New Column, New Columnist
Dr. Hu on Artificial Intelligence & Machine Learning
How do we get to next? Dr. Albert Hu navigates the crosscurrents of artificial intelligence and machine learning (AI/ML)
Tom Green
Demystifying the way forward
Please join us in welcoming Dr. Albert Hu to Asian Robotics Review. The good doctor is our new columnist for a new column that’s popping with fresh ideas and clever insights on our equally fresh new world of artificial intelligence and machine learning (AI/ML).
The blistering pace of AI/ML over the past two years has left many a bit breathless. Suddenly, change is in the air. The experts are saying that AI disruption overall is slowing down considerably. In a lull, so to speak. Big business, it’s been reported, is willing to implement but slow at adoption, while SMEs seem to be totally wary of anything that smacks of AI. The investment is high and ROI is tough to come by.
The timing for Dr. Hu’s new column couldn’t be better.
Like rounding a big, slow curve before taking off again on another AI straightaway, we seem to be collectively stopping to take stock of what’s happened to our world before we hurtle onward once again.
Maybe we’re entering into a period of AI gestation. A time when we’ll assimilate what’s happened to us thus far, or at least begin the assimilation process.
Enter Dr. Hu. Nice opportunity for a knowledgeable and compassionate expert with scads of AI/ML experience—who also just happens to be a wonderful story teller.
He doesn’t have a TARDIS, like that other Dr. Who, but he’s definitely powered by his own internal algorithms that, like the TARDIS, make him larger on the inside than he is on the outside.
About Dr. Hu:
Dr. Albert Hu received his executive management training in growth management and in influence strategy from Stanford’s Graduate School of Business. Dr. Hu’s PhD work at MIT turned out to be machine learning (ML) using Bayesian epistemology framework. That work has impacted the global semiconductor industry, and its IEEE publication is now the most cited article according to Google Scholar in its field in the past 25 years.
Dr. Hu was also the founder, CEO, and board director of a venture capital- and semiconductor industry-funded startup in Silicon Valley. Dr. Hu has observed with great interest that today’s robots converse like a human, bi-pedal like a human, and can beat a human in GO (a game of strategy). What is ahead for us with AI/ML?
Drop by Asian Robotics Review for the good doctor’s take on where we’re headed and why.
Q: You gained a Doctor of Philosophy at MIT in 1992. What, if anything, has surprised you most of all about the importance of Philosophy in relation to the rapidly increasing advancement of AI within society today?
A: At MIT, the degree conferred can be Doctor of Philosophy even if your doctoral dissertation is in the engineering field. This academic policy turns out to be visionary and prescient in view of the rapid advancement of AI and ML. By that, I mean the advancement in AI and ML is such that now we need to re-examine the fundamental questions since the beginning of civilization: existence, knowledge, value, justice, language, perception, reason, and mind; exactly the topics philosophers have been questioning since millennia ago.
From one direction, within AI/ML communities and technology “tool boxes,” philosophy is now part of standard AI/ML vocabulary and algorithms: ontology engineering, epistemic frameworks, natural language, perception, and cognition (both the neurobio- and psycho-social aspects of it).
From the other direction, we now as a society are dealing with the existential threat and moral tension, such as AI-induced biases, introduced by the rapid advancement of AI/ML. These issues are very real, very imminent, and very ancient as well. Therefore, I would like to ask this question, should Philosophy 101 be a compulsory course for all AI/ML majors?
See related:
Back to the Future
Philosophy Grads Sought for Careers in AI!
“Philosophers with dim career prospects are in demand to research the ethics of data tech.” —Financial Times
Q: What do you believe humans will always excel at, regardless of the progress of technology?
A: Philosophy, although analytical epistemology and logical AI may merge into one field. The other is nurturing. Here we humans are in the lead, benefitting from million years of experience in one wet-neural network training another wet one, enhanced by biological evolution. A NN-like algorithm should catch up on training, i.e. nurturing itself without the luxury of Big Data; it should use the least amount of data to train itself in facing an unknown situation or demand.
Q: You volunteer at a local organization in San Francisco called Code Tenderloin. What is their purpose, and what is the most important life lesson you have learnt from volunteering there?
A: Code Tenderloin is a public/private supported community-based organization to help bridge the skill and economic gap between the north and south sides of Market Street. It serves the disfranchised with the resources from leading tech firms in San Francisco and from the San Francisco Mayor’s Office.
AI/ML is an atomic bomb in the economy. It has detonated and the shockwave is now pushing the knowledge divide even wider. If one thinks the internet has created a divide, this new divide can be bigger for not only the blue-collar but also the white-collar workers. A society divided is not just and is not stable. I am doing my share of civic duty. I am learning and observing. I hope I can come up with some useful solutions using the power of AI/ML itself to mitigate the after-shock of AI/ML detonation.
Q: You also have interests and considerable expertise in solar and renewable energy. Where do you believe the most currently untapped potential lies in renewable energy today? Do you believe it can be successfully harnessed in the next 5 years, and will AI play a part in helping to achieve this, in your opinion?
A: Human beings are mammals that have survived with no claws and no thick keratin armor.
We have survived through superior capability to communicate with each other. We have also survived by being agile through expensing more energy per unit weight. The 20th century is when human beings further expanded, following our primordial need, the communication ability with the internet, thus creating this trillion-dollar industry globally.
The 21st century is when human beings further expand, again following our primordial need to have lots of energy in order to move fast, horizontally and vertically through electric vehicles and human drones, for the ubiquitous and clean sources of energy; such solar photovoltaic (PV) energy.
Again, this is another trillion-dollar industry, and it is in its early stage.
The North American Power Grid, inter-connected from the East Coast to the West Coast, South to Mexico, and North to Canada, is perhaps the very largest machinery in human history.
The development and maintenance of it poses extremely complex technological challenges. Companies are already using AI/ML on the power grid.
Q: What ‘one’ fundamental action should any business take today, in order to better prepare itself for successful AI integration tomorrow?
A: At the minimal, a business today should acquire the service from a truly experienced tier-one AI/ML professional to guide them in the killing field of AI/ML-enhanced commerce. AI/ML, as applied to businesses today, is very much like parenting. Good parenting needs experienced tier-one AI/ML professional.
The difference between tier-one and tier-two, between experienced and semi-experienced AI/ML talents is astounding.
An MIT alum recently received venture capital funding in Silicon Valley. Her biggest challenge in the past 6 months? She could find money, but she has difficulty finding top ML talent.
Q: Albert, as a trailblazer and pioneer within the Machine Learning field, how would you explain to a group of primary school children, how Machine Learning will radically benefit their lives, in years to come?
A: AI/ML has so dramatically changed the world they live in, and will change even more so in their lifetime. AI/ML, through robotic and automatic factories and farms, produce for them. AI/ML transports them, through Transportation-as-a-service (TaaS) and through autonomous vehicles and human drones. AI/ML helps them learn better and faster. AI/ML provides good medical diagnosis and care. AI/ML also makes them better philosophers to explore the fundamental questions we have asked ourselves since the dawn of civilization: our existence, our knowledge, what is just among us; if and only if they learn to understand the science and the engineering that had led to the advancement of AI/ML.
Excerpted Q&A reproduced courtesy of Chris Baillie & The Artificial Intelligence Group on LinkedIn