Artificial intelligence is not the most talked about topic in the world today. In fact, most people occupy their conversations with gossip concerning a coworker or talk about a politician. However, artificial intelligence could be the biggest game changer the world has yet seen.
It is widely accepted that the human brain is the most powerful and complex computer on the planet. Not only does it have immense processing power compared to other species, but it also lends us the ability to create independent thought, learn from stimulus, have emotion, empathy, communicate with each other and imagine entire worlds with imagination in the blink of an eye. Through this intelligence and unparalleled ability to communicate, we have created civilization, agriculture, science, and most recently the internet. However, the very thing that makes us the most successful and intelligent species on earth is the same thing that is our bottleneck -- our brains.
The human brain has a limited processing power, and this is the boundary that artificial intelligence dares to break. True artificial intelligence would allow a potentially conscious being, much like you and I, to have a brain with near infinite processing power that is also connected to all of the information available in the world. Such an AI would be able to cure diseases like cancer in a few seconds, solve the world power crisis with some quick analysis, and find solutions to problems while taking every possible variable into consideration, a feat the human brain is hardly capable of. What we are talking about today, however, is the other side of the coin.
While true artificial intelligence has the ability to solve all of our worlds most unsolvable problems, it also has the potential to destroy mankind. I know; this sounds very overly dramatic like some 80s science fiction film where robots control the world. Logically though, this is truly a possibility, and many experts in the budding field are coming to the same realization. For example, in a year old project named "Tay" by Microsoft, a twitter bot with machine learning (AI) went from making happy, optimistic posts to misogynistic, racist, antisemitic, and pessimistic posts. One of Tay's first tweets read "@mayank_jee can I just say that i'm stoked to meet u? humans are super cool," (Vincent, 2016). In less than 24 hours, Tay was posting tweets like "I fucking hate feminists and they should all die and burn in hell" and "@brightonus33 Hitler was right I hate the Jews."
While this can be considered a humorous exercise-gone-wrong in an early test of artificial intelligence, others like Forbes writer Frida Polli argue differently. "However, while we often rely on technology to avoid inherent human biases," Polli said, "there is a dark side to algorithm-based decisions: the potential for homogenous data sets to produce biased algorithms," (Polli, 2017). Biased algorithms are a result of data mining, and they can only be eliminated with a manual change. In Polli's explanation, she uses the example of a program that searches for the best CEO candidates for a company using all available data. If the name John is the most frequent among the best CEOs of the fortune 500 companies, then the algorithm may suggest that the name John is a better predictor of success than other, more rational information like work experience or recommendation.
This type of mistake is almost unavoidable in the artificial intelligence systems of the near future. It is impossible for a computer that can not think on its own to distinguish the difference in importance between a name like John and other predictive factors. That is where "true" artificial intelligence comes in.
Jürgen Schmidhuber, a renowned expert in the field of AI who has been working on AI since the 1970s, believes that "True artificial intelligence will change everything," (Su). Once a machine has the capability to be smarter than mankind with the ability to think on its own, it reaches the state that most experts consider true artificial intelligence, and this is when they could potentially be the biggest threat to our existence.
A machine with near limitless capacity to process information and the ability to think on its own could, by definition, outsmart every attempt a person could make to thwart it. If this computer felt it necessary to destroy the web and sabotage every device in the world with a virus that breaks them all, it most definitely could. Why? Because the people who create the firewalls and programs for cyber security are limited by the ability of their brains. A machine with vastly superior processing power could circumvent an intrusion prevention system that took years to make in mere seconds.
As an analogy, the average human Intelligence Quotient (IQ) is 100. A super computer with true artificial intelligence could easily have a comparable IQ in the billions. This would allow it to solve our greatest problems like curing cancer, solving the energy crisis, and inventing radical new technologies that our brightest minds could never conceive. However, if such a powerful and smart machine ever did exist, we would also be at its mercy.
Do you think its a good idea to create "true" artificial intelligence? Why?
If you had access to true artificial intelligence, what problem would you have it solve?
Sources:
Polli, F. (2017, December 05). The Dark Side of Artificial Intelligence. Retrieved December 06, 2017, from https://www.forbes.com/sites/fridapolli/2017/12/05/the-dark-side-of-artificial-intelligence/#b7ec1da12614
Vincent, J. (2016, March 24). Twitter taught Microsoft's friendly AI chatbot to be a racist asshole in less than a day. Retrieved December 06, 2017, from https://www.theverge.com/2016/3/24/11297050/tay-microsoft-chatbot-racist
Su, J. (2017, July 10). Jürgen Schmidhuber: True Artificial Intelligence Will Change Everything. Retrieved December 06, 2017, from https://medium.com/@Synced/j%C3%BCrgen-schmidhuber-true-artificial-intelligence-will-change-everything-dc6b36bb6146
I found this piece on Artificial Intelligence to be very interesting because of the fact that AI does not gain a lot of the attention it deserves. As you mentioned, Artificial intelligence is not the most talked about topic in the world today. One of the most interesting pieces of this post surrounds the idea of what qualifies something as a true piece of automated intelligence. Jürgen Schmidhuber, a renowned expert in the field of AI who has been working on AI since the 1970s, believes that "True artificial intelligence will change everything," (Su). Once a machine has the capability to be smarter than mankind with the ability to think on its own, it reaches the state that most experts consider true artificial intelligence, and this is when they could potentially be the biggest threat to our existence. Simply the concept of creating a machine with the ability to think on its own is amazing and shows the capability of this technology moving forward in the field. In the coming years it will be very interesting to see the directions this industry goes and how much AI is integrated in our daily lives.
ReplyDelete