Can We Build a Brain?
⯀ The new series “NOVA Wonders” takes viewers on a journey to the frontiers of science, where researchers are tackling some of the biggest questions about life and the cosmos. One of these questions was about the development of artificial intelligence, "Can We Build a Brain?"
Undoubtedly, artificially intelligent machines are taking over. They’re influencing our everyday lives in profound and often invisible ways. They can read handwriting, interpret emotions, play games, and even act as personal assistants. They are in our phones, our cars, our doctors’ offices, our banks, our web searches…the list goes on and is rapidly growing ever longer. But how does today’s A.I. actually work—and is it truly intelligent?
For that matter, what is intelligence? The world’s brightest computer programmers are trying to build brighter machines by reverse-engineering the brain and by inventing completely new kinds of computers, with exponentially greater speed and processing power.
A recent episode of PBS's NOVA Wonders, looks at how far we’ve come and where machines are headed as their software becomes ever more cerebral. How close are we from a world in which computers take over—from diagnosing cancer to driving our cars to targeting weapons? If we place more and more of our lives under the control of these artificial brains, what are we putting at risk?
The participants in the episode include:
The large number of users for the Microsoft chatbot, Xioice is a very interesting part of the episode. Xiaoice has had over 30 billion conversations with over 100-million people.
"Xiaoice is a lot of people's best friend, including me," states Di Li, a senior engineer at Microsoft and one of Xiaoice's creator.According to the program, many users are unable to tell that Xiaoice is not a person, or perhaps more importantly, that is not important to them. Many of her users treat her no differently from a real friend.
- Rodney Brooks - MIT Professor Emeritus
- Lili Cheng - Microsoft
- Greg Corrado - Google
- David Cox - Harvard University
- Pedro Domingos - University of Washington
- Rana el Kaliouby - Affectiva
- Oren Etzioni - Allen Institute for Artificial Intelligence
- Ali Farhadi - University of Washington
- Geoffrey Hinton - University of Toronto
- Hsiao-Wuen Hon - Microsoft
- Christof Koch - Allen Institute for Artificial Intelligence
- Eric Kolve - Allen Institute for Artificial Intelligence
- Ranjay Krishna - Stanford University
- Yann Lecun - Facebook
- Di Li - Microsoft
- Fei-Fei Li - Stanford University
- Dharmendra Modha - IBM
- Roozbeh Mottaghi - Allen Institute for Artificial Intelligence
- Roberto Novoa - Stanford University
- Peter Rander - Argo AI
- Peter Singer - New America Foundation
- Sebastian Thrun - Stanford University
The large number of users for the Microsoft chatbot, Xioice is a very interesting part of the episode. Xiaoice has had over 30 billion conversations with over 100-million people.
"Xiaoice is a lot of people's best friend, including me," states Di Li, a senior engineer at Microsoft and one of Xiaoice's creator.According to the program, many users are unable to tell that Xiaoice is not a person, or perhaps more importantly, that is not important to them. Many of her users treat her no differently from a real friend.
"Can We Build a Brain?" also points out how Geoff Hinton's use of decades-old neural network technology on the ImageNet challenge kicked off the deep learning revolution. "The person who evaluated the submissions had to run our system three different times before he really believed the answer. He thought he must have made a mistake, because it was so much better than the other systems," states Hinton.
The impact of deepl learning's application for AI is historically marked by the defeat of world Go champion Lee Sedol by a team from Google DeepMind.
"t was only by bringing deep learning, in particular, to this area that we were able to build artificial systems that were able to see patterns on the board in the same way that humans see patterns on the board," states Google's Greg Corrado.
In the past few years, deep learning has invaded our everyday lives without most of us even knowing it.
According to Christof Koch from the Allen Institute for Brain Science, "We've suddenly broken through a wall. When I started in this field, none of that was possible. Now, today, you have machines that can effortless, in real time, recognize people, know where they're looking at. So, there has been breakthrough after breakthrough."
Beyond algorithms', "Can We Build a Brain? also explores the development of neuromorphic computing.
Dharmendra Modha, of IBM states, "Today's computers fundamentally separate computation from memory, which is highly inefficient. Whereas our chips, like the brain, combine computation, memory and communication."
The show also presents a nice review of co-host Rana el Kaliouby's journey from Egypt, to heading her own AI startup Affectiva, which uses computer vision, machine learning and deep learning methodologies to train algorithms that classify emotions.
"It is very unusual, especially for women coming from the Middle East, to be in technology and to be leaders. I remember this one time, when I was supposed to be presenting to an audience, and I walked into the room, and people assumed I was the coffee lady," she relates.
Related articles
Facebooks Yann Lecun adds, "The change in performance on ImageNet was tremendous. So, until 2012, the error rate was 26 percent. When Geoff Hinton participated, they got 15 percent. The year after that it was six percent, and then the year after that it was five percent. Now it is three, and it's basically reached human performance."The impact of deepl learning's application for AI is historically marked by the defeat of world Go champion Lee Sedol by a team from Google DeepMind.
"t was only by bringing deep learning, in particular, to this area that we were able to build artificial systems that were able to see patterns on the board in the same way that humans see patterns on the board," states Google's Greg Corrado.
In the past few years, deep learning has invaded our everyday lives without most of us even knowing it.
According to Christof Koch from the Allen Institute for Brain Science, "We've suddenly broken through a wall. When I started in this field, none of that was possible. Now, today, you have machines that can effortless, in real time, recognize people, know where they're looking at. So, there has been breakthrough after breakthrough."
Beyond algorithms', "Can We Build a Brain? also explores the development of neuromorphic computing.
Dharmendra Modha, of IBM states, "Today's computers fundamentally separate computation from memory, which is highly inefficient. Whereas our chips, like the brain, combine computation, memory and communication."
"It is very unusual, especially for women coming from the Middle East, to be in technology and to be leaders. I remember this one time, when I was supposed to be presenting to an audience, and I walked into the room, and people assumed I was the coffee lady," she relates.
The program concludes with an examination of some of the warnings about AI from the likes of Bill Gates, Stephen Hawking and Elon Musk. They present a fairly optimistic viewpoint.
" I want to challenge Elon Musk. Show me a program that could even take a fourth grade science test," states Oren Etzioni from the Allen Institute for Artificial Intelligence.
Comments
Post a Comment
Thank you for your comment!