it will be the year of pre-trained models and third-party data sets

Data is the language of AI. Data scientists need large amounts of high-quality training data to train machine learning algorithms to get the predictive and analytical results they need. After all, there can't be any machine learning without learning, and learning can't take place without clean, well-labeled data.

With the ever-growing need for data across an increasingly wide range of applications for machine learning, it's clear that 2019 will be the year that pre-trained machine learning models, third-party data sets and models, and open source training data will move front and center.

Indeed, we're already seeing major AI cloud computing vendors take steps to move beyond simply providing infrastructure technology to build reusable data sets applicable across many different industry sectors.

At the 2018 AWS re:Invent conference, Amazon announced that it is now offering pre-trained machine learning model packages in its AWS Marketplace. With this, people are able to choose from a variety of free and paid algorithms and models that cover a wide variety of categories, including computer vision; natural language processing; speech recognition; text, data, voice, image and video analysis; and predictive analysis.

Other models are industry- or application-specific, such as Amazon Forecast and Amazon Personalize, which use Amazon's own expertise in forecasting and providing personalized recommendations. The company went a step further announcing it can now use data sets in the medical industry, and it soon expects to add law data sets.

Not to be outdone by Amazon, Microsoft announced updates to its Azure Machine Learning platform for domain-specific machine learning modeling, and it placed greater emphasis on AutoML capabilities -- namely the ability for the system to automatically select and optimize machine learning algorithms and perform feature extraction, algorithm selection and hyper-parameter sweeping.

Google, IBM, and others are also enhancing their platforms with more of these out-of-the-box machine learning capabilities, broadening the adoption of AI and machine learning beyond data scientists and developers to line-of-business workers. These providers are learning that AI development shouldn't be limited to the people and companies with the largest and best-trained data sets. In 2019, the power of these data sets will be available to all.

Chatbots will dominate customer service

In 2018, companies adopted AI-enabled chatbots on the front line of customer engagement. Many of the use cases written about and discussed in 2018 included adoption of AI technologies for customer support. Companies that have AI systems to help with customer support are already seeing positive ROI, greater employee and customer satisfaction, and quicker time to resolution.

While 2018 was the year of chatbot adoption in customer service, 2019 will see the absolute dominance of the chatbot across most customer interactions for consumer-oriented companies.

Companies using AI-enabled chatbots to handle and filter customer inquiries are finding that their call center and customer engagement employees are being freed from routine first-tier support requests, enabling them to handle escalated customer issues that require more time or personal interaction.

Online travel booking providers are now increasingly using AI, which helps provide tailored suggestions based on customers' recent searches and booking history. AI-powered chatbots are acting as front-line support, offering valuable customer engagement. Customers are able to converse with these bots at any hour to get help with a variety of issues, such as booking or changing a flight, booking a hotel room, making restaurant reservations, or getting suggestions of local tourist attractions based on user preference and seasonality.

Restaurants have also jumped on the chatbot bandwagon and are using them to boost engagement and interaction with their customer base, while also reducing customer support costs. AI-enabled bots are proving valuable at a variety of things, like helping manage reservations, respond to customer inquiries and customize customer orders, freeing up staff to spend time with the customers currently in the store.

This trend will expand significantly in 2019, with companies in a wide range of industries adopting chatbots. Whereas the 1990s were dominated by the telephone call center and the 2000s saw the shift to websites and email as a means of interaction with customers, this upcoming decade will see the self-service chatbot be the predominant method for first-line interaction with customers.

A move toward responsible and ethical AI

2018 was the year that the abuse of customer data came home to roost. Facebook is under significant pressure for how it exposed user data. Companies like Equifax saw their brand sullied by data disclosures and Twitter, Google and others are finding it harder to convince their users to part with data to benefit the company's advertisers. Ordinary people are losing their faith and trust in corporate stewardship of their data. For better or worse, this will influence AI in 2019.

From autonomous vehicle crashes and automated social network bots to the increasing use of facial detection technology, companies are realizing that the use of AI technology has a far-reaching impact. In 2018, we already heard calls for companies to use AI in more ethical and responsible ways.

2019 will be the year this comes to a crescendo. While Google and other companies formed AI ethics committees and organizations, more will need to be done. 2019 will be when we will see the first laws and regulations from governments on the use of AI in public spaces and the use of people's information. 2019 will also be the year that corporations will start forming real AI ethics oversight and governance bodies.

Perhaps 2019 will also be when companies will have to disclose how customer data is being used to train machine models and give individuals and companies the ability to opt out if necessary through new regulations. Regardless of how it shakes out, 2019 will be the year that ethical and responsible AI becomes a headline.

The momentum continues

As AI adoption continues to increase and companies implement cognitive technologies across a wide range of uses, the excitement and expectations will only increase. Rest assured, AI in 2019 will feature unique breakthroughs, disruptions and products, and it will be exciting to cover all these developments.