The Future of AI: 2022 Trends

As we in the industry kick off another important year in the development of AI solutions, let’s look at the trends that will guide (dare I say, should guide) our thinking and execution.

‍

‍

1. Shift away from replacing humans with machines to augmenting human performance (from AI working alone to humans and AI working together).  The goal of machine learning for most applications has been the replacement of human effort with machine effort.  The focus will shift to machines performing tedious, tactical tasks (information retrieval) enabling humans to focus on higher-level, strategic tasks (e.g., decision making).
2. Shift away from delivering a complete set of developer tools and technologies for IT teams to delivering comprehensive AI-enabled solutions for line of business teams. Much, if not most, of the AI industry has been focused on developing robust tools for internal IT teams or consulting organizations to  apply the technology for a specific use case in an enterprise application.  The 2022 trend will be to develop specific AI-enabled solutions that can be immediately implemented without any development effort.  For example, one begins with an empty Excel spreadsheet and puts their data in to operate on it.  The trend will be to provide AI-enabled solutions where the business user can start to apply their own data/content and immediately begin to operate on it.
3. Shift away from a focus on the robustness of the technology to the robustness of the user experience (i.e., communicating using human language).  The AI industry has focused most of its attention on the building of AI platforms and associated tools.  In 2022, expect to see AI vendors deliver more “out of the box” solutions where the focus will be on providing a rich and familiar user experience similar to those we find in the best enterprise and consumer applications.
4. Shift away from custom solutions to configurable (off-the-shelf) solutions. Much of the AI industry has focused on tools for data cleansing, annotation, labeling and training of machine learning models by AI experts and developers.  In 2022, expect to see more AI vendors focused on moving up the technology stack and providing user-facing solutions on their platforms.  Expect these solutions to provide an opportunity to configure an AI-enabled app to a company’s specific needs, workflow, and contents--without the need of IT resources.
5. Shift away from black box AI results to explainable AI results.  A persistent challenge of AI-enabled solutions has been the inability to explain how AI produced the results it produced or, sometimes, even to reproduce the same results.  Expect AI vendors to offer greater transparency into AI-produced results, including evidence and explanations for results produced by their platform and its algorithms.
6. Shift away from supervised learning models per use case to the centralization of institutional knowledge across multiple subject domains while simultaneously optimizing for individual business processes. Historically, supervised learning models have required training the AI, often with a finite set of content, with a specific vocabulary, and for a specific use case. The time and effort required to do so is often time- and cost-prohibitive.  Expect AI providers to focus on platforms that centralize data/content for use across multiple business processes.  For example, a sales account executive, product manager, and customer support representative might all draw upon centralized intelligence to solve their particular business problems using the same institutional knowledge for different purposes.
7. Shift away from a point solution approach to a suite of AI-enabled business applications built on a common platform. The focus of platform development has been on the platform itself and the tools to build point solutions on the platform.  With the focus shifting to AI-enabled solutions targeted to line of business teams with specific business problems that can capitalize on the same institutional knowledge, expect to see a corresponding shift away from point solutions to suite-oriented solutions, for example, a sales and marketing suite of AI-enabled solutions. More attention will be focused on the platform’s ability to support the suite and a rich user experience versus the platform and tools alone.  (When a company purchases MSFT Office 365 Office, they are far less concerned with the platform and far more concerned about the productivity apps and the common user experience they share across the underlying platform.)
8. Shift away from solutions coded by a team of AI experts to zero code solutions.  Much has been written about the “democratization” of AI and the delivery of platforms that can be used with no additional coding, frankly a benefit more often enjoyed by IT than business users.  A population of users does not simply gain access to technology because no coding is required to use it.  Honestly, it’s less about democratization and more about the “universality” of AI.  Universal AI will be embedded in a suite of configurable business solutions that do not require coding. However, actual democratization occurs when the universe of institutional knowledge of an organization is available to the universe of business users in an organization across a universe of business solutions.  That is Universal AI.
   
   
   
   

‍