Soon, a technology neophyte will be able to build incredible AI use cases by simply speaking to a cognitive agent. The AI industry will move toward a more democratized approach to who can access and implement cognitive solutions.
This blog is part of our ongoing series, IPsoft’s 2019 AI Trends, detailing what we believe will be the dominant developments and movements in the Enterprise AI market next year. These blogs will be published regularly through the end of the year.
Not long ago, building an interactive website required deep technical know-how and programming skills. Building the site took a long time, and if you wanted it to be responsive (mobile-friendly), you had to either build multiple sites or have developers code around each element.
Starting in 2019, the same evolution toward democratization will begin in earnest within the Artificial Intelligence (AI) and Machine Learning community. AI-focused technology companies (including IPsoft) are creating intuitive development environments that will allow citizen developers to create amazing AI use cases by simply speaking to a cognitive interface. You will not need to understand Natural Language Processing, machine learning classifiers, Long Short Term Memory (LSTM) models or semantic memory, or any other complex mathematical, scientific or technological principles. You will simply tell the AI solution that you want it to help guide customers through service requests, or help shoppers find the right pair of jeans, and the tool itself will teach you how to make the use case work.
As a citizen developer, you will be able to create simple AI use cases in hours — the same use cases that only a few years ago might have taken a team of developers and data scientists several months to create. You will only need a deep understanding of how your business runs and how you would like the AI to interact with your back-end systems.
Do-it-yourself AI solutions won’t completely alleviate the need for people with development and data science backgrounds. Those experts will remain crucial to integrate AI with back-end systems to access customer and product data. Although AI will begin to be delivered in simple formats, most legacy systems will remain necessary, which means you will need professionals who understand how to connect AI with legacy systems to provide the best possible experience.
In theory, as a citizen developer, you could use Robotic Process Automation (RPA) to complete the journey, but you might need API integrations in order for the process to be efficient. In that case, you would simply explain to your technical experts that you need AI to be able to do specific things in the system, and the experts will build the connections that allow AI to push and pull information as needed. You can step away, grab a coffee, and allow the experts to fill in the gaps where technological innovation remains a meritocracy.
The critical business aspects of the actual AI application (how it works, how the conversations sound, how customers interact with the interface) will be created and adjusted by subject matter experts — people with little-to-no technical know-how whatsoever. Maintaining the system will also fall squarely on the shoulders of the SMEs. Rather than sending instructions to technical experts who do not interact with customers or understand how the business runs, you and your business experts will be able to work directly with AI to make it smarter, or make processes simpler, all through the power of conversation.
The democratization of AI will not be fully realized in 2019. For example: You will not be able to completely use a conversational agent to build a fleet of customer-facing digital colleagues who can execute processes with expertise…yet. You will, however, be able to focus on specific use cases and flesh this process out while your technology experts look over your shoulder to make sure the process is functioning properly. It won’t be as quick as building a website, and it won’t be as simple as speaking your desires to a virtual developer. However, 2019 will be the year we take our first substantial step toward a future in which AI can mostly be created by and for people with no technical background.