New Jersey Artificial Intelligence

Explore the critical principles of AI ethics, including transparency, fairness, and privacy. This webinar examines real-world ethical dilemmas and provides strategies for crafting ethical guidelines through collaboration.

This webinar highlights how to leverage ChatGPT as a personalized assistant. It is a great introduction to the tool for those getting started with generative AI, reviewing effective prompts and using practical examples.

As we mark the first anniversary of the LibraryLinkNJ AI Ambassadors training, we are proud to reflect on the impact this initiative has had on libraries across the state. Starting with a group of 20 dedicated library professionals, the program has helped equip more than 750 individuals with the tools and knowledge to navigate artificial intelligence in meaningful ways.

This year brought many highlights. At an NJLA panel presentation in May, Ambassadors shared practical applications of AI in libraries with 66 attendees, sparking thoughtful discussions on its potential and ethical challenges. In August, two seminars presented in collaboration with the State Library Agency Artificial Intelligence Tools initiative (SLAAIT) further extended the reach of the program. Notably, Mary Martin, ShaBree Henry, and Heather Caldwell trained over 400 PALS Plus staff, demonstrating the program’s ability to scale and adapt to different audiences.

Throughout the year, we’ve seen how interactive workshops, hands-on activities, and ethical conversations resonate deeply with participants, creating an environment for both learning and exploration. These efforts have strengthened libraries’ ability to adapt to evolving technologies while maintaining their core mission of serving communities. The AI Ambassadors program’s success is a testament to the dedication and creativity of its participants. Thank you for helping us build a future where libraries and technology grow together!

discussion with Andy Woodworth and Heather Lubchansky

Focusing on the cooling hype around AI and its implications, including opportunities for education and practical applications. Andy highlighted how the AI Ambassador program could help communities better understand and experiment with the technology during this period. Discussions on watermarking AI-generated content touched on OpenAI’s plans, with Heather questioning its effectiveness against manual circumvention and noting artist tools like Nightshade and Glaze that protect digital works from AI training. Andy emphasized watermarking’s greater relevance in business for authenticating workplace outputs, with academic benefits being secondary. He also noted that educators are shifting away from essay-heavy assessments to alternative, low-tech methods for evaluating student learning.

Read the Full Article

shared by Heather Caldwell

The AI Risk Repository, developed by researchers at MIT, is a comprehensive and continually updated database cataloging over 700 risks associated with artificial intelligence.

Read the Full Article

shared by Andy Woodworth

This article discusses the evolution of generative AI from knowledge-based tools to action-oriented agents capable of executing complex, multistep workflows. These agents can plan actions, utilize online tools, collaborate with other agents and humans, and learn to enhance their performance. This advancement opens new possibilities for businesses, such as virtual assistants planning personalized travel itineraries or programmer agents developing and deploying software features based on natural language descriptions. The article emphasizes that while the technology is still emerging, it holds significant potential to automate complex tasks and drive productivity and innovation across various industries.

Read the Full Article

shared by Mary Martin

The article discusses how AI has revolutionized the study of ancient texts, particularly the Epic of Gilgamesh. Traditionally, scholars painstakingly reconstructed fragmented cuneiform tablets to piece together this ancient Mesopotamian narrative. The introduction of AI has significantly accelerated this process by analyzing and matching fragments more efficiently, leading to new discoveries and a deeper understanding of the text. This technological advancement not only enhances the study of the Epic of Gilgamesh but also holds promise for the broader field of archaeology, offering innovative methods to explore and interpret ancient artifacts.

Read the Full Article

shared by Heather Lubchansky

Researchers at Oxford University have developed a method that identifies when AI outputs are likely incorrect, achieving approximately 79% accuracy—about 10 percentage points higher than previous techniques. This improvement could enhance the reliability of AI systems in various applications, reducing instances where AI tools confidently present false information.

Read the Full Article

shared by Andy Woodworth

Examines the integration of artificial intelligence (AI) into college admissions and advising processes. It highlights the potential for AI to streamline administrative tasks, enhance student engagement, and provide personalized support. However, the article also raises concerns about over-reliance on technology, the risk of depersonalization, and the ethical implications of AI-driven decision-making in higher education.

Read the Full Article

shared by Heather Lubchansky

Explores the phenomenon of "model collapse," where AI systems trained on data generated by other AIs experience a decline in performance. This degradation occurs because AI-generated content often lacks the diversity and quality of human-created data, leading to a feedback loop that diminishes the model's effectiveness over time.

Read the Full Article

shared by Andy Woodworth

Explores the growing importance of artificial intelligence (AI) competencies in higher education and career readiness. It highlights that employers increasingly seek candidates proficient in AI tools and technologies, prompting colleges and universities to integrate AI education into their curricula. However, the article also notes that not all students are eager to adopt AI, with some expressing reluctance or uncertainty about its role in their future careers.

Read the Full Article

shared by Andy Woodworth

Virgin Media O2 has developed "Daisy," an AI-powered virtual grandmother designed to combat phone scammers by engaging them in prolonged, unproductive conversations. Daisy mimics the persona of an elderly woman, discussing topics like family and knitting, thereby wasting scammers' time and reducing their opportunities to defraud real individuals. This initiative not only disrupts scam operations but also raises public awareness about phone fraud. Early tests have shown Daisy keeping scammers on the line for over 40 minutes, highlighting the potential of AI in enhancing consumer protection.

Read the Full Article

shared by Mary Martin

Bernard Marr explores the significance of courteous interactions with AI systems. He highlights that while AI models like ChatGPT lack consciousness, the manner in which we engage with them can influence their responses. Polite prompts often lead to more accurate and helpful outputs, as these models are trained on vast datasets reflecting human communication patterns. Marr also discusses the broader implications of our behavior toward AI, suggesting that respectful interactions can promote positive communication habits in human-to-human exchanges. He emphasizes the importance of maintaining ethical standards and empathy in our dealings with AI, as these practices can shape the development and deployment of future technologies.

Read the Full Article

As AI technologies become more prevalent in work, creative endeavors, and the spread of online misinformation, libraries are positioned to educate communities about ethical AI use and to counteract the negative impacts of AI, such as deepfakes and the erosion of trust in democratic institutions. The article reflects on library programs being implemented to demystify AI, including the LibraryLinkNJ AI Ambassadors!

Read the Full Article

• Generative AI in a Nutshell (Henrik Kniberg)
• 'Game of Thrones' AI Chatbot Allegedly Drove Teen to Suicide (USA Today)
  
• US Government Commission Pushes Manhattan Project-Style AI Initiative (Reuters)
• VFX Artists Expose AI Scams (Corridor Crew)

Join the LibraryLinkNJ AI Ambassadors for a session exploring the practical use of AI tools in academic libraries. Each panelist will introduce a specific AI tool—NotebookLM, ChatGPT, and Microsoft Copilot—along with brief demonstrations. The discussion will focus on how these tools can enhance library services, streamline workflows, and improve engagement with patrons and students.

Hear from Jon Braun, Andy Woodworth, Annie Chatterjie, and Joseph Pascale during the first series of Breakout Sessions on Friday, January 10th. Registration is required to attend.

In this issue, we’re spotlighting LiveChatAI, a powerful tool for creating custom AI chatbots with a localized corpus. This service allows you to define the exact pages, documents, and resources included, making it ideal for library-specific implementations.

With LiveChatAI, you can:

• Point it to your website, create tailored questions, and embed the tool seamlessly.
• Index platforms like LibGuides and LibAnswers using sitemap URLs for enhanced library services.
• Gain granular control over your chatbot's functionality compared to similar tools like Ocelot.

shared by Jonathan Jiras

Did you know that LibraryLinkNJ publishes newsletters on a range of topics? Consider subscribing and take a peek at the latest issue when you click the links below.

• On Our Calendar - Updated Weekly
• LibraryLinkNJ News - Updated Monthly