Will The Master Algorithm Ever Be Created?

We hear the conspiracy and rumor surrounding the master algorithm, but is there some truth in this? Read more here…
The Master Algorithm might sound like a science fiction novel, but it’s a concept gaining real traction in technology circles. Currently a principle rather than a practical application, it is being hailed as a global game-changer. It’s claimed it might have the power to automate numerous jobs, while minimizing human error.

What Is The Master Algorithm?

The master algorithm rose to prominence following the publication of a 2015 book by Pedro Domingos. An expert in machine learning research, Domingos theorized computers will one day replicate the thought processes of the human brain.  Instead of innumerable algorithms governing individual aspects of our lives – from curated content suggestions to production line quality checking – one overarching algorithm may organize our lives and automate thousands of mundane tasks. In theory, society will improve immeasurably, as tasks like transportation become increasingly automated.
Domingos based his book on decades of research by fellow computer scientist Geoff Hinton, who co-created a learning algorithm christened backpropagation. This deep learning algorithm is used by many of the world’s leading technology companies including Google and Microsoft, though typical applications include interpreting images and determining which adverts to display in social media timelines.

Is The Master Algorithm Possible?

Although the master algorithm is a widely accepted concept, there are divergent schools of thought about how it will be realized. Some people believe human knowledge is stored in the connections between neurons in our brains, and reverse engineering this could bring machines to the brink of consciousness. Machine learning advocates disagree, and a group known as symbolists reckon a learning algorithm could become capable of combining rules to reach conclusions when undertaking deductions.
Others argue evolutionary biology perfectly encapsulates the master algorithm at work. A fourth school of thought suggests we should look to mathematics. Specifically, advocates point to Bayes’s theorem regarding the impact new evidence can have on an established hypothesis. These probability distributions are expected to drive autonomous vehicles, both literally and metaphorically. Bayesian theory already underpins various established technologies like email spam filters.
There’s serious money to make in developing the master algorithm. This is why Google bought British backpropagation startup DeepMind for £400 million in 2014.  But what are these companies investing in, and how do they see this theoretical concept changing our lives?

The Future Of Machine Learning

In short, machine learning has the power to make decisions for us. The ultimate extension of curated formulas was first used by Amazon in the 1990s when the company started recommending products we might like. The more the master algorithm knows about us, the better it can guide and support us.
Anyone who’s watched Terminator 2 or read dystopian sci-fi novels appreciates that machine learning has a fundamental flaw. Sentience might enable computers to question their own status or seek to improve it. Isaac Asimov devised three laws of robotics to prevent machines rising up against us. But there are clear and present dangers involved in giving computers the capacity to think for themselves. Even though, research into the master algorithm will likely speed up in the years ahead, for however long the perceived benefits outweigh the potential risks.

To create a platform redundant and secure enough to protect your data, visit VPS.NET today.