Is it ‘Human vs Machine’ or ‘Human and Machine’?
This question about whether it is ‘Human vs Machine’ or ‘Human and Machine’ has been widely debated for the last few years. Although AI and algorithms have been around for a few decades, it is only in the last few years that they have suddenly gained prominence across almost aspects of business and life.
For me, the current corona virus pandemic reopened this debate, given how AI and related technologies are being used in unique and innovative ways in a life and death situation. Whether it is the use of facial recognition technologies in China, smart helmets, drones or Alibaba’s Covid-19 diagnostic tool, AI is in the forefront of all these initiatives. Companies like Google’s DeepMind are using AI to understand the proteins making up the virus so that treatments can be developed. In coming days, we will see more and more use cases of AI being used to fight this battle. Even otherwise, AI has gotten wide acceptance in our day to day life, whether it is Amazon, where we buy the recommended books or in Google Maps where we blindly follow the suggested route or Netflix where the next binge-watching show is ‘recommended’ to us.
It’s clear that the discussion is no longer about how Artificial Intelligence or machine learning are affecting our lives and our decisions, but rather about what the future will look like: thought leaders are divided on this topic. Some predict that algorithms will take over most of human workforce; while others believe that the machines will be complimentary and humans will still be in-charge.
However, from the earlier days of IBM Deep Blue defeating the world chess champion Kasparov to Google’s AlphaGo defeating Chinese Go Master, if we look at success it is not only of machine but also of humans who could train the machine and optimize it to desired results. Google’s search engine or Facebook’s facial recognition software, they all use elements of human assessments to fine tune and improve accuracy. Even when we look at the last few months, AI has definitely helped whether in prevention of the virus spread or accelerate search and testing of new potential drugs but ultimately, it is the human element - doctors, nurses, scientists, law enforcement who are making the real difference; with the help of technology, of course!
Coming back to my original question - there are very strong arguments on both sides but if you ask my personal opinion, I drift towards the latter – Human and Machine are complimentary. However, the very nature of AI means huge investments in technology and capabilities to collect and process vast amounts of data, learning from the data and thus, likely to be controlled by a few organizations or governments. This would mean that those who control this intelligence would have immense power to influence everyday decisions in our lives.
The possibilities related to AI are real and this belief has only been solidified in the last few months across various life altering use cases. However, we must not forget that the use of AI needs to be within ethical boundaries and clear policies on AI governance need to be agreed and defined to further accelerate the influence of this technology on solving some of the most pressing problems of the world we live in.
Interesting and insightful, Ajit! I concur that human oversight will remain critical for AI deployments. Institutions will need to ensure stricter security and governance controls to secure AI-driven automation. Strong security, compliance, and governance will create trust in AI and it cannot be achieved without humans supervision - at least for now.
Human is first because machine can be made again
Interesting thought.... In a three-entity equation of Human vs Virus vs Machines, Machines would surely be on the side of humans. For humans, Virus are more lethal than machines.
Answer to your original question depends on who controls who.
The AI is able to detect correlations based on a multitude of parameters. It is important that we learn to understand these results. Therefore, technologies are needed to explain the results of AI #explainableai.