Is Sentient Artificial Intelligence a reality now?
The Curious Case of Sentient Artificial Intelligence
“Everything we love about civilization is a product of intelligence, so amplifying our human intelligence with artificial intelligence has the potential of helping civilization flourish like never before – as long as we manage to keep the technology beneficial.“
- Max Tegmark, President of the Future of Life Institute
Welcome to the 21st century. An era practically transcending into Keanu Reeves’ Matrix - rendered so, by an increasingly sentient artificial intelligence (AI). The almost meteoric rise of AI over the last few decades has truly revolutionised the manner in which the benign Homo sapiens interact with technology. From helping compose music to offering voice assistant services like the popular SIRI, AI is gradually acquiring an omniscience like never before.
In layman terms, AI entails an evolving set of technology that imbues humble machines with the ability to simulate human thinking. Not in its entirety, but certain elements of human thinking – such as reasoning and learning – can be easily simulated by machines or applications today, thanks to AI.
Image Courtesy: Duy Pham Nat
The mighty Google’s search algorithms, for instance, are based on machine learning (a form of AI). This empowers systems to learn and pick patterns from available data and even make predictions about it. Google’s Assistant, Apple’s Siri, Microsoft’s Cortana are all premised on the dynamics of AI.
But have we really reached what many deem as “technological singularity”, a point where AI truly becomes sentient? A juncture at which AI acquires a consciousness that marks an important distinction between humans and machines? At this point in time, machines or robots will cease to be mere tools. For Ray Kurzweil - an inventor, computer scientist and futurist – technological singularity would bring an overhaul in life as we know it today. We haven’t reached it yet. So maybe we are still on the path to turn that ultimate fantasy to reality, although we have made several strides. To understand it better, it’s necessary to learn what AI actually is about and then decide if such sentience is actually possible.
Understanding AI: Is it all ‘Real Steel’?
Common parlance makes us think of robots and only robots when it comes to Artificial Intelligence. Our minds picture ‘Terminators’ or ‘Wall-Es’ or something like the ‘Big Hero 6’. Rarely ever, do we make the effort of actually getting to the heart of the matter.
So let’s reconfigure our minds a little and picture this: our ‘body’ sans the ‘mind’ is a mere entity. An entity, with no sense of direction or purpose or ability. Our hands and legs and other organs are merely tools with certain functions but without any ability to actually perform them.What makes our organs run or function is our mind. AI is pretty much like that. It is not the robot but what resides within the robot – the brain. The brain empowers the robot to be what it eventually turns out to be.
Siri, for instance, is AI. The voice we hear, the one which interacts with us is merely a personification of the AI on which it is premised. Moreover, AI has different variants which warrant special mention.
Types of AI: What kinds of ‘Chappies’ are there?
Artificial Intelligence should be viewed as an overarching framework within which its different manifestations fall. Of course sentient artificial intelligence would be one superior kind, but it is not the only one. What distinguishes one AI from its other variants is its ‘calibre’. That is, what an AI can essentially do.
Image Courtesy: Seanbatty
Calibre 1: Artificial Narrow Intelligence: This variant of AI is sometimes also deemed as weak AI. Because it is capable of performing only one or a specific kind of function. Because it specialises in a particular field and doesn’t really bother multi-tasking, it is narrow in its scope and functions.
Calibre 2: Artificial General Intelligence: This variant possesses intelligence capacities of a human. This is why it is also referred to as Human level AI (or strong AI). All intellectual tasks that the human mind can perform - AGI can perform as well. Reasoning, learning, problem-solving, abstract thinking and so on are some of the things it can easily do. But we haven’t yet reached here because quite obviously, AGI is more difficult to achieve than ANI.
Calibre 3: Artificial Superintelligence: This variant, like its name, beckons the future. Sentient Artificial Intelligence is all about machines being able to possess an intelligence way more profound than ours. Nick Bostrom, an AI thinker, says that superintelligence is about “an intellect that is much smarter than the best human brains in practically every field, including scientific creativity, general wisdom and social skills.” It really is a machine possessing an intelligence of a magnitude beyond our understanding potential.
The Present Scenario: A Glimpse
Dr. Hiroshi Yamakawa, Director of Dwango AI Laboratory, a leading AI researcher says that in the present, AI is performing specific functions alone. That we have only been able to reach ANI. Your precious little phone which allows you to talk to Siri or Google Assistant, is in essence an ANI sweatshop. Google Search is one gigantic ANI system. The Spam email filter, which smartly tells you what is and isn’t a Spam? The tech embedded in cars that would make some people fall for ghosts (we are talking about self-driving cars here)? Google Translate, Voice Recognition? And oh, ever wondered how Facebook figures out who you might want to befriend? It’s all ANI.
Image Courtesy: Panorama Marketing & Media
But the buck doesn’t stop here. Because many are eagerly figuring out ways to make the move from ANI to AGI. Dr. Hiroshi Yamakawa talks about how this move, although extremely tough, isn’t impossible to make. It would require achieving massive computations prowess and the ability to make machines simulate human general intelligence. He says that this is tough because it can’t be broken down into specified tasks easily. Nonetheless, we have made certain strides in this direction.
Image Courtesy: MikeRenpenning
Image Courtesy: GDJ/2402 Images
Deep learning (a form of machine learning) models itself on the manner in which synapses and neurons function in the human brain. It has been able to ascertain how they change and respond to differences in stimuli or input. This has proved to be pivotal for voice recognition software. In fact, machine learning and AI will together pave way for more complicated forms of AI to come into being. Sentient Artificial Intelligence would probably be a result of this in the near future.
No Man’s Land: Hailing the Age of Ultron?
“I’m increasingly inclined to think that there should be some regulatory oversight, maybe at the national and international level, just to make sure that we don’t do something very foolish. I mean with artificial intelligence we’re summoning the demon.” —Elon Musk , at MIT’s AeroAstro Centennial Symposium
Image Courtesy: Graphic Competitions
With great power comes great responsibility. And such concerns have always hovered around the field of AI. This has augmented further with the possibility of sentient artificial intelligence coming into being soon.But the problem here really isn’t about whether or not AI would turn malevolent. That we’d end up creating vicious robots that would destroy humanity. This, in fact, is one of the misconceptions people hold when they think of the cons of achieving ASI. Competence is the real concern.
When AI becomes truly sentient, the major problem would be aligning its goals with ours. Because the machine would never really have issues with reaching its goals. It is only when its goals would conflict with ours that problems would arise. A missile is a goal-oriented machine with a target to hit. But if what it wants to hit is something misaligned with what we want to achieve, then the problem isn’t really with the machine. That is why there’s a huge focus on a beneficial AI movement.
A Few Concerns
Nonetheless, intelligence is nature’s gift to us. Creating sentient artificial intelligence powered with an intelligence superior to ours involves major ethical concerns. It involves questions about job automation which would disempower many. It involves concerns about several social problems, like the ones about women and minority rights.
What kind of change will a super intelligent machine be able to bring about in the lives of these social groups? Will it bring an overhaul in our prejudiced mind-sets that give birth to the problems these groups face? Recently, Saudi Arabia became the first nation to give citizenship rights to a robot named ‘Sophia’. In a country where women have barely been accorded their natural rights, this move seems sadly ironic.
Quantum of Solace: The Brighter Side of the Story
However sceptical some of us might be about the possibility of achieving a fully sentient artificial intelligence, we cannot deny the wonders it will usher in. AI today has become a norm in our lives. We probably cannot imagine a life without the conveniences AI provides us. A sentient AI would probably double this convenience. It will remove human errors and make up for all the things we cannot do.
AI is already doing wonders in the field of medical surgery and space exploration. Haptic Robotic Arm (with powerful tactile sensors) is being used for delicate surgical procedures beyond human prowess. Mining and fuel exploration can be turned into simple procedures with AI involvement. Its daily applications abound. Siri, GPS, the way Facebook tags us in other’s photos, face recognition etc. will all become more complex with better AI.
Image Courtesy: Pinterest
Although unemployment is a major concern, repetitive jobs can be done away with - through constructive use of AI. ASI could resolve problems related to global warming. It could cure cancer and even AIDS. And although we don’t know if it’ll be for the good or not, ASI could even usher in immortality!
Stairway to Heaven or Highway to Hell?
AI is definitely transforming our interactions with technology. It is being used for a variety of purposes in a myriad fields. And there’s no doubt that it has revolutionising potential. Although a fully sentient artificial intelligence or ASI is yet to be achieved, the day when it becomes a reality isn’t very far. What happens then is for us to know and find out.
Image Courtesy: Pinterest
We cannot predict if it’ll prove to be a stairway to a technological heaven or a highway to a high-tech hell. We can only hope to progress in a benign manner. And we can only envision a future where AI and humanity coexist peacefully.
Read the full article