Computers that learn in unsupervised ways and don’t need to be told what to do in advance is the next phase of computer science, according to innovator and engineer Steve Wozniak.
The Apple co-founder said at the World Business Forum in Sydney this week that this kind of artificial intelligence is important to the future of computing, and we’ve spent years just pre-programming and giving instructions to computers to carry out certain functions or tasks.
“Computers have never really theoretically in the past … programmed themselves. And how do we program ourselves? What is the term for that? It is learning,” he said.
“When you learn, you program yourself, even when you were a baby just learning how to hold a spoon or something. You are programming yourself and computers will get to that stage.
“We are getting closer and closer where computers seem to be doing the things that we call thinking.”
Wozniak said although advances in computer power and speed over the years have been impressive, just having computers do a trillion or even quadrillion things per second is not coming up with a method for solving a problem on its own.
“The human brain has an intuition; we organise a problem and say, ‘here is a method to go about that to get it solved.’
“Then if we say that computers can really think for themselves and program themselves, they can make themselves better faster than we can. They can come up with solutions to the real world faster than we can.”
Wozniak’s concerns about AI’s potential implications on humanity in the far future are shared by other technology leaders such as Stephen Hawking, Elon Musk and Bill Gates.
“Once machines have independent thought and can program themselves, that’s the turning point.
I am hoping the computers only reach our level, so we never have to say they are smarter and better than us.”
Innovative thinking comes from clearing your head
Wozniak looked back at his days working for Hewlett Packard as an engineer where he learnt how to forget what he spent most of his days ingrained in and start with a clear mind to find new approaches to problems.
“Pretend you don’t know what you know about your business. If you want to start up the business fresh, from scratch, every bit of it from retailers to product descriptions, how would it work in the most beautiful way?” he asked.
“It’s not how would we modify what we have today a little to get there, but is there another alternative approach?”
He gave the example learning to use the Texas Instruments calculator when it first came out, a new technology he wasn’t adjusted to. Wozniak spent day after day using the HP calculators that were based on Reverse Polish Notation (RPN), where numbers where entered or layered on like a stack and then were added, subtracted, times or divided by at the end.
This meant equations weren’t entered into the calculator in sequential order, like they are written on page from left to right. Wozniak couldn’t quite work out that the Texas calculator operated this way at first, having operated in such a different way for a long time.
“We laughed, [and thought] it was a toy. We had a big equation on a card, so big … you could solve it on a HP calculator with Reverse Polish Notation. And I took this little, toy calculator and … I started looking at the equation.
“I realised I will never be able to back out and keep track of where I am with this huge, long equation that takes 10 minutes to calculate. But I thought the Texas calculator must be able to solve the equation otherwise they wouldn’t sell it, so what could the method be? I wiped my slate clean and I thought, ‘Oh my god, I’ll just type it in the way it’s written from left to right.’”
He then tried to explain to the other HP engineers how to compute the equation using the new calculator, but “not one of those engineers could do it”.
“They had learned a complicated way of using their brain to assemble the equations, and they couldn’t give up that way of doing it.”
Remember when we were once poor?
Wozniak reflected on his days where he and Steve Jobs were slumming it as students, without any money to fulfill their dreams. This forced them to do some lateral thinking about how to build products cost-efficiently.
“Two of the most important aspects of all the great creations that I was ever responsible for were because I didn’t have money and had to figure out how to build things with the fewest parts and the lowest cost and to make things possible that weren’t because I didn’t have money,” Wozniak said.
He said Steve Jobs managed to find a few hundred dollars here and there, just enough for them to build electronics and computers and bring their ideas to life.
While working in Sunnyvale, California on programming computers during his year off from study to help save money for his education, Wozniak was offered an opportunity to build a computer for the company.
“One of the executives heard that I used to design computers. He asked if I had ever built one. I said, ‘Oh, no, no. I could never afford even a single chip, I don’t have any money for that’. But he had connections with chip making companies, and if I designed the computer he would get me the parts.”
The Stanford Linear Accelerator Center was also a key part in helping Wozniak drive his passions without much money and access to resources. He often drove to the centre with Jobs to hunt around for free information on computers and electronics.
“It was the top physics research place in the world then. There wasn’t any access to computer information, there were no books on computers in the book stores.
“So I would drive down to Stanford Linear Accelerator Center, where the smartest people in the world never locked doors.”
He and Jobs heard about a telephone that could dial to any country in the world, which in those days you wouldn’t even dial from San Jose to San Francisco because it was so expensive. Their curiosity lead them to hunt around for any free resources they could find and learn how to build such an advance system for its time.
“I would always on a Sunday find a door open in the main building and go to the tech library and read magazines about computers. I did a lot of my education that way.”