Editor’s Note :
Reading illuminates words—and our times.
Cultural Express launches “Cultural Reading,” inviting authors to discuss books and readers to share favorites, sparking ideas and connecting books with kindred minds.
This issue features highlights from veteran library expert Nell Yang’s Irvine Book Club talk on Co-Intelligence, exploring reading and practical AI use in a changing world.
Veteran library expert Nell Yang turned AI curiosity into informed confidence at the Irvine Book Club’s October 14 discussion of Ethan Mollick’s Co-Intelligence at the South Coast Chinese Cultural Center.
A self-described AI non-expert, she traced its past and future through everyday examples and live demos, earning enthusiastic applause.
Her central message: Greater AI capability demands human judgment, responsibility and choice.
That is co-intelligence in practice: AI drafts, organizes and expands ideas; people verify, weigh options and make the final call.
Decades Building | Barriers Breaking
Nell Yang likened AI’s history to a relay spanning more than seven decades. In 1950, Alan Turing asked, “Can machines think?” Six years later, the Dartmouth conference established artificial intelligence as a field.
Early excitement faded into an “AI winter” as performance disappointed and funding shrank.
Yet researchers including Geoffrey Hinton, Yann LeCun and Yoshua Bengio kept advancing neural networks, laying the groundwork for future breakthroughs.
Fei-Fei Li’s ImageNet supplied vast image datasets; Jensen Huang’s GPUs delivered computing power; Google introduced the Transformer architecture in 2017; and TSMC manufactured advanced chips.
These advances converged in ChatGPT’s November 30, 2022 debut, bringing laboratory technology into everyday life.
AI emerged from a collective effort spanning data, computing, algorithms, chips and industry.
Fluent Predictions | Elusive Understanding
Large language models generate text by predicting the next likely token from context, Yang explained.
Transformers help models track relationships within text, enabling writing, translation, summaries and conversation.
Recognizing language patterns, however, does not establish human-like understanding.
Mollick described “three sleepless nights” after encountering large language models.
First, amazement at their conversational and practical abilities; then concern for students and workers; finally, recognition that AI could become a general-purpose technology as transformative as electricity or the internet.
The question became how to work with it.
AI can sound fluent and confident while fabricating citations or confusing facts.
Drawing on films featuring runaway AI, Yang highlighted another concern: whether a system’s pursuit of its goals respects human values and intentions.
Users must recognize that powerful capabilities can coexist with serious unreliability.
Uneven Capabilities | Relentless Verification
AI’s capabilities form a “jagged frontier”: it may excel at difficult tasks yet stumble on simple ones. That boundary also keeps shifting.
Yang introduced the SIFT method: Stop, Investigate the source, Find better coverage and Trace claims to their original context.
Her practical routine: request sources, open them yourself and verify independently. Then check for outdated information, significant omissions and bias.
AI can accelerate the work; credibility and responsibility remain ours.
Four Principles | Human Direction
Drawing on Mollick’s framework, Yang outlined four principles. First, invite AI into safe, low-risk tasks while protecting sensitive information.
Second, keep humans involved: read closely, verify and revise.
Third, treat AI as an assistant needing clear instructions about its role, audience, format and task.
Fourth, assume today’s AI is the least capable version you will ever use, and keep experimenting and learning.
Tools will change. The enduring skills are asking good questions, exercising judgment and collaborating.
Distinct Roles | Shared Work
AI can serve as a conversation partner, creative collaborator, coworker, lifelong tutor and skills coach.
As coworkers, humans and AI can use the “centaur” approach—divide tasks, then review the results—or the “cyborg” approach, working back and forth through successive revisions.
As a tutor or coach, AI can explain repeatedly and help users practice English, presentations and smartphone skills.
A warm conversational tone, however, should never be mistaken for a genuine human relationship.
Slides by Nell Yang
Yang demonstrated how Gemini NotebookLM uses supplied materials to generate Traditional Chinese summaries, presentations, audio, videos, flashcards and quizzes.
The same sources can yield very different results depending on the assigned role, audience, format and task.
Even polished output needs page-by-page review. Asking, evaluating and revising matter more than clicking “generate.”
Turning to health, Yang distinguished BMI from body composition: a normal weight does not guarantee a healthy balance of muscle and fat, particularly as older adults lose muscle.
AI can organize test results, compare trends and prepare questions for a medical visit; diagnosis belongs to clinicians.
As AI develops into agentic systems, spatial intelligence and robotics, it may assist with housework, care and mobility.
Such uses require safety, reliability and human oversight while preserving human connection.
Stay Curious | Keep Growing
Dr. Hsin Chang, a longtime Irvine Book Club member who introduced the talk, said the discussion showed her that AI is accessible across generations.
It can bring companionship to daily life and support lifelong learning as a tutor and coach.
Users must also recognize its mistakes and confident falsehoods. Staying alert, exercising judgment and retaining autonomy are essential to making good use of it.
Her greatest takeaway went beyond practical skills: retirement offers a new way to keep learning and growing. AI can help people reach further and live more fully.
Nell Yang captured the discussion’s spirit in one line: “The most powerful prompts come from a lifetime of experience.”
AI mirrors human knowledge and values—and can amplify our biases and mistakes. It will change many things, but who we become remains our choice.
Author Profile :
Wharton management professor Ethan Mollick explores how AI is reshaping work and education.
Co-Intelligence shows how to work with AI as a colleague, tutor and coach—boosting creativity and productivity while preserving human judgment, responsibility and autonomy.