Tag Archives: artificial intelligence

Learning Lists

Learning Artificial Intelligence

Last year, my partner and I designed Gump, a Voice-Commanded Bipedal Robot for our Engineering Graduation Project. Gump won first place in the Engineering Faculty for Best Project and is currently featured in Engineering News Magazine at Concordia.

The moment we saw our robot walk on two legs for the first time and respond to the sound of our voice was miraculous.

That moment is when we knew we wanted to work on artificial intelligence for the rest of our lives.

Learning Artificial Intelligence with code(love)

Learning Artificial Intelligence with code(love)

That moment is when we knew we wanted to work on artificial intelligence for the rest of our lives.

As technology grows, so does the information that is available to us.

We began working on artificial intelligence in the NLP (Natural Language Processing) domain, something that allowed machines to understand and process human language patterns. We worked on speech recognition and now we are doing text analysis using artificial intelligence.

We’ve learned that the possibilities of using artificial intelligence are endless. Learning artificial intelligence is crucial to understanding how the 21st century will unroll—Chris Dixon, a prominent venture capitalist, argues that the next 10,000 startup ideas are clear: take x and add artificial intelligence to it.

With artificial intelligence, you can get unlimited insights from Big Data. Imagine measuring the mood of iPhone Users on Twitter after Apple’s latest product launch, gauging the world’s opinion on a major event like Russia’s invasion of the Ukraine, or predicting a company’s performance based on previous financial data, without any need for a human to relay that information to us.

Imagine a world where machines can automatically process information, and present it to use in a relatable and relevant form. This is what artificial intelligence proposes.

Artificial intelligence can help us all be more informed in less time, allowing us to make quicker decisions by getting the insights and analytics we need from Big Data.

Learning Artificial Intelligence

Learning Artificial Intelligence

Because of all this, I’m excited about artificial intelligence. I want you to be as excited as I am, so I’ve decided to highlight some ways to get started with artificial intelligence using resources available online

1-NLTK with Python

NLTK stands for Natural Language Tool Kit, an open source library with lot’s of functions to help you use NLP and artificial intelligence. to solve a problem or work on a project. You must know Python to get started with NLTK. To begin your path, I recommend the Coursera class offered by Stanford University. You might want to brush up on your Python skills with this great book for Python and NLTK beginners and then level up and learn more about Natural Language Processing by looking at this book.

2- IBM Watson Sandbox

Watson is IBM’s artificial intelligence., famous for defeating two human contestants on Jeopardy by having only the Wikipedia database to answer questions. It is considered the gold standard for artificial intelligence nowadays. IBM released a sandbox version for using Watson’s APIs for 30 days, which allows you to hack away and create a cognitive application.

If you want to use Watson for your next big thing, you won’t be able to unless you’re a Series B Funded startup, but that will change in the future. The access IBM has given to Watson now bodes well for the future. The ecosystem is still evolving, but it is the gold standard of artificial intelligence, and IBM is working to make it more accessible.

3- Udacity Courses

Sign up for the free courses on Udacity for Introduction to Artificial Intelligence, and the three Machine Learning Courses offered there. There are tons of resources and a community there to answer all your questions. We personally believe that Machine Learning should be a mandatory course for every Software Engineer and Computer Science Major.

Check out the forums as well, they’re very insightful and they will help guide you step by step.

4- MATLAB’s Neural Network Toolbox

This one is costly, but if you can get a MATLAB license with Neural Network Toolbox license as an add on, you have a very powerful system to play with. NNT is an Artificial Neural Network that has amazing capabilities for your project like speech recognition, image recognition, and object detection. We loved playing with NNT for our robot, Gump.

With a few lines of code, you can create artificial “neurons” in your program. The neurons can be assigned to do several things. We created eight neurons with MATLAB, each one assigned to a specific voice command. After that, we trained the neurons by recording over 5,000 voice samples from every person we could find on campus. The more voice recordings each neuron had per voice command, the more it grew in knowledge. When we tested it, the response was just perfect! MATLAB is immensely powerful.

Those resources will help you get started with learning artificial intelligence.

Leave comments below if you found this helpful, or know of other resources!

This post was written by Yaz Khoury, the founder of an A.I startup called Summarit that uses artificial intelligence to summarize articles for students.

Longform Reflections

Bringing the Dead Back to Life

“IN THIS TEMPLE AS IN THE HEARTS OF THE PEOPLE FOR WHOM HE SAVED THE UNION THE MEMORY OF ABRAHAM LINCOLN  IS ENSHRINED FOREVER”

01000110010011110101001001000101010101100100010101010010

Escaping Biology  

Ray Kurzweil is convinced that his father is going to come back from the dead.

He’s also convinced that he will be one of the people who will help bring him back. His father died when he in his early 20s, and they have not seen each other in 40 years, but Ray has kept enough of his old documents, and photos to remember his father by.

He’s not going to wait with a cache of relatives around the decayed body of his father, and apply electrical currents that will bring the blood racing back into his father. His father will not stumble around the room, dazed at the years he would have missed. They are never even going to physically touch, or look at each other. Instead, they’ll interact with glass screens.

His father will be a virtual avatar that will so closely resemble the biological consciousness that defined Ray Kurzweil’s fantastic quest, that Ray will not know the difference between his virtual and biological father.  It will be programmed through past artifacts to act, and think about the future like his father would have. His virtual father will be one of the holy grails of computing theory, an artificial intelligence that will have passed the Turing Test of being indistinguishable from a human.

Ray graduated from M.I.T, and he is now the Director of Engineering at Google, which puts some of the fiction out of the science component of this idea. He has been the leading proponent of the technological singularity, a moment in time when artificial intelligence will begin outpacing human intelligence, and where society will be upended by the combination of man and machine. As technology has gradually accelerated, the clarity of the vision of virtual life becomes more and more well-defined. Smartphones with more processing power than the supercomputers of years past are now the norm. Is it unfathomable that robots with human-like minds and spirits could be so far off?

Ray’s alma mater is a driving force behind this powerful thought. Every day, almost unbelievable things are accomplished through the manipulation of scientific research at M.I.T, from computers that can convey the sense of touch over different continents, to printers that will be able to spit out entire houses. It has been described as a “Hogwarts” for Muggles, a place where “magic” is common-place.

The metaphor is a beautiful one, but it begs another comparison that may not be so flattering. In the Harry Potter series, we are confronted with the search for the Deathly Hallows, among one of which is the Resurrection Stone. The Stone is able to bring the dead back to life, but at a terrible cost. The reincarnated dead are never truly like they were in life, which leads to a sense of dissonance from the living. This ironically leads to death for all, as the living choose to join the dead in their true form, rather than to accept living with a broken figment of a departed soul.

Questioning Identity

Will virtual intelligences ever be anything more than a figment of a real person? The question examines everything humans have always assumed about human nature: that we are unique, and that we are defined by our uniqueness against non-humans. We possess a strange combination of social interaction, physical manipulation, and processing power that is hard to define, so we often use comparisons to living things that are distinctly not human to define ourselves.

We are not cows. We are not dolphins. We are not chimpanzees, even though that is getting uncomfortably close.

The closer robots get to piercing that space, the more uncomfortable humans get with them. This is the “uncanny valley”. The more robots look, and act like humans, even if we distinctly know they are not, the more we revile them. Like the broken souls of the Ring, poorly designed robots can lead us to hate, and to pain, because they lead us to question who we truly are.

Virtual life that humans can accept must pass the Turing Test. It must fool a human into thinking that it too is a human, that it is really he or she. When Ray sits down to talk with his reincarnated father, he cannot be talking with a robot, but with a real, living human being that he has been yearning to speak to for forty long years.

Ray Kurzweil believes that will happen within a couple of decades.

That robot masquerading realistically as a human being will have all of the characteristics of the thoughts and patterns laid down by his father ages ago. It will be, Ray observes, more like his father than his father truly was due to robotic precision.

This is where the discomfort truly sets in. If robots can so accurately simulate the human condition, what truly makes us human?

The definition of human nature, and what is human or not has important implications. It is hard to imagine that the atrocities of the 20th century could have ever occurred if groups of humans regarded other groups with the basic respect and dignity they accorded to their own group. Indeed, the victims of atrocities have often been relegated to sub-human status: recall the Hutus of Rwanda who insisted on calling their Tutsi victims “cockroaches” during the infamous Rwandan genocide.

Who we are, and what we are here for defines a large part of the human condition. Any threat to the stable framework we have evolved of what is human and what is not represents a seismic change that brings opportunity, and risks.

How many among us would not want to hear the thoughts of Martin Luther King Junior on the civil rights issues of today—the thoughts of Harvey Milk on gay rights in present-day America, the thoughts of Abraham Lincoln on how he would run the country today, the thoughts of Kurt Vonnegut on modern-day warfare—how many among us would want to bring a loved one back, if only for one conversation, nevermind an infinite many…

Yet a virtual intelligence that can evolve beyond our comprehension, and that threatens the very foundation of our identity is nothing to be trifled with. It may be inevitable, but the ramifications are unclear. All that can be known, from a human perspective with our limited processing and predictive powers, is that there will be change of some sort.

Human Change

I would bring my grandfather back. He raised me when I was a baby, and he was ahead of his age. He insisted that my mother and her sisters get an education, even if that was not a popular view at the time. He survived fighting World War Two with his spirit intact. He is a large part of who I am, even if we never got the chance to have a full conversation.

When God burned down Sodom in Genesis, he commanded that the people not look back. In many ways, this command is a wise wall to insulate one from the tragedy and certainty of death. Nevertheless, it is a wall we have all attempted to scale, one way or another.

Lot’s wife famously disobeyed this command, peering back at the ruined city. As punishment, she was turned into a pillar of salt. As Vonnegut wrote, “And Lot’s wife, of course, was told not to look back where all those people and their homes had been. But she did look back, and I love her for that, because it was so human. So she was turned into a pillar of salt. So it goes.” So it goes.

Even if we are all turned into pillars of salt by this new possibility and the desires it creates, it occurs to me that nothing could be more human. Knowing that a part of you, and your loved ones endures, even if you do not, is something that is a constant desire of human nature. It is why we plant stones in the ground for those who passed, why in Chinese tradition we burn food and money for our ancestors, and why we build things for the future we will never enjoy ourselves.

In creating virtual reincarnations of the dead, we will be satisfying the most human of urges: the desire that you and the people you love will endure, and leave a legacy of ideas for future generations.

It may very well be the one defining trait that distinguishes us as humans.

————————————————————————–

Want to check out more of Ray Kurzweil’s story?

 http://abcnews.go.com/Technology/futurist-ray-kurzweil-bring-dead-father-back-life/story?id=14267712

Want to know more about the technological singularity?

http://www.singularity.com/

Inspired to pick up coding and understanding machines?

http://www.code-love.com/2014/01/26/five-brilliant-resources-for-learning-how-to-code-design-and-think/#comment-5