Guest post by Chwee Kan Chua, IDC
While several artificial intelligence (AI) winters have come and gone in the last couple of decades, today’s AI hype is an actual reality due to the perfect storm of digital data, high performance compute, and new machine learning techniques and software. The AI hurricane has gone from Class 1 to 3 in the last 24 months and accelerated AI development at a pace never seen before.
The positive impact of AI can be clearly seen – from autonomous vehicles to disease diagnostics.
We are witnessing new real-world business values being delivered with AI, but production-ready AI solutions are only streaming in and in some areas, these are barely a trickling stream. Based on recent IDC research, the three key challenges faced by businesses getting AI into production environments are:
- Shortage of data scientists or AI specialists, with some organizations not having any. In fact, 73% of organizations surveyed indicated they have none.
- Overwhelming effort needed in data preparation and exploration, largely due to data quality problems.
- Lack of both data science and domain expertise convergence to generate any meaningful insights.
For AI to not stall into another winter, a paradigm shift has to occur to address the three abovementioned issues. The shortage of data scientists cannot be fully addressed by educational and training programs, especially finding the perfect balance of data science and domain expertise. In a nutshell, organizations without AI expertise will eventually be outcompeted by their AI-enabled peers.
Organizations need to change their perception and look beyond, especially regarding the concept of a “citizen data scientist”.
Since business users are already well versed in their respective domains, why not enable them with data science capabilities as well? Think of how intelligent GPS routing software allows us to get to our destinations quickly, accurately, and with a good forecast time based on traffic and weather conditions, all without needing to know how to read a road map. Intelligent GPS routing/mapping has changed how everyone gets to a destination, and automated machine learning will do the same for all business users.
This Copernicus-type change is already happening today. These automated machine learning platforms are transforming organizations by empowering business users and domain experts with extensive self- service machine learning capabilities that require little or no data science technical skills
DataRobot is a pioneer in this revolution, having already delivered an AI platform that helps break down the final barriers for business users to have access to machine learning. Everyone in the organization, from domain experts to business analysts, regardless of industry, can now extract insights and predictive models from their digital data quickly, with clearly explained reasons, rationale, and transparency. The business value cannot be ignored – processes can be optimized, fine-tuned, and deployed rapidly as new data becomes available, transforming how an organization can innovate and compete continuously.
In short, DataRobot has removed the technicality of AI modeling and facilitated automated machine learning with best-of-breed algorithms.
One of the best ways for organizations to embark on this new revolution is to start with a well-versed business case with data-rich processes. Business users will be able to see immediate actionable insights from the automated machine learning results. Organizations should no longer be stuck in a mindset that they need to hire an army of technical AI experts; business users empowered as citizen data scientists will become ubiquitous with this AI paradigm shift.
About the Author: