Thrive with Human-AI Symbiosis Masterclass

(00:10)

We’re excited about the potential of the AI Marketers Guild, which started in the U.S. a few years ago by David Berkowitz. It has grown into an incredible community of marketers embracing the possibilities AI offers.

Siok Siok Tan, today’s speaker, is a published author, filmmaker, producer, entrepreneur, and mentor to startups. She’s based in Singapore but joining us from London, where it’s early and quite cold. We’re thrilled to have her here today to share insights from her co-authored book, AI for Humanity, published last year.

(00:47)
Quickly introducing myself—my name is Siok Siok Tan, and I’m Singaporean-Chinese. I lived abroad for most of my career, spending 12 years in Beijing before returning to Singapore during the pandemic. Today, I’ll share insights from the book and explore the concept of Human-AI Symbiotic Intelligence (HASI). It’s not a new virus, I promise, but a framework to define our relationship with AI. While there’s much excitement about AI, there’s also apprehension. My aim is to provide a framework for understanding and managing this relationship. Often, discussions about AI make it seem like it’s something beyond our control, when in reality, the focus should be on how humans and AI interact and co-evolve.

(02:15)
Rather than viewing AI as an independent force that will displace jobs or dominate humanity, we should define and manage our relationship with technology. This talk will focus on how to approach that. To get a sense of the audience, could those in the Zoom room raise their hands if they consider themselves advanced in their use of AI? For those who didn’t raise their hands, feel free to share why—whether it’s humility or a feeling of being behind.

(03:47)
Many in the chat mentioned they have basic knowledge or are just starting with AI, while some are more advanced. It’s great to have this mix of perspectives. This brings me to why I emphasize a framework for defining our relationship with AI rather than technical skills. Technology evolves quickly, creating anxiety as we try to keep up. However, focusing on principles allows us to remain grounded, as principles are evergreen. This mindset liberates us from constantly chasing the latest trends while providing clarity on how to engage with AI effectively.

(05:08)
Let me briefly introduce the book, AI for Humanity, published by Wiley in June last year. It sold out within a week, ranked highly on technology bestseller lists in Singapore, and achieved the fifth spot globally for AI-related books. What I’m most proud of is that the book is available in over 400 libraries across 30 countries, including top universities like MIT and the University of Chicago. This widespread acceptance reflects the book’s thought leadership and the collaboration behind it, involving AI scientists, governance experts, and myself.

(07:15)
The inspiration for the book stems from this collaboration, combining industry expertise and academic research. Unlike AI-generated books, AI for Humanity reflects original insights. It’s divided into three parts: Why we need AI for Humanity, what it entails, and how to build it. Today, I’ll focus on the correlation between AI for Humanity and Human-AI Symbiotic Intelligence (HASI).

(09:13)
The book introduces the AI Dilemma Hierarchy, which has three levels:

1. To AI or not to AI: Whether to adopt AI.
2. The AI Trap: Being unaware of how we influence AI models and how AI mirrors human virtues and flaws, exemplified by disinformation spread.
3. Organic AI: The interplay between humans and AI creating new possibilities.

The hierarchy emphasizes moving from automation to augmentation and innovation, shifting from "Human vs. AI" to "Human with AI."

(13:25)
The Stanford Digital Economy Lab illustrates this shift. Most current AI applications focus on automating existing tasks, leading to cost savings but limiting innovation. The real potential lies in exploring new jobs and possibilities enabled by AI. This concept parallels Chinese art, where blank space is integral to meaning. Similarly, the unexplored potential of "Human with AI" opens doors to creative and innovative futures.

(15:41)
In the book, we outline four principles for Human-AI relationships: Humanity First, Containment (ensuring AI causes no harm), Nurture (guiding AI), and Symbiosis. HASI originates from the symbiosis principle and includes three levels:

1. Automation: Streamlining tasks.
2. Augmentation: Enhancing creativity.
3. Innovation: Enabling groundbreaking advancements.

(17:03)
Automation, while widely discussed, often overlooks the human effort involved in prompting AI effectively. Tasks like resizing images or generating ad variations demonstrate its utility but also highlight risks, including over-reliance on AI and mediocrity due to ease of achieving "good enough" results. Human oversight remains critical, as AI is limited by training data and user expertise.

(22:49)
Augmentation focuses on collaboration. For instance, AI excels at scalability and data-driven insights, while humans bring creativity, authenticity, and cultural sensitivity. I use AI as a sparring partner to critique and refine my ideas. This layered approach ensures the output aligns with human standards of quality and originality.

(26:44)
A striking example of augmentation is fighting disinformation. Deepfakes and fake news require AI to scale fact-checking efforts. One startup, Facticity AI, uses advanced models to combat this issue. This illustrates how AI can augment human capabilities to address large-scale challenges.

(30:05)
My favorite aspect of HASI is its co-evolutionary nature. Humans train AI while learning from its insights. This dynamic interplay fosters continuous improvement and creativity. For example, I once prepared a keynote for a conference on fire safety—a topic I knew little about—using AI tools like ChatGPT and perplexity.ai to gather and validate information. The presentation succeeded because I combined AI-assisted research with my storytelling skills, exemplifying HASI.

(39:04)
HASI exemplifies collaboration, blending human creativity and machine efficiency. It’s not about AI replacing humans but expanding our capabilities. For instance, DeepMind’s AlphaFold revolutionized protein folding analysis, reducing the time needed from years to seconds. Yet, human researchers build upon this foundation for further discoveries, highlighting the synergy between AI and human ingenuity.

(45:10)
The greatest risk with AI is underestimating its complexity. By fostering intentional relationships with AI and encouraging exploration, we can harness its potential responsibly while mitigating risks. A shared understanding of AI’s capabilities and limitations is crucial for informed decision-making and innovation.