In the fast-paced lanes of commercial drones and artificial intelligence, the speed of innovation often leaves many companies trailing behind. The pace is relentless. Yet, it is simple. You must incessantly experiment and adapt if you want to survive. This adaptation must be based on objective observations of truth. And the truth is that advances in machine learning have made the need for software developers in their current form, a soon to be relic.
This week Erik Brynjolfsson from MIT and Tom Mitchell of Carnegie Mellon, released their research in the Science Magazine, entitled, "What Can Machine Learning Do: Workforce Implications". The authors describe a shift or digital transformation, on par with the advances of electricity becoming widely available, primarily driven by machine learning. The authors specifically elaborate:
Until recently, creating a new computer program involved a labor-intensive process of manual coding. But this expensive process is increasingly being...replaced by a more automated process of running an existing machine learning algorithm on appropriate training data.
It is entirely possible, right now, for a trained computer system to be more accurate and capable than those that are programmed by humans using manual coding methods. This advancement will make software creation more accessible to the masses. With the current trend of open-source software development likely to continue to grow rapidly, many more human ideas and scientific breakthroughs will result.
However, this will not mean software developers will become obsolete, but rather their roles will dramatically change. Future developers will spend more time designing new human-inspired programmatic systems while deciphering other systems that are computer generated. On the other hand, "trainers" will take on the prior role of the developer. These trainers will feed machines experiences, ones which will help them to learn various functions and tasks independent of their trainer. Additionally, the trainers who are the best will be those who respect their creations, and seek to work alongside them, rather than enslave them. Such wisdom will encourage an equilibrium between man and machine, preventing unnecessary miscalculations and future disasters.
Finally, the good news. In spite of machine learning advancements, many human-sourced inventions, new discoveries, products and services, and industries will be created almost overnight. As a result, many opportunities will manifest for those who are able to afford the necessary computing power required to feed algorithms mountains of data. In the near future, massive amounts of computing power will become more affordable, allowing the majority of human minds to tap into the digital mind, creating infinite digital and neural connections beyond what exists. In spite of the debate concerning the impact of automation, a worthy strategy may consist of running ahead of the technology, as opposed to being overwhelmed by it. The developer must untether themselves beyond maintenance and input of code, and explore becoming trainers of autonomous systems, rather than face the prospects of doing otherwise.
Ean Mikale, J.D., is a Dronetrepreneur, current participant of the NVIDIA Inception Program for AI Startups, IBM Global Entrepreneur, member of the National Small Business Association Leadership and Technology Councils, and has been featured in the London-based Global Banking and Finance Review Magazine for his work in social finance. Follow him on Linkedin, Instagram, and Facebook: @eanmikale