Artificial Intelligence — Will It Outpace Humans?
The Artificial Intelligence (AI) discourse comprises two camps. One who view it as the hope of tomorrow; a breakthrough which, combined with human.
Updated
On this page
Key takeaways
- ***The Artificial Intelligence (AI) discourse comprises two camps.
- It could collate and sift through its learning, selecting the choice that had the highest potential for a positive outcome.
- There is no crystal ball to state if one will overtake the other, but for now, we will just have to allow technology the upper hand and to let AI steadily improve and run amok in our society.
The Artificial Intelligence (AI) discourse comprises two camps. One who view it as the hope of tomorrow; a breakthrough which, combined with human capabilities, can augment and transform every walk of life. Then there are those for whom AI is the harbinger of a dystopian future. One that, more often than not, resembles some variation of the Matrix. The reality? AI is still in its infancy. It is evolving because we are continually seeking ways to simplify our existence. In that case, could we go too far? Could AI, at any point, become smarter than us or is its existence dependent on the cognitive abilities and native intelligence of humans?
A central tenet of the 4th Industrial revolution, AI has been more transformative than anything we have ever seen, at times challenging what it means to be human.
From consumer teach to enterprise-grade solutions, artificial intelligence is firmly ingrained in our innovations. It might be easier to collate a list of the part of modern society that is yet untouched by AI than the other way round. AI is influencing every facet of our daily lives, helping improve efficiencies and augmenting human capabilities.
The Science Behind Artificial Intelligence #
In recent years, artificial intelligence has been popularized by software’s like Sir and Google Duplex, though it has been in the works since the 1950s. The human brain has been the primary inspiration for AI researchers, with the quest for replicating cognitive abilities of the brain, being the most elusive of the goals.
As it stands, machines are already observing, analyzing, and learning from data and mistakes just like humans can. Instead of being programmed in what to think, AI uses statistics, mathematical algorithms, and artificial neural networks to automatically model and find patterns in data to predict an outcome. Given the amount of data we generate from the touchpoints in our daily life, it is little wonder the use of artificial intelligence has exploded in recent times.
Abstract Thinking – Can AI Emulate It? #
Recently Google released a speaking AI that mimics human conversation, and a google subsidiary, Deep Mind, created an AI that beat two world champions in a board game called ‘Go’. The reason the latter is remarkable is that ‘Go’ is one of the most complex strategy games. There are so many potential moves and game states that it makes chess look like toddler play. Yet, ‘Alpha Go’, Deep Minds innovation, conquered it by keeping the memory of the games it has played and studying them as experience. It could collate and sift through its learning, selecting the choice that had the highest potential for a positive outcome. Does that sound a lot like a human brain, learning from its experiences and using that to decide future interactions?
But given all this, the reality is more humbling. While there is substance behind AI’s hype, its underlying hypothesis still centers around the availability of data. The most sophisticated of machine learning algorithm is at its core a computer software that requires the right data sets to problem solve. From facial recognition software that compares billions of pixels to AI-powered Rob-analysts in banking software’s, the same rules apply.
Technology evolves, progresses, and becomes more efficient. Deep neural networks enable AI algorithms to analyze data through multiple layers with greater accuracy. Generative adversarial networks let AI create data based on the data set it has trained on, while reinforcement training lets machines learn from feedback-based trial-and-error interactions. However, the fundamental principle remains the same – if you can translate the task into data, AI will do it.
But can one break down common sense into data? Can we create datasets that enable AI to think in abstracts when situations do not follow the predetermined course? How do we mathematically define, curiosity?
Artificial Intelligence vs Our Cognitive Abilities #
In continuing with the argument around data, one thing we can certainly say–Humans are dreadful at storing massive amounts of data and are terribly slow in processing larger sets of information. In the time it would take a human brain to translate 200 words, an AI-based algorithm would translate at least a million. The fact is computers surpass us in any operation that involves mathematical computation or data processing.
However, the human brain is exceptionally adept at deciding based on common sense, instinct and scarce information. We can transfer learning across parallel events, use limited information to determine the next best course of action, and when all else fails, we use creativity to construct new responses. What comes naturally to humans is the hardest to translate into machine learning.
The Future as It Stands #
Artificial Intelligence is the cornerstone of our existence and is here to stay. Machines do obviate the need for a human workforce in some sectors, but they are not a complete replacement. The future does not lie in the hands of either. It is based on the collaborative and complementary intelligence of both; the cognitive abilities, leadership, and creativity of one combined with the speed, scalability, and quantitative strengths of the other, is the way forward for businesses and humanity alike.
For sure, as technology advances and gives AI the wherewithal to handle generalizations on its own, the heated debate on its regulation needs to take center stage. Black-box models of AI can increasingly make predictions and decisions as humans do, but without the ability to communicate the reasoning behind it. Their outcomes could potentially be on patterns which we as humans cannot perceive. How does one then justify a course of action determined by a machine?
In either case, we are at a dawn of technological evolution – one where man and machine will need to learn to live together, where the existence of one entirely depends on the other. There is no crystal ball to state if one will overtake the other, but for now, we will just have to allow technology the upper hand and to let AI steadily improve and run amok in our society.
Related reading