The Thinking Algorithm
What is it that frightens you? Is it the fear of the unknown? Or its the fear of accepting your destiny?Are you the Neo to Morpheus, not ready to accept your role? Are you Luke to Darth Vader when you have not realized your destiny? Or are you Einstein or Oppenheimer or Fleming – just waiting for lightening to strike?
What if you had an aide?
Your faithful C-3PO or WALL-E who could point you in the right direction, but still could not do what you could?
That’s the thinking algorithm! The also that anticipates your quest!
Even though most of us intermingle with algorithms every day, they are rarely the focus of public parlance. We are not cognizant of algorithms, thanks to their invisible nature and the lack of common knowhow of what an algorithm is and what it does. An algorithm is similar to a recipe where a set of instructions when followed produces the desired end result and can be easily replicated again and again.
So what exactly constitutes the ‘Thinking Algorithm’?
We just described that an algorithm is a set of instructions giving a desired output. Now imagine if the same algorithm could think and modify the instructions and outputs in real-time. This is what forms the basis of all cognitive, recursive and predictive analytics prophecies. ‘Intuition’ plays a big role in the thinking process. Prior to questioning our intuitions, we need to comprehend that what our mind’s x-ray vision is looking at is an intuition or some form of a ‘thinking algorithm’ as viewed from the inside.
Before we can question our intuitions, we need to recognize that our mind is looking at an intuition which is some form of a perceptive algorithm, as seen from the inside—rather than a direct observation of ‘How Things Really Are’.
People hold on to their intuitions, because they cannot visualize their intuitions as their thinking cognitive algorithms happen to look from the inside. The result being a paradox of sorts because the cognitive algorithm is in direct conflict with the real world representation of the way things are. One key learning here is the algorithm is also part of a feedback loop which is constantly changing the way the thinking algorithm thinks!
Researchers at the University of Tubingen in Germany are working on a paper titled ‘A Neural Algorithm in Artistic Style’– where an artificial intelligence system based on a Deep Neural Network creates artistic images of high perceptual quality. The system uses neural representations to separate and recombine content and style of arbitrary images, providing a neural algorithm for the creation of artistic images(Citation: arXiv:1508.06576) maybe someday leading to the possibility of transforming your ordinary run of the mill photo to a Da Vinci, Picasso, Van Gogh or Munch.
Enter Messrs. Meta-Algorithm (or machine learning): Consumers and brands from all walks of life will have specific needs and requirements, which means the thinking algorithm needs to constantly tweak its cognitive intuitions to meet the end user’s constantly changing needs. This is where the meta-algorithm sets in. It programs the algorithmic mind to make tweaks to the set of input instructions incorporating the changing query and specificities of the end user. What comes out is a completely modified new result with continuous iterations where none of the original ingredients of the algorithm are present. This democratized process where the original recipe evolves into several new forms is what today’s world of customization, predictive analytics and machine learning entails. The key takeaway here being the new output of the algorithm has been created by the tweaked formula in the meta-algorithms interspersed and mixed with the influences and choreographies’ of the various algorithm beta testers.
Today, popular wearable devices like Fitbit and Jawbone can create a potential revolution in bringing drugs to market faster. The biopharmaceutical industry is a living example where the thinking algorithm is at play. We at BioXcel, have developed our own targeted interactive recursive mapping algorithm (ApolloTRIM) to produce high quality clinical leads from raw unstructured data from millions of data sources. The algorithm helps derive relationships between disease pathophysiologies, molecular targets so as to identify intervention points for treating the disease and further identify potential mechanisms of such curative agents. We have already used ApolloTRIM to discover novel connections between genes and cellular and molecular mechanisms relevant to a disease – for example in unraveling the various molecular targets relevant to immuno-oncology and CNS orphan diseases.
We are part of an era where we all ‘live in the lab’ and various algorithmic outputs are being experimented on us resulting in game changing breakthroughs in almost every field. Like Trinity to Neo – It is the question that brings you here? What is it that we seek? Why are we special? What are we trying to do? While these maybe philosophical questions – it abounds in practicality on modern medicine!!!
What are we trying to do? Solving important medical problems? Make profits for our ventures? Make our healthcare system meaningful? All of the above?
Do you want a GPS to guide you through this maze?
It’s time we become “Alice” enough to follow the rabbit and look forward to the day when an algorithm will create an algorithm and yours truly will simply think and have an algorithm put his next blog together.