As the Federal government looks to emerging technologies such as artificial intelligence (AI) to help improve citizen services and lower government costs, applications that enable this new computer-based intelligence via application programming interfaces (APIs) might be the least disruptive and inexpensive way for agencies to get started.
The Trump administration’s plans to establish the Select Committee on Artificial Intelligence as part of an overall effort to coordinate Federal AI research and development initiatives while working with industry and academia is a sign the government is getting more serious about the importance of AI.
Some of the issues holding back Federal agency adoption of AI technologies are the high cost of supercomputing resources and the reliance on legacy systems that were not developed with AI applications in mind, Gary Newgaard, public sector vice president with Pure Storage, a provider of data storage technologies, told MeriTalk in an interview in May. Pure Storage has teamed-up with NVIDIA, a leader in the graphics chip space, to develop a comprehensive AI infrastructure to make the technology more accessible to government.
However, as AI moves toward becoming a core component of modern applications, AI-enabling certain enterprise applications might be a route an agency can take to add intelligence at an affordable price point. Amazon Web Services, Google, IBM, and Microsoft as well as a handful of smaller companies offer AI platforms that expose APIs, allowing developers to make existing applications more intelligent. For instance, many of these companies offer a diverse selection of pre-trained services that enable organizations to add computer vision, speech, language analysis, and chatbot functionality.
AI refers to the use of computers that simulate human abilities and perform tasks that people typically have undertaken. Related concepts and names include cognitive computing, predictive analytics, robotic process automation, and machine learning (ML), which refers to systems that learn from data and improve over time.
Amazon AWS AI Services offer some of the same machine learning capabilities and algorithms that power Amazon internal systems such as Echo with Alexa and Amazon.com’s recommendation engine. Developers of all skill levels can deploy visualization tools and wizards that guide them through the process of creating machine learning models without having to learn complex ML algorithms and technology. Once the models are ready, Amazon Machine Learning makes it easy to obtain predictions for applications using simple APIs, without having to implement custom prediction generation code, or manage any infrastructure, according to AWS.
Google Cloud AI Services also offers machine learning services with pre-trained models, and a service that lets organizations tailor their own ML models. Cloud AutoML is a suite of machine learning products that lets developers with limited ML expertise train high quality models by leveraging Google’s transfer learning and Neural Architecture Search technology.
IBM Watson Services offers the ability to build an AI Assistant for a variety of channels, such as mobile devices and messaging platforms as well as embed AI, machine learning, and deep learning into applications to derive better insight from data.
Microsoft Cognitive Services uses machine learning to let developers integrate intelligence into apps, web sites and bots. The suite of tools includes image-processing algorithms, pre-built scripts for processing natural language, algorithms for converting audio to text, advanced search APIs, and knowledge services that let developers map complex information and data to solve tasks such as intelligent recommendations and semantic search.
AI has enormous potential for government, according to an issue brief by the IBM Center for the Business of Government and The Partnership for Public Service, The Future Has Begun: Using Artificial Intelligence To Transform Government. “It can improve agencies’ effectiveness, make data more understandable and easier to use, and help citizens navigate government services.” And it could save government up to 1.2 billion work hours and $41.1 billion annually, the brief states.