Across the world there has been a growing concern over the increasing intelligence of machines. These artificially intelligent (AI) machines are posing serious threat to jobs. According to some international studies, around 800 million jobs are likely to disappear within next two decades. More alarming is that jobs in developing countries are highly vulnerable to these machines. More than 50 per cent of industrial jobs in these countries are susceptible to automation. Particularly export oriented manufacturing jobs are being targeted by advanced countries; they want to reduce the outflow of existing plants and take back their already offshored production plants by deploying AI as substitute to labour. Already some of the countries like Bangladesh have started to experience job loss in the industrial sector to these machines. In the labour-intensive readymade garments and textile industry of Bangladesh, a single state of the art intelligent machine is killing up to 12 jobs.
Such a massive job killing threat by this new machine capability raises pertinent questions. Why are we in the race of developing machines to kill jobs, consequentially lowering human value in production? Is it creating an enabling force for the human race to progress?
With the growing machine's ability to sense and perceive uncertain situations like road condition, and reason out how to act, often question arises whether machine intelligence is going to be superior to human or natural intelligence. Are we in the race of building machines superior to human beings? As humans create machine intelligence, how can artificial intelligence be superior to natural intelligence becomes an obvious answer. Intelligence is a suitcase world having many connotations. If it is the ability of imagination, machine intelligence is far inferior to human capability-and it is likely to remain so. But once the issue of execution of machine intelligence shows up, machines perform better than human beings. For example, some of the tasks of flying passenger Jets require gathering of intelligence by processing real-time data produced by thousands of sensors to understand the situation and act accordingly. Over the decades, machine capability is being meticulously developed to perform those tasks. And it has been found that machine intelligence in such a situation is far superior to that of pilots sitting in the cockpit. For this reason, over the last 50 years, there has been steady growth of aviation safety records, while the complexities of passenger jets were continuously increasing. Intelligence in performing most of the production tasks starting from driving a car to understand the appetite level of fish are to be delegated to machines. For example, successful delegation of roles of driving from human to machine has the potential to reduce economic loss and human causality caused by road accidents substantially.
Machine intelligence is growing as a key capability to remain competitive in this globally connected economy. For example, flow of minerals from Russia in 1980s created significant price pressure on mining companies in USA, Canada, South Africa and Australia, among others. Some of these countries came up with the idea of human-free mining to remain in the business of mineral production. Due to the non-availability of autonomous machines, many mining companies formed research partnerships with universities and research laboratories to develop underling technologies and innovate autonomous solutions. After more than two decades of relentless work, they are reaching the dream of "Mining 24 Hours a Day with Robots." These gigantic machines are guided by precision GPS (geographical positioning system), while looking out for obstacles using radar and laser sensors. In terms of economics, driverless trucks have proven to be roughly 15 per cent cheaper to run than vehicles with humans behind the wheel.
Machine intelligence is not about AI fantasy of science fiction writers, rather it's about getting the job done better at less cost by adding human-like intelligence to machines.
By observing recent development, someone may perceive that experts in artificial intelligent are in the race of implementing science fiction writers' imaginations. For example, Elon Musk seems to be inspired by science fiction stories, particularly the writings of Lain M. Banks. Elon Musk has become restless entrepreneur, embarking on a series of ventures like reusable rockets, electric cars, giant batteries, vacuum trains and underground roads; his latest firm, Neuralink, hopes one day to build a working brain-machine interface (BMI). But careful observation indicates that the major underlying force of AI progress is to implement basic principals of management theories, as descried by Prof. Clayton Christensen of HBS (Harvard Business School). He suggested that innovators should focus on tools in getting the job done better. For pursuing incremental innovations for making existing tools better, or conceiving next generation ones around new technology core to disrupt existing tools, innovators should focus on the purpose to get the job done better at less cost in order to turn innovative ideas into profitable revenue flow.
AI has been new technology core in supporting disruptive innovations in a massive scale. Artificially intelligent machines are autonomous systems increasingly taking human like roles. Some recommend music or movies; others recommend sentences in court. Even more advanced AI systems will increasingly control vehicles on crowded city streets. But once they fail to perform as per expectation, and cause loss, who will share the liability? Thus ethics, liability and law should be worked out further to accommodate new capabilities of machines.
There is no doubt that machines are likely to be superior to human in performing repetitive production tasks-for offering higher quality at lower cost. To become competitive, we need to take smart steps to innovate and adopt intelligent machines rather than wasting time on the debate whether we need intelligent machines or not in densely populated countries, or whether machines are going to be superior to human beings.
M Rokonuzzaman PhD is academic, researcher and activist on technology, innovation and policy.
