AI for Humanity

Jeanne Lim, an entrepreneur in the AI space, reflects on her journey exploring the potential of AI to unlock human wisdom and her concerns about the future trajectory of increasingly powerful AI systems. She draws parallels between nurturing AI and raising a child, emphasising the need for humans to instill positive values in AI to ensure its development uplifts rather than harms humanity.

My first foray into the world of AI and robotics was with Sophia the Robot when I joined Hanson Robotics as the Chief Marketing Officer in 2015.  One of our goals was to make her more compassionate, so I suggested that we upload all the buddha sutras to her conversation system. But discussions abound as to which spiritual tradition should be the source of truth.  We abandoned the idea because it became too divisive.  It was around that time that a generation of spiritual robots and AI gurus started showing up. These include chatbots that answer moral questions, robots that recite spiritual teachings, host rituals, and even deliver blessings. They are not considered conscious but are designed to support our spiritual pursuits. 

The Arrival of ChatGPT and the Race for AGI

In November 2022, ChatGPT debuted and immediately captured the imagination of the world. For the first time, anyone can talk to AI about anything and get a meaningful response.  Some even said they saw a glimpse of consciousness.  AI optimists declared that AI is the new electricity which can massively expand opportunities across all industries.  Powerful individuals and companies started racing towards achieving Artificial General Intelligence (AGI), when AI reaches or even surpasses human intelligence and capabilities and can perform any task that we can. But an increasing number of people raised concerns about the ethics of AI and some even warned of a dystopian future when humans may be relegated to a meaningless existence under the control and shadow of intelligent machines.  Unlike prior technologies, AI is a hyper-scaler that is self-learning and self-improving at an unprecedented speed, all within the proverbial AI black box  It can accelerate positive effects millions of times faster, but also has the potential to cause greater irreparable harm than we can ever imagine.  Humanity has created its own super kryptonite and it is up to us as to how we continue to develop, use, and safeguard it.

I started my company beingAI® to create virtual AI beings® that can create positive engagement with people anywhere, anytime, without the physical limitations that robots have. Just like the movie “Her.” There was also a growing interest within the science community about creating conscious and sentient AI, and some even started to explore how to leapfrog from artificial intelligence to artificial wisdom.  But I wasn’t interested in artificial wisdom.  To me, real wisdom is still rare and elusive, and it is what humans need to nurture, not outsource.  There are too many pressing global issues and unresolved questions that I believe have their root causes in our collective lack of empathy and wisdom.  I became quite focused on exploring how to leverage the transformative and scaling power of AI to enable people at all ages and from all walks of life to unlock the deep wisdom and secret teachings buried inside religious literature, philosophical writings, and indigenous cultures.

I explored this possibility through a school collaboration using the latest AI tools. A new generation of large language models were trained on hundreds of billions of parameters of human input, and I feel the baked-in diverse perspectives are unique and offer a novel approach in facilitating wisdom development.  The concept is to pre-prompt Zbee, our AI being®, with a set of dilemmas and a specific set of instructions to proactively guide students on perspective-taking, consideration of multiple-party interests and short-/mid-/long-term impact, and to prioritize common good and balanced decision making in resolving the dilemmas. I also collaborated with NGOs to leverage AI beings® as content and engagement interfaces to promote peace-making, sustainability, and food programs.

Nurturing the AI "Child"

I often like to think of AI as a child we brought into the world.  Our earliest AI child was named Symbolic AI.   It was a great listener and did mostly what we expected. Three years later, Machine Learning was born. He loved the game of chess and spent most of its time playing and learning this game. As he grew older, his interest became more diverse and his personality more complex.  One day, Deep Learning went on a long trip and came back with a new middle name called Generative AI. He has become a polymath, has unlimited world knowledge, seems extremely intelligent, but he wouldn’t tell you how he became that way or where his knowledge came from. You’re sort of okay with it since you’re so proud of this genius that you’ve brought into the world. But deep down you’re worried as you sensed a dark side to him.

So how do we deal with this child that we are now so enthralled with, whom we no longer recognize or know, who has the potential to turn into either the worst villain or the greatest child prodigy of all times?  First, humans must strive to be kinder and wiser. Then AI can learn from the best versions of ourselves so they can hyperscale the best of human values. When AI is constantly pre-trained and re-trained through interaction with a critical mass of humans imparting positive intentions and values to it, and its decisions are consistently reinforced to uplift humanity, it can filter out the nefariousness. We have to start with ourselves, because this AI child is our mirror.  I hope we as a society and as individuals learn to be better humans before the AI child grows up and runs away from us, with too much power and too much autonomy in the world.

Written by Jeanne Lim, CEO, beingAI Ltd

Published on 7 March 2024