The Financial Express
Swasti Lankabangla Swasti Lankabangla

Artificial intelligence: Bangladesh\'s predicament

| Updated: October 20, 2017 23:27:46

Artificial intelligence: Bangladesh\'s predicament

We no longer have to wait with baited breath or scrounge every nook and corner for any artificial intelligence (AI) scent: it is already in the works, though in fits and starts. For some it is in the smartphone, for others in a variety of everyday Internet programs. This should not be surprising as the centennial anniversary of the robot, which coincides with our 50th independence anniversary, 2021, approaches. Karol Capek, who created something like it, called that contraption 'an artificial person'. In his native Czech language, 'robot' means 'forced labour'. At a time when our own ready-made garment workers have often been called just that, it is important to reassess the robot's role in a country like ours. How it stands today as the symbol and epitome of the highest scientific knowledge is a cruel play on those forced workers, whom they are designed to replace.
Who would have believed we would have to wait this long for both AI and robot dissemination to take place? George B. Devol built the first formal robot, Unimate (for 'universal automation'), in bucolic Kentucky in 1954, then sold it to General Motors in 1961 to manoeuvre hot metals from die-casting machines in its Trenton, New Jersey factory. Almost simultaneously (in 1956, precisely), the AI term was coined by John McCarthy, Marvin Minsky, Allen Newell, Claude Shannon, and Herbert Simon in a Dartmouth College scholarly exchange. But only with the Fourth Industrial Revolution unfolding slowly before our very eyes today have they grabbed attention, not to mention mileage.
Few others have spelled out the AI trajectory's contemporary relevance as methodically as Anand Rao. In "A strategist's guide to artificial intelligence" (in the Strategic + Business Summer 2017 issue), he disaggregated AI practices and evidences into three meaningful, self-explanatory, albeit, overlapping clusters: technologies that assist or augment human intelligence, and those that remain autonomous of humans. He applies them to five automatic arenas of AI application today, especially in industrialised countries, but quickly dissipating into their less developed counterparts: healthcare; arts and communications; personal finance; mobility; science and environment; and management.
His findings, obviously relevant to a typically developed country, stand to reason: assisted intelligence can already be found in medical imaging (that is, the healthcare arena), and automated financial processing (personal finance); augmented intelligence in precision planting advice (science and environment); and autonomous intelligence in bomb-disposal robots (science and environment). By mid-century he sees these three clusters respectively advancing enough to produce doctorless hospitals, creative arts engines, and self-driving vehicles.
We, in technologically colourless developing countries, need not feel left out of the loop: just as communications antennas dot our paddy-fields and a smartphone has become as common as a lungi or a sari, so too will all three types of AI contraptions seep into our own everyday lives, albeit, less prosaically as the lungi and sari, more selectively among these people, and become even more extracting (rather than complementary or supplementary), meaning, it will tax our pocket more than alternative contraptions (but we will still, under social pressure, purchase it). Many of the causal factors may already be familiar today.
In the first place, given the irreversible patron-client/owner-producer/manager-worker schisms in society, any technology that helps the patron, owner, and manager to lessen dependence on the client, producer, or worker, the more likely it will be on the front-burner of changes. Imagine a ready-made garment (RMG) factory-owner being able to displace the producer with a machine: not only would manager-worker tensions evaporate automatically, but output might also multiply. With autonomous intelligence technology as his dream contraption, until he has it at his fingertips, he might experiment more conceptually with assisted intelligence technology (perhaps with some equipment enhancing worker productivity), then augmented intelligence technology (the machine adding a production twist that the worker cannot manage in order to boost production and sales).
Bangladeshi entrepreneurs have not as yet shown the stomach to be daring with technological improvements: a bulk of our factories run on old machines, meaning, our research-and-development (R&D) plate is rather thin or empty. Yet, the competitive instinct might still lead him (or, increasingly, her) into smuggling in a new technology from abroad, thus accidentally sparking a production revolution through imitation or adaptation. This is the second trigger, built upon the simple and straightforward logic that Dhaka boasts of BMW automobiles when we have neither the roads to run them on with satisfaction nor the professional technicians to repair them when needed. Even if for the demonstration effect it may generate, we will somehow have the latest technological mark-up somewhere in our starkly unequal society.
A final trigger may be more pragmatic and automatic: once we see our exports or competitive advantage slip, we will jump to the nearest upgraded technology available, just to ensure the cookies keep coming in. In other words, unless push turns to shove, our sticky-feet will not be going anywhere.
Whether by the back- or front-door, all three types of AI technologies cannot but become a part of the Dhaka landscape (if not scattered elsewhere across Bangladesh). With so many hospitals, thus fulfilling the critical competitive criterion, we have already plunged into medical imaging: by mid-century, some robot in some posh neighbourhoods will be administering our medical prescriptions, even the most banal medical treatments, like circumcising boys.
Though Dhaka's road network may simply thwart any experiment with self-driving cars, even by mid-century, our cumbersome anti-terror outfits, which already fiddle with precision planting devices, suggest automated machine translation, or other mechanical reproduction of human activities by mid-century is very much possible. Crude bomb-disposing robots flourishing by mid-century would spark the decentralisation of corporate functions, another arena where job-growth will slowly yield to machines (watch out, MBA-supplying universities). An immediate future corporate owner residing in Purbachal (perhaps in the world's tallest or second-tallest building, if our dream architects have their way), can direct his/her manager in Uttara to coordinate field-operations in Tongi/Gazipur/Munshiganj without stepping outside of his/her home/office. The machines connecting them may be pipe dreams no longer. The more the resort to such kind of augmented intelligence technology, the sooner will be our rendezvous with an autonomous intelligence destiny.
Even before the invasion of robots, we will be in AI centre-field; and once we are, the socio-political questions will spiral: how many can be operationalised before sparking labour unrest. Since labour unrest can push technological progress and profit-making back to the medieval age, it is not an issue to be taken too lightly in Bangladesh. In short, without incrementally enhancing education and job opportunities, any AI resort could be playing with fire. How adept we are at fire-fighting may be but a facetious indicator if the Fourth Industrial Revolution and artificial intelligence is for us or not.
Dr. Imtiaz A. Hussain is Professor & Head of the newly-built Department of Global Studies & Governance at Independent University, Bangladesh.


sepnil-desktop-ad-the-financial-express sepnil-mobile-ad-the-financial-express

Share if you like