Is the future well-being of mankind dependent on how we ‘train’ machines? Machines, as has been the case for centuries now, have eased our lives, made tasks less cumbersome to execute, and standardized the quality of output and efficiency. But we now live in times when machines have brains of their own — and with great duties come greater responsibilities. It is, hence, so much more important now to identify where we seek to be and how machines can be used, and to what extent, to expedite the process.

Amplifying Intelligence for Cognitive Computation

Industries worldwide are starting to focus on perceptual computing. And therein lies the difference between human beings and machines – perception. While machines are getting smarter every day, the one thing they lack is cognitive understanding and the power to take decisions on their own. Artificial intelligence (AI) has for decades tried to innovate ways to inject problem-solving capabilities in machines, but they have been met with mixed results. At this juncture, more and more AI researchers and universities striving to make path-breaking innovations are involving neuroscientists in their endeavor. One very important facet of empowering machines with decision-making caliber is data analytics. So, while neurologists give their inputs on evolving machines for finer tasks, it is the data that would feed the algorithm framework in order to lend perceptions. As machines are gradually starting to emulate and replicate human cognitions, some of the powerful algorithms, supported by strong neuron-based deep learning coupled with fast computing processors, are already driving autonomous cars, piloting drones, recognizing faces and voices, identifying words in speech, and so on. The dawn of the cognitive era is already upon us.

AI: Opening Up a World of Possibilities

AI is soon expected to identify people who are at risk of developing mental ailments in the future. It appears like amplified intelligence cannot only leverage AI but also take it to a whole new level of problem-solving, and, in due course, even innovation. We are generally skeptical of such possibilities that take control off our hands, and the fears are not entirely baseless. Stephen Hawking has repeatedly asserted that AI spells doom for mankind, and his apprehensions are very likely based on whether we are technologically ready to let machines take perceptually significant decisions which may have all-encompassing effects.

What I believe, however, is that ultimately, machines will evolve to the extent that they can be trusted decision-makers. But between now and then is a truly complex journey. For example, when AI is entrusted with security and surveillance tasks, we might come across a situation where a backpack-carrying individual in a public place can’t be classified with conviction as a tourist or a terrorist. So, while the initial scenario for the AI machine will be to determine the identity of the individual in question, the more important task would be to decide what to do with the information. This is where amplified intelligence becomes crucial.

AI without human intelligence and self learning is a framework which is likely to be found wanting in critical, make-or-break scenarios like the one I mentioned. A machine has to be fed with data, adequate yet not in excess, for it to grow a perception that can emulate that of a human being. The key components of perceptual computation are past experiences, contextual understanding, and situational awareness. Data analytics enlists the experiences which will feed the algorithm framework to AI. Contextual understanding is a complicated facet as machines don’t cardinally function as human beings. For instance, in spoken language, communication doesn’t solely rely on the message but also the tonality of it. Amplified intelligence will imbibe this understanding in machines by exposing it to a wide variety of possible options. Finally, and perhaps most importantly, situational awareness is all about basing a decision on circumstantial permutations and combinations. The aptness and accuracy of a decision depends wholly on the situation it is being taken in. The backbone of amplified intelligence is formed by these three elements and how seamlessly they are fused in order to provide AI the required impetus.

Paving a Sustainable Future

The advent of AI and machine intelligence has opened up avenues for individuals and industries. As AI advances, machines will function more independently. But it is the interdependence of man and machine that has to be harnessed in order to pave a future that is safer, easier, faster, more efficient, and less erroneous. Merging human brains with machines will usher in cognitive computing. And some of the market bigwigs are proactively endowing resources in order to harness this man-machine interdependence. Collective amplified intelligence is nothing but super intelligence that is born out of a marriage between AI and human intelligence, according to Microsoft’s Dr. Hsiao-Wuen Hon. I agree with him wholeheartedly. Look at the progress that has already been made. Today, AI is at the cusp of delivering complex duties in aerospace, healthcare, financial, and marketing domains, to just name a few. In other words, we are rapidly reaching a point where we are ready to treat machines as our sixth sense.

It makes me happy to be involved in this rapid transition of perceptual computation across spectrums. Providing services on data fusion, edge analytics, image processing, machine learning, and Internet of Things (IoT) are crucial means to attaining a smarter, self-reliant world that neutralizes problems on global scale. But like all good things about to happen to mankind, governance is an important aspect we must never discount in the long run. AI can and will enhance our talents and thought processes, but only when it adheres to guidelines that we lay down in order to defeat the demons that Mr. Hawking has been equivocal about in his widely documented apprehensions.