Artificial Intelligence- Benefit or Risk?

By: Tejasvini Ramesh- Web Developer, Interviewer

Artificial Intelligence (A.I) has begun developing a heightened existence in our society. This ground-breaking ability allows technology to possess human-like intelligence and has incredible potential to enhance our livelihoods - or become our worst nightmare.

Avengers: The Age of Ultron

On May 1st, 2015, Marvel Studios released Avengers: The Age of Ultron, the gripping, action-packed sequel in the 'Avengers' franchise. The film revolves around Ultron, a peacekeeping program powered by artificial intelligence. Ultron was created by Tony Stark (Robert Downey Jr.) and Bruce Banner (Mark Ruffalo) to independently operate the Iron Legion, a group of androids designed to shield citizens when the Avengers battle villains. Previously, the Iron Legion was controlled by JARVIS, an artificial intelligence system tasked with running Stark Industries and the Iron Man suit's computer system. After viewing a harrowing vision- including the Avengers’ deaths, and a calamitous invasion from the Chitauri (cybernetically enhanced aliens), Stark develops the idea of "putting a suit of armor around the world." He decides to work with a reluctant Bruce Banner to configure the world's most competent artificial intelligence system to protect the planet. After some slow progress, the Avengers decide to celebrate their recent victories through a party. Aware of the ongoing experiments, Stark directs JARVIS to watch over Ultron. As the Avengers enjoy their night off, Ultron suddenly activates, maintaining a fully aware consciousness, despite what Stark and Banner intended. Ultron determines that the only pathway toward world peace is to eradicate humanity, claiming that the world must "evolve" to end our planet's turmoils. Ultron destroys JARVIS, kickstarting the series of attacks to initiate humanity's extinction.

Photo Credit: Computer Robot, WIX Media

Is A.I Research Necessary?

Although A.I seems like an alien concept, it's actually all around us. From Google Voice to SIRI, A.I is a lot less complicated and a lot more common than Hollywood's portrayal. However, according to Max Tegmark, President of the Future of Life Institute, today's A.I is known as narrow/weak A.I. This is because it is only capable of performing a simple task, such as facial recognition or internet searches. According to Tegmark, many scientists hope to achieve general A.I (AGI), the technology that outperforms humans at almost every single cognitive task. These systems constantly undergo self-improvement and evolution, which could eventually lead to an intelligence "boom," leaving the human acumen in the dust. However, general A.I. may have the ability to help us rid our world of war, disease, and famine.

How Did Ultron Go Rogue?

On the flip side, this creation may also be our last - unless we learn how to teach the A.I to work for our goals. We should take serious consideration before attempting advanced A.I. Having doubts about its potential for superintelligence leaves a risky margin for error. If we proceed without caution, then our society may end up becoming a victim to its creation in real life - just like how the world did with Ultron. This is concrete evidence as to why we need to conduct more research on the consequences of A.I, thus helping us prepare for and eradicate these negative consequences in the future - especially since A.I is advancing faster than ever before.

Ultron: The Rogue Artificial Intelligence

With all the discussion on the development of A.I, is 'Ultron' a likely scenario with this revolutionary technology? Probably not. But we can't turn a blind eye to the fact that Ultron represents the epitome of rogue A.I, seeking to overthrow and exterminate its primary creators - humans. A fundamental theme in The Age of Ultron is the conflict between organisms and A.I - but to what degree could this dispute exist in real life? In a study conducted by Volodymyr Mnih and Google Deepmind, a neural network was challenged to learn how to play video games. According to E. Paul Zehr, an author at Discover Magazine, the idea was to see if the software could learn how to apply the lessons learned in one game to master another. For the majority of the games examined, the A.I program surpassed humans. The list included various video games such as Boxing, Robotank, and Tutankham. But why is this information so valuable? Why should I be worried about an A.I gamer taking over the world? Well, according to Zehr, the key takeaway from this had absolutely nothing to do with games. The A.I software was adapting to a situation that its program did not prepare for. It was thinking on its own, formulating its own thoughts without assistance. This process is the first step toward rogue A.I - mainly because it allows for the software to build its own opinions, just like Ultron did.

Can Our Future A.I Be Dangerous?

According to the majority of researchers, A.I is not going to exhibit human emotion. However, in the worst case, A.I could be programmed to do something mortifying. For example, A.I weapons are programmed to kill. In the wrong hands, this could result in global calamities- just like the scenarios we saw in The Age of Ultron. As a result of wanting to defend the Earth from alien invasions, Tony Stark created Ultron as an A.I commanding system that would run the Iron Legion, also another example of A.I. With the capabilities of self-awareness and opinion, Ultron was able to use the Iron Legion to cause mass destruction. Although these scenarios are futuristic, we still cannot ignore the fact that A.I is a beneficial, but dangerous use of technology.

Photo Credit: Robots With Guns; WIX Media

Artificial Intelligence as a Risk

The A.I may be programmed to do something useful, but it doesn't use proper methods toward achieving goals. According to Max Tegmark (Future of Life Institute), this can happen if we fail to align the A.I's goals with our goals. For example, asking an obedient autonomous vehicle to take you to work as fast as possible, may not be the wisest thing to do. You'd probably get there chased by helicopters, police, and a pile of speeding tickets. As said by Tegmark, the system literally does what you asked for. If A.I is assigned ambitious projects - such as removing all the trash from the oceans, it may ruin our ecosystems and view our attempts to stop these side effects as a threat.

Future of A.I

According to several experts, the primary fears about artificial intelligence isn't necessarily the indignity, it's the competence. If we don't know how to align A.I goals with our own, we would be in grave danger. Many famous experts in science and technology, such as Bill Gates and Stephen Hawking, have expressed interest in A.I consequences and benefits- specifically, the question of whether the benefits outweigh the risks. But let's be real, it's probably going to take us a long time before we can develop A.I capable of so many things. However, we need to start research now and determine how we want to make progress with this revolutionary technology - especially since it may take decades to conduct enough research to ensure safety.

Why We Need To Start Research Now:

Artificial intelligence is capable of doing so much more than humans, and as a result, we don't know how it would react to us. As Ultron said, the best thing we could rely on is "evolution." Humans are at the apex for a reason - when we're surpassed in our abilities, how would we remain at the top? There are so many questions left unanswered as well. How would we decide who gets to use A.I? Can we control people from conducting unsafe A.I experiments? Is it possible to regulate A.I use to make sure it doesn't get into the wrong hands? These are just a few of the questions that we don't have the answer to. Considering that many A.I programs can formulate their own thoughts, we have to look at the consequences - not just the benefits. Regarding the development of A.I, we must conduct more research - especially considering the rapid progress rate. After all, we don't want to create a system that leads us to our own demise.