Chapter 17
The Next Chapter: Academic Libraries Enter the Intelligence Age
While in graduate school, I switched from a career in user research to becoming an academic librarian. This change was influenced by the thousands of layoffs155 in the tech industry and my strong belief that artificial intelligence (AI) literacy would become an essential part of the information literacy skills taught by university librarians.
Lacking a technical background in machine learning but aspiring to be a tech educator, I established a new guiding light—a supernova of purpose and possibility—without a clear idea of where it would lead me. After landing my first job as an academic librarian, I started diligently following the latest AI developments156 and keeping my team informed about new trends.157
As I returned from the holiday break, however, I noticed a significant shift had occurred in the conversation and speculation surrounding leading AI labs and wondered what it might mean. The AI race had become more like an AI war,158 and academic libraries would need to navigate this new terrain, wherein generative chatbots evolve into superintelligent agents.159
Let’s review: At the end of 2024, OpenAI160 and Google DeepMind161 unleashed a flood of new products and features. These included ChatGPT’s Canvas, Project, Sora, Search, and an Apple Intelligence integration with the iPhone assistant Siri. Google also released Gemini’s 2.0 models and added Deep Research capabilities, which can write a freshman research paper. It also rolled out Veo 2, Project Astra, Project Mariner, and Notebook LM updates.
OpenAI also announced a new reasoning model called o3, which quickly grabbed news headlines162 and sparked a huge debate.163 Due to o3’s notable performance164 on artificial general intelligence (AGI) benchmarks—tests that attempt to evaluate the capabilities of AI systems—reports emerged165 about the potential achievement of AGI.
No one agrees on how to define AGI, but if you ask ChatGPT, it’s a “theoretical form of AI capable of performing any intellectual task that a human can, with general problem-solving and reasoning abilities.” If you take a look at the recently leaked documents166 outlining the partnership between OpenAI and Microsoft and their definition of AGI, they describe an AI system capable of generating up to $100 billion in profits. It’s like if Clippy167 became Mark Cuban.
The Moonshot Goal
On January 5, 2025, OpenAI’s CEO, Sam Altman, intensified168 the AGI conversation through a blog post,169 stating, “We are now confident we know how to build AGI.” He emphasized that 2025 may see the debut of the first AI agents in the workforce, which was followed by the release170 of Operator on January 23, a research preview of an agentic system that can complete tasks inside a cloud-hosted browser window.
Altman also said that OpenAI is shifting its focus beyond AGI toward developing artificial superintelligence (ASI). (I’ll let you ask ChatGPT to define that one.) These bold claims reminded me of a recent tweet from Logan Kilpatrick, a product leader at Google AI Studio, on X.
He remarked that171 a “straight shot to ASI” appears “more probable by the month.” This viral post172 also referenced Ilya Sutskever, former Chief Research Scientist at OpenAI, who he believes recognized this direct approach’s potential before leaving OpenAI to start173 Safe Superintelligence and raise $1 billion. This is likely why Altman, Sutskever, and Kilpatrick envision an ASI future ahead.
With so many influential figures discussing ASI, it suddenly feels necessary to bring up the concept of technological singularity. This idea, popularized by futurist and inventor Ray Kurzweil in his 2006 book The Singularity Is Near,174 signifies the moment when AI surpasses human intelligence, leading to rapid and profound societal changes. In 2024, Kurzweil published a sequel titled The Singularity Is Nearer,175 which reaffirms many of his previous predictions.
He believes that AGI will emerge by 2029, or at the latest,176 by 2032, with the singularity anticipated around 2045. It’s worth noting that a sci-fi novel by Cory Doctorow and Charles Stross comically described177 the singularity as the “rapture for the nerds.”
To understand why this prediction seems increasingly sober and possibly realistic—rather than the wrong side of TikTok or a Silicon Valley hype cycle gone wild—it’s important to recognize that advancements in AI have far-reaching implications.
At CES, NVIDIA CEO Jensen Huang stated178 that we are nearing a “ChatGPT moment for general robotics.” He said that these rapid breakthroughs would drive agentic systems, self-driving cars, and a humanoid robot revolution,179 ultimately transforming the entire technology industry.
The Sputnik Moment
At this time, I can imagine that you are wondering how we shifted from discussing the recent advancements by OpenAI and Google to exploring the possibilities of AGI or ASI. To connect the dots, I often cite the recent series of statements by former Google CEO and chairman Eric Schmidt, coauthor of The Age of AI180 and Genesis.127
In late 2024, he said that in the tech industry, it’s believed that within the next five years, AI systems will be able to write their own code and continue to self-improve, possibly leading us to a definitive version of AGI. He says that six to eight years from now, which will fall between 2030 and 2032, “it’ll be possible to have a single system that is 80-90% of the abilities of experts in every field.” That’s 80-90% of a Nobel prize-level physicist or chemist.
In a recent CNBC TV interview,181 Anthropic CEO Dario Amondi said182 that in the next two to three years, we can expect “a country of geniuses in a data center,” echoing a manifesto9 he wrote in 2024.
To provide additional context for these spectacular predictions and statements, major tech companies, corporations, and utility providers are projected to invest183 roughly $1 trillion in capital expenditures over the coming years to support AI development. And if that number wasn’t already mind-boggling, President Donald Trump revealed Project Stargate184 on January 21, alongside key figures from OpenAI, SoftBank, and Oracle.
If completed, this investment would be185 “the largest AI infrastructure project, by far, in history.” It involves up to $500 billion over the next four years. The goals are to improve AI infrastructure in the United States, reinforce US leadership in AI technology, create over 100,000 jobs, and help us achieve AGI before China.186
That same week, a Chinese AI lab released a new reasoning model on its chatbot, DeepSeek, which exploded in popularity187 and climbed to the top of Apple’s App charts. A major venture capitalist referred to this as a “Sputnik moment.”188
The Intelligence Age
In his popular AI-focused newsletter, One Useful Thing, educator and Co-Intelligence189 author Ethan Mollick recently wrote,190 “The flood of intelligence that may be coming isn’t inherently good or bad—but how we prepare for it, how we adapt to it, and most importantly, how we choose to use it, will determine whether it becomes a force for progress or disruption.
The time to start having these conversations isn’t after the water starts rising—it’s now.” So, the big question is, where does the academic library fit into this picture, and how might it adapt to the intelligence age?191 How might the daily lives of university librarians change as a result?
No one knows what will happen next. In the short term, hundreds of academic libraries have developed192 AI literacy guides for their institutions. Many have hosted events and workshops that teach important AI concepts, tools, and skills; others have introduced virtual tutorials193 on generative AI.
Many will celebrate National AI Literacy Day194 on March 28 by creating book displays, facilitating panels, showing documentaries, and hosting AI-themed art contests. In the long term, there will need to be efforts to provide equitable access60 to the latest AI tools and services and empower students with the knowledge and skills to use them ethically and effectively.
If students will soon have a polymath in their pocket195 —a Benjamin Franklin-style AI companion, if you will—this tool will expand their access196 to information, help connect ideas across various disciplines, manage important projects and tasks, and impart timeless wisdom and humor. Consequently, academic libraries will transform into campus hubs for innovation and collaboration.
Librarians will take on roles as AI facilitators and educators who guide students on their journeys, helping them navigate their research interests and greater causes. The world has no shortage of wicked problems—those challenging issues without clear solutions, such as climate change, public health crises, or social inequality. Many unanswered questions still exist in fields like science, physics, math, biology, and medicine, to name a few, that will require bright, young minds.
As AI systems continue to advance rapidly, academic libraries must keep working to create equal opportunities and help students gain essential AI literacy skills for this exciting new era. These efforts will promote prosperity in many different places and ensure everyone has access to these intelligent machines and the infinite stream of information they offer. We cannot let the digital divide become a vast chasm in AI access.
Ensuring that people from both urban and rural communities have the skills to navigate the AI landscape is crucial. We must prepare everyone for this new age of AI, empowering students from all walks of life to use the power of intelligent machines to pursue the “new” American Dream.