The late Arthur C. Clarke, noted science fiction author, said: “We have to abandon the idea that schooling is something restricted to youth. How can it be, in a world where half the things a man knows at 20 are no longer true at 40 – and half the things he knows at 40 hadn’t been discovered when he was 20?” The latter portion of this quote is very appropriate for the computer industry. For example, let’s look back just ten years. What can you do today with a computer that you couldn’t do ten years ago?
- Talk to it and have it talk back to you!
- Have it translate a foreign language!
- Turn on or off your lights, TV, A/C, etc. using your voice!
- Have it learn through trial and error rather than explicit programming!
- Have it recognize faces and images!
The tremendous increase in computing power that enabled these new applications has been fueled by the increase in the density of computer chips which follows a rule called Moore’s Law, described by Gordon Moore of Intel in the mid 1960’s. This law states that the number of circuits on a computer chip will double approximately every two years, and this has held now for the past 50 years.
In the mid 1960’s computers filled whole rooms and required special power and cooling. A large mainframe computer cost millions of dollars. In contrast, IBM earlier this year announced a computer that is the size of a grain of salt and can be manufactured for 10 cents! It has its own memory, power and communications and is as powerful as a desktop computer that you might have had in the early 1990’s. Those machines in turn were more powerful than that room full of equipment from the 1960’s! This improvement in price/performance has fueled the advances in technology we see today, including Artificial Intelligence.
Artificial Intelligence or AI has been in the public consciousness for many years. Most of us remember the 1968 movie 2001: A Space Odyssey where the evil computer HAL took over the spaceship and attempted to kill the crew. Well- meaning or not, HAL clearly possessed what most of us would consider Artificial Intelligence. Here are several other more recent examples from the real world:
- In 1997 a computer developed by IBM called Deep Blue beat the reigning world chess champion, Garry Kasparov. However, Deep Blue derived its playing strength by using brute force, not by exhibiting any real intelligence. At the time it was the 259th most powerful supercomputer in the world.
- In 2011 another IBM creation called Watson beat two human Jeopardy champions on TV. This was an impressive achievement but also was accomplished with brute force. The Watson employed a cluster of 90 powerful servers, and the overall hardware configuration costs about three million dollars.
Modern AI uses more sophisticated techniques than these earlier examples. Here are some relevant terms:
- Machine Learning is a branch of AI that aims to give machines the ability to learn a task without pre-existing code.
- Deep Learning is a subset of machine learning that has networks capable of learning directly from data which can be either labeled or unlabeled. Deep Learning is often made possible by using artificial neural networks which imitate human neurons or brain cells. This technology has fueled many of the recent advancements in AI.
- Artificial General Intelligence (AGI) is the intelligence of a machine that can successfully perform any intellectual task that a human being can. This is the ”Holy Grail” of AI. We haven’t achieved this goal yet (and it’s possible that we may never achieve it). However, some experts believe that AGI is only about a decade away.
Here are some examples of current Artificial Intelligence applications which use the above techniques:
- A team from the United States, Germany and France taught a computer to detect skin cancer by showing it 100,000 images of moles and then compared its performance to a group of 58 dermatologists. In the end, they found that the AI could diagnose melanomas, the most dangerous form of skin cancer, better than the dermatologists. On average, dermatologists correctly identified 86.6 percent of melanomas. The AI, in comparison, accurately diagnosed melanomas 95 percent of the time.
- Exoplanets are planets that exist around stars other than ours. An artificial neural network is being used to recognize and classify exoplanets that are similar to the rocky worlds in our own solar system, like Mars and Earth. This system can recognize what elements are in a planet’s atmosphere by analyzing the light from the world’s parent star passing through that atmosphere. From there, the system can find elements that might indicate the world is hospitable to life, such as oxygen or methane.
- A computer was programmed to study images of tens of thousands of live human stem cells. Over time, the computer learned to look at an image of a typical cell and figure out its internal organization. The end result was a 3D model of a living cell that lets scientists study the interior structures of a cell even when they can only see the exterior and the nucleus.
- Researchers at Intel and the University of Illinois Urbana-Champaign taught an artificial intelligence algorithm how to take the data from darker images and reconstruct them so that they’re brighter and clearer. The team claims the algorithm can now amplify low-light images the equivalent of up to 300 times the exposure, without the noise and discoloration that programs like Photoshop might introduce or the need to take two separate images.
- Artificial neural networks are playing an important role in the development of self-driving cars, a technology that is currently getting close to commercialization. Waymo, a subsidiary of Google, has been at the forefront of this technology for some years. The company is planning to roll out more than 62,000 autonomous Chrysler Pacifica hybrid minivans, modified with their self-driving technology, as they launch an autonomous ride-hailing service in the U.S. later this year.
Also there is testing of self-driving cars going on right here in Tampa. A company called SAE International sponsored a “Demo Day” in May of 2018 where the public could ride in a self-driving car. They closed off a portion of the Selmon Expressway that day for this SAE International hopes to better understand evolving public attitudes about self-driving technology and the experience of traveling in an automated vehicle.
- Using an artificial neural network, Google developed an application called Google Duplex which can make restaurant reservations and hair salon appointments over the phone. Not only can it carry on a conversation with a human being to complete these tasks, but the voice of the AI is indistinguishable from a human. This application should be available on smartphones later this year.
As impressive as these applications are, they are only the beginning. Over time, AI applications will become increasingly more sophisticated. There is growing concern as to what will happen if AI’s eventually achieve or surpass human intelligence. Only time will tell whether Artificial Intelligence will ultimately be a blessing or a curse to civilization, but right now the future looks bright.
Here are links to relevant sources that cover some of the items in this article.:
- Watson Vs. Jeopardy Contestants
- Skin Cancer Detection Using AI
- Classifying Exoplanets
- Waymo/Self Driving Cars
- Google Duplex
Bruce Gobioff retired I.T. professional joined OLLI in 2015. He has taken OLLI courses in history, politics, literature, media and science. Bruce will teach The Computer Revolution: Now and in the Future at Lake Magdalene United Methodist Church starting February 27, 2019. He is a member of the OLLI Board and the political SIG.
3 Replies to “The Computer Revolution and Artificial Intelligence”
Wow, glad I’m a senior!
Hey, Bruce! Is this class instructional or just telling me all that I don’t know or understand what the computer will do? I try to understand but don’t always get it committed to memory.
The future does look promising, especially with the application to medicine and health care. I did not know about the application of AI to detect skin cancer. Thank you for bringing the topic to us with the blog article and your upcoming course.