The Student Newspaper of Glendale Community College

El Vaquero

The Student Newspaper of Glendale Community College

El Vaquero

The Student Newspaper of Glendale Community College

El Vaquero

Artificial Intelligence: Robots Coming Soon

The year is 2011. Why are there no robots among us?

That was the question that part-time faculty member and self-proclaimed geek Stephen Weese answered via a robotics lecture titled “Where’s My Robot Companion?” on Oct. 25.

“There’s something about computers and humanity that is very compelling to me,” he said. “We [ask questions like] ‘can a computer relate to a human being and on what level and how? Can we do this, and will it happen before I die?’”

Weese discussed the technological advances currently available in the field of artificial intelligence, and how those developments are leading up to the prospect of computerized companions in the near future.

Story continues below advertisement

“Basically, [AI is] a computer or a machine simulating a human thought and actions, and being able to do things intelligently like a human being,” said Weese. “Which is funny because humans often do things unintelligently.”

Machines and computers are currently used to solve technical problems and doing things that bore or disinterest humans, like cleaning or repetitive tasks. However, having computers and robots similar to C-3PO from “Star Wars” that emulate humans is still problematic.

“For computers to be able to function like a human being, they have to be able to sense like us,” said Weese. “Getting a robot to perceive the world like we do is very difficult.”

The main senses computerized robots need to have to exhibit human qualities are touch, sight and hearing, depending on what kind of task they’re built for. Also, they must be able to respond to humans in some way.

“If we make it like a human being, then we do want it to talk, because that’s our primary way of communicating,” he said.

Weese said simple tasks, such as picking up a glass of wine, would have multiple steps that are easy for humans but would be rather complicated for a robot to emulate.

He proposed the example of drinking from a wine glass. A robot would have to identify the coordinates of the glass in three dimensions (something humans do automatically), extend its arm and pick up the glass with enough force to lift it without smashing it. Actually drinking it and placing it back down would require programming multiple steps as well.

“Do you know how hard that is?” said Weese.

Aside from programming mechanical knowledge, computers need to be able to reason and make inferences. A car-driving robot, for example, would need to be able to make split-second decisions if unforeseen circumstances arise, such as boxes falling from a car.

More complex tasks, like getting a sandwich from a shop, would require a large amount of commands and programming to deal with the circumstances and actions involved with the task.

“We humans work in such a general way. We don’t need too specific pieces of information to identify things,” he said. “It’s worse than a baby, because even a baby human has built-in instincts to recognize objects. We have to teach [a robot] everything.”

Weese performed a demo using a friend’s iPhone 4S of Siri, a personal assistant application that recognizes the human voice and performs commands and answer questions based on what was said, as an example of how far AI has gotten in the past few years in terms of voice recognition.

He asked Siri to execute several commands, such as asking the phone to send him a text saying “I hate you,” find the location of a theater where the movie “Paranormal Activity 3” was playing, and asking it directions from GCC to Chicago. The audience asked Siri what the age of the Earth was, 22 divided by seven, and the location of the largest Muslim population in the world.

He asked Siri, “Do you love me?” and received a response of “How can I tell?”

After the demo, Weese discussed expert systems, which are programs that excel in a certain area.

“It’s kind of funny how humans are really good at doing general stuff but really specific things, computers are better than we are – in most cases anyway,” said Weese.

Expert systems are systems that can do one thing over and over, and do it well. Such a computer can excel at things such as chess and trivia, like a human expert.

Notable examples of expert programs are Deep Blue, a chess playing computer that defeated world champion Garry Kasparov in 1997, and Watson, a computer system that plays Jeopardy! and defeated champions Brad Rutter and Ken Jennings in an exhibition match on Feb. 14 this year.

He then showed clips of integrated AI systems, including a Japanese robotic maid, a prototype of a self-driving car, and a female companion that responded to touch and voice that even reacting to inappropriate groping and physical abuse.

Weese concluded by saying that robot companions are around the corner, but that the costs are prohibitive.

“It’s a question of when,” he said.

However evolved robots get, humans will remain exceptional beings.

“As human beings, we really are amazing!” said Weese. “I can’t stress it enough, it’s so crazy the stuff that we do — and looking at AI makes you realize it a lot more.”

About the Contributor
Angel Silva
Angel Silva, Managing Editor
Angel Silva is a former writer for El Vaquero.  He was a Mass Communications major who transferred to Northridge in Fall 2013.  Angel is a technology and music enthusiast and likes to write in his spare time.
Activate Search
The Student Newspaper of Glendale Community College
Artificial Intelligence: Robots Coming Soon