Introduction to AI
Part 1 of a series on AI for public library staff
Interests and concerns about how artificial intelligence (AI) will impact librarianship and the services that libraries provide are increasing. The field is vast, ever-changing, and, at times, overwhelming. This article was shaped by the results of a recent poll of library staff, and explores different aspects of AI and the intersections with public libraries including some basics of AI, along with ethical, legal, and environmental considerations. Future installments in this series will explore how public library staff use AI, how libraries are helping patrons use and learn about AI, and a look ahead at the potential future of AI in libraries.
What is AI?
Columbia University Engineering uses this definition: “Artificial Intelligence is the field of developing computers and robots that are capable of behaving in ways that both mimic and go beyond human capabilities. AI-enabled programs can analyze and contextualize data to provide information or automatically trigger actions without human interference.” AI is found in many of the technologies we use and can range from autocorrect to cars that assist us in driving. It also includes online shopping sites that recommend products, chatbots that help us navigate a complex website, robot vacuum cleaners, and of course personal assistants like Siri. AI has penetrated daily life in the modern world.
What is generative AI?
According to Coursera, “Generative AI, also referred to as GenAI, allows users to input a variety of prompts to generate new content, such as text, images, videos, sounds, code, 3D designs, and other media. It ‘learns’ and is trained on documents and artifacts that already exist online. Generative AI evolves as it continues to train on more data. It operates on AI models and algorithms that are trained on large unlabeled data sets, which require complex math and lots of computing power to create. These data sets train the AI to predict outcomes in the same ways humans might act or create on their own.” Popular generative AI platforms include ChatGPT, Gemini, Claude, CoPilot, and image-generating products like Dall-E, Adobe Firefly, and Midjourney.
While generative AI models have taken the world by storm in the past few years, the concept of artificial intelligence is not a new one. Ancient philosophers and mathematicians theorized about technology that would become a reality centuries later. In the modern era, mathematician and computer scientist Alan Turing gained fame for his work with machine intelligence, and even correctly predicted that computers could one day defeat a chess grandmaster.
To learn more:
If you are interested in the technology and history of ChatGPT, check out the University of Arizona’s excellent tutorial.
Boston University Library’s LibGuide, "Comparing Generative AI Tools" is a starting point to look at the landscape.
The AI Pedagogy Project at Harvard developed a guide as an introduction to the current AI tools that are relevent to use in educational settings.
What are some of the benefits of AI?
AI applications have the potential to influence almost every aspect of our lives and already span multiple domains, from healthcare to business, from technology to the arts. Each application leverages AI's core capabilities: pattern recognition, rapid data processing, and adaptive learning. For example, in healthcare, applications assist in pattern recognition, which allows physicians to diagnose disease. In business, AI can help spot fraudulent emails, which prevent security breaches, saving companies and customers money and time. AI is also powering smart-assistive technology including tools such as Apple’s Siri or the writing software Grammarly, helping people communicate clearly and effectively. In science, AI is developing climate models; the next article in this series will explore specific examples of how public library staff are using AI to assist in their jobs, including using AI as a writing partner to improve accessibility, provide user analytics, enhance catalog records, and generate marketing materials.
What are some of the ethical concerns about AI?
The ethical questions raised by AI are significant and must be weighed against the benefits. For example, how does using AI-generated art to create captivating blog posts about your library events impact jobs and artists?
A few responses to our recent poll raise pointed comments and questions that the library field must consider, such as:
- What are ethical ways to approach AI? I personally think there are very few applications that we can use it in an ethical way that don't harm the environment, authors, and creators/copyright law, and our profession.
- How do we protect the privacy of our users?
- Where is the line between AI-generated content and plagiarism? How do I teach the public about that line?
- Given the devastating environmental impacts AI use incurs that push us closer to irreversible climate collapse, why should public libraries use AI?
- AI is antithetical to our professional code of ethics; generative AI is biased against already vulnerable groups and inaccurate; it is compromising personal data, and it infringes on copyright to name a few.
The University of Texas at Austin’s LibGuide addresses some of the concerns about AI, including the bias of its output, the privacy lost in the harvesting of personal data, the environmental impacts of the energy needs of this technology, and issues regarding unpaid labor, underpaid labor, and author consent.
The environmental impact of AI, both its development and its use, are significant and not fully understood as this technology evolves. According to a Forbes article in May 2024, Chat GPT-4 required over 50 gigawatt hours to train, approximately 0.02% of the electricity California generates in a year, and 50 times the amount it took to train GPT-3. The massive computing power required to run these models and the water needed to cool the servers is significant. Researchers at University of California Riverside have calculated that running some 20 to 50 queries using ChatGPT will use roughly a half liter of fresh water.
Sources to explore the environmental impacts include:
- Scientific American, 2 December 2023, "AI’s Climate Impact Goes Beyond Its Emissions"
- LibGuide from UC Irvine Libraries, "Generative AI and Information Literacy: Environmental impacts of generative AI"
AI tools can also produce hallucinations which IBM defines as “phenomen[a] wherein a large language model (LLM)…perceives patterns or objects that are nonexistent or imperceptible to human observers, creating outputs that are nonsensical or altogether inaccurate.” This IBM article discusses some high-profile hallucinations, where they come from, how to prevent them as well as some implications. On a related note, deepfakes are videos, pictures or audio clips made by AI to look real. Deepfakes can mimic a person’s voice and facial features using audio recording to make it say things the person has never said using a picture of their face or other videos of them. This can be a particularly effective way to spread misinformation, mislead people, influence voting behavior, etc.
AI is forcing some tricky conversations about the boundaries of intellectual property; Educause has a good 5-minute video describing the legal and ethical complexities around data ingestion for AI training, the potential for AI-generated outputs to infringe on existing works, and the challenge of securing copyright protection for AI-created content. Institutions are currently grappling with these issues in the court system, and libraries be watching these issues closely.
For a deep dive into the risks of AI, take a look at this AI Risk Repository categorized by discrimination and toxicity, privacy and security, misinformation, malicious actors and misuse, human-computer interaction, socioeconomic and environmental harms, AI system safety, failures, and limitation.
Up next
As AI continues to take shape and unfold, the public library community will have opportunities to learn and use new tools, enhance and improve workflows, and also grapple with related concerns. The next article in this series will explore how library staff are using AI to assist in their jobs duties. You can also register and join us for a webinar on May 1 which will feature panelists discussing AI and public libraries.