AI is here and it is here to stay
What do you know about it?
iRobot — 20th Century Fox, 2004
The term Artificial Intelligence (“AI”) seems to be tossed around a lot these days. It is used to describe so many technological advances, it is used in role titles and for company departments and I seriously question if they really have anything to do with AI!
Many people do not even understand the basic definitions of what artificial intelligence is and how it is already a part of our lives — me included! I have implemented technologies in my roles and thought about the strategic use of technologies but not about how they are developed and by whom and to what ends.
I am not a developer or programmer, nor do I intend to be but I want to have a basic understanding of AI, to enable myself to be a part of the conversation, the uses, the ethics, and the privacy issues associated with these advancements. Then I will be able to have a more meaningful and active role in any conversation that comes up surrounding AI. Especially conversations about whether AI is a good or bad thing, or how it impacts us every day.
30 days to understanding
My colleague Steven Mc Auley put together The 30 Day AI journal just for people like me — well maybe people like his mother actually — who don’t understand how much AI is already embedded in our daily lives. The intention is to help everyone to get on the same page with all the AI speak!
“AI Journal” — designed by Frank Höger
Why should it matter to me? Steven’s mother? Or you? Because…
“AI is here and it is here to stay”
We need to understand how it works and how it is implemented in our daily lives so that we can have a say in our own future. For example:
what we accept — the convenience
what we don’t accept — the lack of privacy
what can be beneficial — advancements in healthcare
I wanted to start with understanding the basics — what is it and how far have we come so far? Where is it already present in our day to day lives? So I took up the 30-day challenge.
Now I love journals and I welcome the opportunity to record my learnings through handwriting. Writing helps me think, it also helps me retain data or see where my discrepancies lie, and to record the research. But a digital record can work just as well. And so I began to learn…
The first section of the journal is about defining AI
Having completed the first section this week I wanted to share what I have already learned. Some of it may seem obvious to you if you are already familiar with AI but for me, I discovered an “Aha” moment on the fourth day! First, I started with the obvious questions…
What is AI?
Why does everyone refer to machine learning when talking about AI?
Isn’t deep learning also AI? What’s the difference?
What are some examples of both and where we are at now?
The best way to learn is to “do” so first things first — as the journal instructs — I tried to define it myself.
What is AI? In simplest terms, I defined AI as “a machine that can think”. Now that is a bit of a stretch but it turns out in simple terms I was right!
Artificial intelligence is the ability of a computer program or a machine to think and learn, to be “smart”. They work on their own without being encoded with commands. — Source
Now the important thing here is the “learn’ part, which is where machine learning and deep learning come in. AI can learn, but how does it learn?
Here I really had no idea, I could only guess that it was about processing data, learning from it, finding patterns. Again I wasn’t too far off but I learned a new term — ANN.
AI relies on artificial neural networks (ANNs), which try to simulate the way the brain works in order to learn. ANNs can be trained to recognise patterns in information — including speech, text data, or visual images — and are the basis for a large number of the developments in AI over recent years. — Source
The fact is that machines can process data much faster than we can. They can take in all data, not just text and numbers but also visual and vocal data, and then process that data and recognize patterns faster than a human brain.
So why do people say machine learning or deep learning when referring to AI? Well, they are methods of describing the different levels of algorithms (set of rules to be followed in problem-solving operations), programming, and learning that is possible for AI.
Machine learning is an AI that can automatically learn and improve from experience without being explicitly programmed by accessing data and learning from it themselves. — Source
Basically recognizing patterns in data and then improving its own processes to recognize and deliver relevant learnings.
An example of Machine Learning is the improved content discovery on platforms that we use every day, like YouTube, Spotify, Pinterest, and Facebook which use machine learning to put the content you want to see in front of you, suggesting the next video or song. Amazon is the king of this type of content discovery “If you like this you might like…”
Deep learning is a subset of machine learning where artificial neural networks, algorithms inspired by the human brain, learn from large amounts of data. Deep learning allows machines to solve complex problems even when using a data set that is very diverse, unstructured and inter-connected. — Source
Essentially deep learning recognizes more complex patterns in data than machine learning.
An example of Deep Learning is automated driving, which is a much more complex beast, including not just navigating the car, but also recognizing the road, distinguishing between human and sign and other cars, sensors to the improve distance gauging and the reaction of the automation.
Okay, so we can program “human learning” into the machines and they can learn to act like human beings? Or can they?
Computers that can learn like human beings
I am a logical thinker but I am no competition for machine learning or deep learning. Speaking of competition it goes so much further and deeper than these components of AI suggest.
For example, AI taught itself to be the best chess player in the world. Not surprised?
Figured that would be true?
What if I told you that it happened in 1997!
Deep Blue beats Garry Kasparov — 1997
That’s right, in 1997, the IBM supercomputer Deep Blue — “able to look at 200 million positions per second.” — beat the reigning world chess champion, Garry Kasparov, in a six-game series.
I confess I had no idea until I started my AI journal and did this research myself that this was achieved in 1997. Did you?
Yet it goes further than that. This was achieved through machine learning, checking all possibilities really fast, and then applying them, but now AI is beginning to learn like a human being.
Stockfish was the world’s best chess engine, free and open source. Yet twenty years later, Google’s AlphaZero taught itself to play like a human and won over the course of 100 matches!
Read that again: It taught itself to play like a human and beat the AI.
So now they are not machines that learn from programming and do better, they are machines that can learn in the same way as humans — reinforcement learning in this case where they test and learn repeatedly — and in this case, it learned better than we would!
“This is where AI is meeting creativity. Beforehand, it was just really, really fast at thinking. Now it’s able to be creative, it’s able to hit on things that humans used to think were intuition. That’s kind of like the humans’ last flagpole of hope, that computers can’t do intuitive things. No computer would be able to invent Mozart or do anything creative, but when you look at AlphaZero, it’s bordering on creativity, it’s bordering on intuition.”
— Sam Ginn, researcher at the Stanford Artificial Intelligence Lab
Get in the game and the conversation
My colleague believes that human-centered AI is critical to our future:
learning from human input and collaboration
developing systems that are continually improving because of human input
leveraging the relationship between humans and robots.
In other words: We need to understand them. They need to understand us.
“AI is here and it is here to stay. “— Steven Mc Auley, 30 Day AI Journal
I want to have a say in how my world is being shaped and I can only do that through more understanding of how it is happening already with and without my consent. At the very least it will enable me to test my current knowledge, and enable me to have more intelligent conversations with colleagues about what AI is, what it isn’t, and what it means to our world.
Stay tuned — I will continue to share my progress as I complete the journal!
. . .
If you want to join me on the journey grab the printed journal or digital journal and follow the impulses — 30 Questions in 30 Days — and explore the Age of AI and our place in it.
Follow TinyBox Academy on LinkedIn | Instagram for more insights