Author: Scott McNea, Alliant Public Entity
Artificial intelligence, once confined to the realms of science fiction as a futuristic improbability, has rapidly evolved into an integral part of our everyday lives.
Despite its immense utility, the risks to society and humanity, including privacy and surveillance, bias and discrimination, and perhaps the deepest, most difficult philosophical question of the era, the role of human judgment, will continue to be debated.
Some fully believe AI will unlock the full potential of the human capability, assisting humankind with a new level of thought and forward-thinking, and forever alter how we engage with machines around us. And there are some who fervently believe that society should be trepid about how we continue to innovate with these advancements in technology.
AI is designed to simulate the intelligence of human beings within various machines, which then allows them to perform tasks that would otherwise require a human. These tasks include learning, reasoning, problem-solving, perception, understanding natural language and interacting with its environment.
But not all types of AI are human-like—in fact, many of the most powerful systems are very different from humans. There are generally two ways to differentiate it: What it does and how it does it. This article will focus on several types of artificial intelligence that we currently work with as well as its future in the Public Sector.
Generative AI
The most common engagements that we have with artificial intelligence on a daily basis is considered generative AI. These are programs and applications that can create various outputs based on the request and input from the user. This generated content can be images, messages, emails, natural language files, audio files, as well as lists and various other content that can be “generated” by AI. These generative systems can be either standalone or embedded. Standalone AI such as Chat GPT or Google Gemini, where the user goes directly to the system for content, by providing data and making a request. Embedded refers to AI that exists within the program, often working behind the scenes. One of the largest challenges with the utilization of any generative AI, be it stand alone or embedded, is that it is completely dependent upon the information given, per the adage, “Garbage in. Garbage out.” The results of these artificial intelligence sources cannot be relied on with 100% certainty unless the initiating information is trusted and verified.
Discriminative AI
Discriminative AI focuses on classification tasks, distinguishing between different categories within a dataset. Discriminative models, such as Support Vector Machines (SVM) and logistic regression, excel in identifying patterns and making precise predictions based on input features. It works by analyzing past behaviors, such as the social media posts you've liked or the pages you've followed, to figure out what kinds of products or services you might be interested in. Thin of this type of AI as a helpful friend who remembers your preferences and recommends things you might like, such as specific television shows and movies. In marketing, companies use discriminative AI to target specific groups of people with personalized ads, making their advertising efforts more effective by reaching a captive audience.
Abstractive AI
Abstracted AI refers to the generation of human-like summaries, speech or interpretations. Abstractive AI represents a cutting-edge approach to natural language processing, revolutionizing the way we generate summaries and comprehend complex textual data. Unlike extractive methods that merely condense existing information, abstractive AI goes a step further by understanding the context and generating new, concise summaries that capture the essence of the original content. This sophisticated technology leverages advanced neural network architectures and deep learning techniques to analyze and interpret language with human-like proficiency. By synthesizing information from various sources and generating coherent summaries, abstractive AI enables us to distill vast amounts of textual data into digestible insights, facilitating decision-making, research and communication at an unprecedented level of efficiency and accuracy. As researchers continue to refine and expand the capabilities of abstractive AI, its potential to transform numerous industries - from journalism to academia to business - remains vast and promising.
Examples of technologies that use abstractive AI are Siri, Google Assistant and Alexa, as well as language translation and summarization tools such as ChatGPT. These technologies understand natural language queries and generate responses from various questions, problems and criteria.
Although an extremely powerful tool, these systems typically operate within a limited scope, and rely greatly on human guidance and input. Abstracted AI lacks the ability to adapt to new situations or learn from experiences and mistakes; it requires human intervention to assess changes and provide appropriate direction.
Autonomous AI
As opposed to Abstracted AI, which works to generate content based on specific criteria given by the user at each interaction, Autonomous AI can operate independently by making decisions void of human intervention. This type of AI is self-learning and adaptable, and found in self-driving vehicles, autonomous robots, games and trading systems. Autonomous AI is designed to perceive its environment, process the information, and make decisions based on obstacles and challenges that would help or inhibit it from accomplishing a task.
Artificial Intelligence in the Public Sector
The development of AI will continue to improve and accelerate its utilization across a wide array of professional spheres, and public entity is no exception. With proper understanding and implementation, artificial intelligence can be adequately utilized to improve the quality of public entity organizations and the lives of the very citizens they serve.
For instance, AI-powered predictive models can help city governments predict traffic congestion patterns, optimize public transportation routes and allocate resources effectively during emergencies. Additionally, AI-driven chatbots and virtual assistants are being deployed by public entities to automate customer service interactions, providing citizens with instant access to information and assistance while reducing the burden on human support staff. Predictive technologies can assist with city planning processes, saving costs through material procurement, outlining work scopes and schedules, and identifying potential problems years and even decades ahead.
Many large cities are investing in AI technologies in hopes that these systems can play a crucial role in enhancing public safety and security measures. Imagine programs that can talk to anyone regardless of the language they speak and be able to identify distress or inflection in their voice. Law enforcement agencies are starting to utilize AI-powered systems for video surveillance, facial recognition and anomaly detection to identify and respond to potential threats more efficiently. AI algorithms can analyze surveillance footage in real-time, flagging suspicious activities or individuals for further investigation. Moreover, public health agencies leverage AI for disease surveillance, outbreak prediction and epidemiological modeling, enabling proactive measures to contain infectious diseases and safeguard public health.
These are just some of the thousands of ways that artificial intelligence will begin to be utilized within public entity. Fei Fei Li, co-director of the Stanford Institute for Human-Centered Artificial Intelligence, stated, “Artificial Intelligence is not a substitute for human intelligence; it is a tool to amplify human creativity and ingenuity.”
While the adoption of AI technologies in the public sector holds promise for improving efficiency and effectiveness, it also raises important considerations around privacy, transparency, equity and cybersecurity, necessitating careful oversight and ethical guidelines to ensure responsible deployment and equitable outcomes for all citizens.
Before implementing new technology, business leaders should carefully consider the business objectives and associated risks involved. Alliant utilizes our extensive industry-specific expertise and an innovative approach to risk quantification to help clients navigate the ever-evolving risk landscape and make informed business decisions when leveraging AI technologies, implementing best practices and insurance coverage solutions to manage AI risk.
For more information, visit www.Alliant.com/Cyber