What algorithms mean to mankind

Tags: Op-ed
What algorithms mean to mankind
HELPING SCIENCE: Modern CT-scan machines can ‘memorise’ the positions of non-motile cells before a plate is removed for media exchange, and can then continue tracking them when the plate is replaced
Structures are fundamental to this universe and they are everywhere — right from the minutest particle to the cell, DNA, genome, and the biggest galaxy in the universe. These structures exist for a purpose and it is the task of mankind to discover the relationship between various structures (which sometimes have mutual interdependences and which lead to creation of systems, such as our solar system, or a molecule). In simple terms, structures can have sub-structures and super-structures depending upon where one sits and observes. A scientist has to identify not only the structure of a body, but also the effects of its interactions with other independent structures around it (such as bacteria and viruses with healthy cells, or to create new bodies). Knowledge of algorithms becomes indispensable for this purpose for it contains simulative, predictive, preventive, and prescriptive powers.

Most underlying principles of game-theory are related to quantitative approaches to human behaviour un­der varying conditions. The algorithms-based approach allows theorists to look deep into the possible outcomes. An algorithm is a statement of how a particular problem will be approached and resolved (kind of a flow chart). Algorithms help detect patterns, find errors in logic, provide predictive ability to events, and help humans design machines (including artificial life) which otherwise would be impossible. The science-driven knowledge era is mainly driven by machines run on algorithms, which themselves are an interfaced adaptation of ‘natural selection’ between humans and machines. There is an element of dynamism and continuous improvement leading to ever better solutions, and faster processing, and elimination of errors. Scientific pursuit at the nano- and astro-scales is now based on inductive, prescriptive, circumstantial, and deductive algorithms (which means direct physical evidence may not be there), rather than physically identifying the presence of an object as was done till the last century. Two prominent examples are the existence of black holes and Higgs-boson life particle. Larger advances are coming, however, not from improvement in technology, but from our awareness of what software and algorithms can do and are doing. Let me mention a few examples mostly coming from Nature.

Scientists have developed new algorithms to find fresh biomarkers — biological molecules and physiological characteristics — that are precursors to an oncoming health problem or disease. Their discoveries are set to transform the practice of medicine by giving doctors a more objective and quantifiable basis for clinical decision-making. National security agencies use similar algorithms routinely to track ‘noises’ at sensitive nodes for predicting and tracking terrorists-related activity and presence of sleeper cells. Modern CT-scan machines can ‘memorise’ the positions of non-motile cells before a plate is removed for media exchange, and can then continue tracking them when the plate is replaced, avoiding the image blurring that can disrupt statistical analysis.

One of the most beautiful aspects of our genetic code is its simplicity: three letters of DNA combine in 64 different ways, easily spelled out in a handy table, to encode the 20 standard amino acids that combine to form a protein. Predicting a stable structure arising from an amino acid sequence is a huge computational challenge since the change in structure can happen in microseconds.

We know that between DNA and proteins lies RNA. A team of researchers led by Benjamin Blencowe and Brendan Frey of the University of Toronto in Ontario, Canada used the massive data generated by a technology called alternative splicing to train a computer algorithm to predict the outcome of alternative spl­icing in mice. This complex model is an important technological advance, and this time there is no simple table — in its place are algorithms that combine more than 200 different features of DNA with predictions of RNA structure! At another level, using the power of crowd-sourcing in an experiment, Seth Cooper, David Baker and colleagues at University of Washington turned their ‘Rosetta structure-prediction’ algorithm into an online multiplayer game called Foldit, in which thousands of nonscientists competed and collaborated to produce a rich set of new algorithms and search str­ategies for protein structure refinement. Their work sh­ows that even computationally complex scientific problems can be effectively cr­owd-sourced using interactive multiplayer games.

The previous era of hu­man-machine interaction was mostly for elimination of drudgery; mankind is now poised where his knowledge and understanding of phenomena is undergoing a major transformation with the coming of self-correcting generation of machines collaborating with humans. The possibilities are immense; for example, just one faculty has the power to deliver an online course across the globe to millions.


(The writer is a professor of strategy and corporate governance, IIM-Lucknow)


  • Finance ministry must stand its ground on lower EPF interest rate

    In quick succession, yet another controversy is brewing with regard to affairs of the employees provident fund organisation (EPFO).


Stay informed on our latest news!


Amita Sharma

Sanskrit: a victim of academic schizophrenia

J Robert Oppenheimer, the father of the atomic bomb and ...

Zehra Naqvi

God save the child

Childhood is supposed to be the best phase in life. ...


William D. Green

Chairman & CEO, Accenture