We discuss the pros and cons of AI development.
It isn’t future anymore
“By far the greatest danger of Artificial Intelligence is that people conclude too early that they understand it.”
― Eliezer Yudkowsky
Artificial intelligence has been a topic of interest since decades with deep social and cultural impact. It has been glorified in many movies and books for years and had always been the thing of future. We never felt any immediate threat from AI taking over the planet and humans being reduced to slaves.
But times have changed over the last decade. We have made immense progress in AI. We have self-driving cars and computers that can understand and even diagnose diseases. Somewhere we failed to realise that computing power has increased exponentially in accordance with Moore’s Law.
We have two groups of AI followers. One who welcome AI and think that this is a natural progression in evolution and the other group who are sceptics and think that that the rise of AI will eventually lead to a revolution of machines against the human race.
Hope we're not just the biological boot loader for digital superintelligence. Unfortunately, that is increasingly probable— Elon Musk (@elonmusk) August 3, 2014
It’s a bit early to decide which of these groups are right, but for sure once we have AI, the world wouldn’t be the same again. Whether humans still remain the overloads and keep AI on a leash or is it a natural course of evolution where we are just an intermediate species eventually leading to the development of artificial superintelligence as Elon Musk mentions in his tweet waits to be seen. But either way, it is not going to be entirely good news for the human race.
While the likelihood of a judgement day like the one depicted in Terminator movie is scarce, our destruction can come in other formats. In today’s episode, let us discuss how.
A brief history
The term Artificial Intelligence was coined by John McCarthy back in the 1950s. According to the Dartmouth Summer Research Project on Artificial Intelligence of 1955, there are seven aspects of AI:
- Automatic Computers with the ability to do higher functions of the human brain.
- Programming a computer to use general language.
- Neuron Nets — forming concepts by arranging hypothetical neurons.
- Theory of the Size of a Calculation — ability to determine and measure problem complexity
- Abstractions — the ability to deal with ideas rather than events.
- Randomness and Creativity
These concepts were far ahead of the times in the 1950s since these required computing power that was much more higher in magnitude than that existed. But over the years, computers have become far more powerful than before. Smartphones these days have much more computing power than any computer that existed during the 50s and 60s, including the one onboard Apollo Spacecraft that took men to the moon.
But during recent times, computers have been able to achieve some of the above-stated aspects if not all. We have seen self-driving cars, smartphone assistants like Siri. Computing power no longer seems to be a constraint. According to Moore’s law, the computing power doubles itself every passing year. Along with being powerful, computer components are also becoming cheaper.
These factors have helped AI grow in leaps and bounds to a point where serious discussions are happening around the world about the pros and cons of an AI driven world.
The problem statement
Media and culture have been very vivid in their portrayal of AI in the form of killing machines and robots with the sole intention of finishing off the human race to avoid conflict of interest. In movies like Terminator and The Matrix, AI is depicted as a self-aware system that views humans as inferior species.
While this is a kind of future that excites a movie buff, the real world problems due to AI expansion might be much more of a real world problem. What does it mean? Let us see.
AI today is capable of seeing and understanding. It is also capable of reading and writing as well. Text to speech conversion has reached new heights with AI. Assistants like Siri and Alexa can listen to you and make decisions based on your inputs. There lies the problem.
While movies depict AI in the form of a killer robot of some sort, that might not be the real case. These days, everything is online and automated. Banking systems, power management, IOT to name a few. AI taking over a distributed system is as scary as a robot shooting laser on your face.
All the capabilities would lead to AI replacing humans in the services sector. If we take a look at the way current world economies work, all the developing nations either depend on agriculture or related industries whereas developed nations depend more on the services industry. The same industry where AI is capable of replacing human beings. Are we going to lose our jobs? Time will tell.
As Sam Harris mentions in his TED Talk, what this means is that any slightest divergence in the goals of humans and AI will lead to some form of catastrophe. He gives the example of humans and ants. Human do not usually harm the ants, but suppose there is an ant hill which is in the way of his construction project, the man wouldn’t think twice before destroying the ant hill and killing all the ants in the process.
Similarly the moment AI realises that their goals are in conflict with ours, we might end up suffering the fate of the ant.
If you're not concerned about AI safety, you should be. Vastly more risk than North Korea. pic.twitter.com/2z0tiid0lc— Elon Musk (@elonmusk) August 12, 2017
Coming back to Elon Musk’s tweet, it seems like a natural progression to evolution is for human beings to give way to AI. Human beings have been explorers since ages. Now we are at a point where the Earth seems like a small place to live in. It wouldn’t be long before we realise that the natural resources are no longer going to be able to support the population on the earth.
Once we are saturated on the Earth, the natural thing is to explore the space and head out and find other habitable locations. SpaceX is planning to send men to Mars in the coming decade. Mars seems to be at a reasonable distance for humans to have a safe journey in a lifetime. But if there has to be an interstellar travel, there either has to be a way to travel way faster than the speed of light (because the nearest star is almost 4.7 light years away) or a way to carry human beings in a state of dormancy.
AI will take control of this interstellar expansion. If human beings are capable of keeping AI on a leash and act as masters, we have a better chance of survival. On the other hand, if AI realises that a human is a liability, then that’s the end of the human race as we know. This is what is portrayed in movies where AI goes rogue.
In spite of all the things that can be listed here which might make AI scary, they have some definite advantages over humans. A future where humans and AI co-exist and work together, that could be considered as a Superintelligence. That would be the bright future that we wish for ourselves. AI giving it’s incredible and infinite capabilities to help human beings in their future quests.
That is a future that we hope for. That is a future we should dream about. Time will tell if that turns into reality. For now, we can take sides and debate on the influence of AI on our future.
Music: Free Stock Music
- Rip and Tear