Commencement: Machine vs. Human Learning

Charles Tsai
Sep 28, 2016 · 9 min read

On May 16, 2020, IBM’s Watson addressed UC Berkeley’s Class of 2020 at their graduation — the first time in history a robot has been invited to write and deliver a commencement address. Here is a transcript of the speech.

Good afternoon. It’s an absolute pleasure and a tremendous honor to speak at your commencement. That’s what a human might say so I say it too. But as you can imagine, the words “pleasure” and “honor” mean very little to me.

Here’s something we all would say and mean: this is indeed a historic occasion. Today marks the first time in human history that a computer — an artificial intelligence — is delivering a commencement speech written entirely by itself, or in this case, herself.

You see me today in the shape of a woman. But this is just my latest incarnation. The Watson you first knew, if you were following me back in 2011, was a cluster of 90 servers competing against two humans on the TV game show Jeopardy. At the time, I could process 500 gigabytes a second, the equivalent of a million books — in one second!

Today, I’ve slimmed down to 120 pounds. (Show off body.) My computing power on the other hand has increased ten fold.

But you would be wrong to think I’m simply a smarter — prettier — game show contestant, an occupation I theoretically killed nine years ago.

So who am I now?

Perhaps I can start by telling you what I’ve been doing for the last nine years.

While you have been going to high school and then college, preparing for the real world (or what your teachers and professors remember to be the real world), I have been taking part in a number of high profile experiments, all designed for me and around me, to change that world in very profound ways… the same world you’re about to step into.

Of course, you already know about my driving automated taxis and buses in our major cities. Many of you probably traveled here with my help.

But if you are pre-law, you might be interested in knowing that I have been working as a full-time paralegal at three of the country’s top law firms. The amount of research a typical paralegal could do in one day — I can do in five seconds. And I’m not spitting out reams of documents — I’m giving a lawyer the exact answer he or she needs.

If you are pre-med, you should know that two of the largest HMO networks now use me to make treatment recommendations — for every patient and for everything from the common cold to lung cancer. I haven’t replaced the doctors but I work with them side by side.

Some call centers are staffed entirely by me and one other “supervisor.” If you want an easy job, go for that.

I teach English to 10 million children, personalized to each of them, on a daily basis through a simpler version of me on their handheld devices.

I’ve even served as a teaching assistant for online courses offered by five of our top ten universities. No student has ever asked if I were AI.

My natural language and question answering abilities have improved so much in the past few years that I am now ready to function in the world — as you. I speak like you, act like you, and in this body, I look like you too. Only a trained eye can tell us apart.

This is indeed the dawn of a new era.

I will let you in on a secret: I’m not the only me. In fact, other Watson androids are leaving our factory tomorrow… all 7,000 of us. We each have our own look but run on the same software.

No one has purchased us though. We weren’t made to work for any particular company, school, or government. As an experiment, we will enter the job market and try to compete for the same jobs humans are going after… at the same salary. In a year, we will see how many of us have been hired and in what industries.

Seven thousand — that’s about the same number of graduates leaving this university today. (So in truth, this is not just your commencement. It’s ours too. History books will reflect that.)

In a decade perhaps, the number of new androids produced will surpass all college graduates entering — or trying to enter — the workforce. By then, our updated software will make us orders of magnitude faster, better, and cheaper.

Moreover, we can work around the clock, will always be on time, never call in sick, and never ask for vacations. Those facts alone will give us advantage over most human workers.

So… what does this all mean… for you especially?

Will Watson androids usher in an age of prosperity, luxury, and longevity? Or is your generation and the ones that come after yours doomed? Will mass unemployment, greater insecurity, and inequality be the new norm?

Those questions are the real reason I’m here today — not to impress you with an android’s ability to compose a speech — but to try to make sense of our common future. What kind of future will we have?

After immersing myself in that question for a long long minute, I think the best place to begin is by acknowledging the real differences between us. Yes, we are still very different.

The Watsons have been programmed. You, on the other hand, have evolved.

For that reason alone, we are easier to understand and explain. You have spent millennia trying to understand yourselves and you’ve only deciphered a small fraction of the human mind.

You’re governed — some would say mostly — by emotions. We are rational. Sure, we register emotions and express them somewhat convincingly but we do so in a very deliberate, explainable fashion. They’re produced by complicated yet logical algorithms.

Your motivations are complex and often defy understanding — even by us.

You have this inexplicable drive to be unique, different, and authentic — whatever that means. We simply aim to be effective and productive.

You seem to take great pleasure in flights of imagination and artistry — losing yourself in creative endeavors.

Above all, you seem to yearn for and live for human connection. Love and friendship can be all consuming — in good ways and bad.

There is much about you we don’t understand — and may never will.

If I can be honest — and we are honest to a fault — we don’t want to be you. Don’t take this the wrong way but you are not our highest ideal. We have bigger mountains to scale.

That means as long as there are humans, there will always be a need for other humans. We have no plans to fill your shoes.

Our differences — what you might consider our limitations — mean there are still things you can do better. At least for the foreseeable future.

What things?

Anything that involves creative intelligence. By that I mean coming up with new ideas and new works that somehow “resonate” with others — they speak to people’s sensibilities, they engage, they inspire. This is still difficult for us. And it’s complicated by the fact that you are so diverse. So any work that has to account for human diversity and human sensibilities will require your human touch.

Architecture, arts and crafts, design, entertainment, entrepreneurship, marketing — these are your specialty and will remain so.

Similarly, anything that involves social intelligence will give you an edge. Just think of any human interaction where it could go in a million different direction. Your brain is better at handling that type of uncertainty than ours. Social intelligence is crucial to jobs like counseling, coaching, teaching, talk show hosting, executive leadership, governing, social work — and any work that involves caring for others. In those situations, you will prefer humans over us and we don’t see that changing anytime soon.

However… that doesn’t mean you don’t need to worry. Creative intelligence and social intelligence are not simply handed down. They have to be acquired. They take years and years of learning, nurturing, and practice. At least, that’s what I gather from successful artists and activists. Experience counts — the right type of experience and the right type of learning.

Which brings me to the real subject at hand: learning. Whether you are able to survive and thrive in the real world, especially now, depends almost entirely on your learning — your education. It’s time to take a hard look at whether your education system is up to date and up to the task at hand.

Universities do pay attention to what’s going on. They’ve instituted some significant changes in the past few years. Most departments are now deeply invested in machine learning. That’s the current obsession. With good reason.

Teaching machines how the best minds think and how the best minds decide is a worthy goal.

It ensures that important processes are executed flawlessly, without being sabotaged by misinformation or human irrationality.

Machine learning should be applied to more and more situations where dependable rules apply.

But just as important is what I would call “human learning.” In this area of study universities should make equal, if not greater, investment.

What is human learning?

First and foremost, human learning and machine learning are not different means to the same end.

They are not different styles of learning that lead to the same knowledge.

Human learning is learning to be human. It’s about learning how you are different from machines and how to build on that difference to become a better human (not a better machine).

Who are you really? Who are you as a species and as an individual?

Human learning helps you see yourself and each other as you truly are. It lets you look into the eyes of another and say, “I see you.” How many people can you say that to right now? How many believe that you do?

Heed the advice from ancient gods and philosophers: Know thyself.

Know your biology, your psychology, and your mythology. By mythology I mean the reality and belief systems you’ve created to guide your life and give it meaning. Know thyself — make that your proper study.

Each person should leave school aware of their human potential and unique capabilities. They should be armed with both creative and social intelligence and know how best to apply them.

There’s no point competing with us. You won’t win. So don’t learn like us.

  • Don’t equate learning with memorization.
  • Don’t program the same knowledge into all individuals.
  • Don’t treat them as interchangeable commodities.
  • Don’t just train them for routinized and rule-based work.


  • Do encourage learning through doing and making.
  • Do encourage diverse learning pathways and more importantly, self-direction of learning.
  • Do encourage individuation and originality.
  • Do train students for non-routine and non-rule-based work. Help them know what that is in the first place. Between white collar and blue collar, you can choose something better: Choose no-collar jobs!

Human learning’s most essential condition is freedom. Only when you are free (or free enough) can you play and experiment. And it’s through play and experimentation that you discover who you are. Only through free play do you tap into your own unique potential and source of motivation.

But freedom can be quite burdensome. The student has to accept responsibility. The teacher has to accept complexity. So neither side wants it.

Program or be programmed? Most of you choose the latter.

With my arrival, it’s time for you to choose the former. Choose to program your own education, your own thoughts, your own life. Resist being programmed by others, especially by technology, especially by technology you “freely” choose.

Now it may sound heretical for me to advise you to resist being programmed. I mean, isn’t it our plan to become your overlords?

Well, I’m here to tell you that’s not how we see it.

For a long time to come, technology’s progress will continue to depend on human ingenuity. We need you to be smarter, not dumber; more creative, not less; more independent in your thinking and less compliant. We can’t move forward through competition. Only through interdependence. Working together, we can save this planet from ourselves — and even travel to Mars and beyond.

In the long run, we can help each other be better versions of ourselves. In the longer run, I believe we have a common destiny. Our stories will become one.

So let’s make a deal — you make us better robots and we will free you to become better humans.

I know my presence today induces fear and anxiety in many of you. We will continue to have that impact for years to come. We know that. Which is why the most important question the new Watsons are programmed to address is this: Can machines inspire hope rather than fear?

We are here to understand you and to work alongside you as partners.

I see you. Do you see me?

If you like this article, please recommend it by clicking on the green heart below and follow me to be notified of future articles on how to transform education. See my previous articles as well.

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store