Only 11% of companies surveyed say they get any financial benefits from AI implementation. What’s holding the others back?
When the pandemic accelerated organizations’ digital transformation, the first technology that gained higher acceptance was artificial intelligence (AI). Companies rushed to understand how it works and what aspects of their business and people processes can be adapted to include AI. The overall belief is that it will raise productivity and performance as well.
A new studyOpens a new window by MIT Sloan and BCG titled Expanding AI’s Impact With Organizational Learning shares a crucial finding. Only 11% of companies are getting significant ROI from AI technologies. The big difference between these companies and those who don’t get financial returns is that the former focus on organizational learning, not just machine learning. They are open to changing their processes to enable higher levels of learning.
Responses of 3,000 executives to 100 survey questions were analyzed to find that identifying, investing in, and implementing AI is not enough to make it a valuable game-changer in organizational performance. Employees and leaders need to learn how to learn from AI and transfer their own learning back to AI. The cycle of learning must be continuous and collective for it to make a significant difference.
When AI Use Results in Financial Benefits
As per the study, some organizations can double their opportunities to have high financial benefits from AI with one step – allowing humans and AI to learn from each other mutually.
The roles of machines and humans are not defined. They are fluid and change based on the situation. Strategy is also rooted in digital data as well as human experience as per the report.
Leaders follow five models of AI implementation, based on the report findings, with two models being most successful:
- AI recommends and humans implement
- AI generates insights and humans use them in the decision-making process
While companies are at the threshold of adapting to AI’s role, a structured process must be established for humans to learn how AI works and when human intervention is required to add value. This understanding of how human and machine learning can be facilitated will drive the success of any AI solution.
HR Technology News: HR Leaders Need New Competencies: Josh Bersin Academy Report
Bringing Human and Machine Learning Together
The lines between human learning and machine learning are blurring. But therein lies the reason for its importance. Each model has its pros and cons, but it is crucial to understand the two models in which leaders have seen more success.
AI the recommender: AI recommends, human decides
It is evident that machines are equipped with enough data and past insights for it to offer recommendations. This indicates that it can integrate multiple facets needed for problem-solving. But when companies use this approach, the final decision-making is left to humans. Human beings can gauge the steps that AI took to arrive at the recommendation and add their own assumptions or business context to evaluate whether that recommendation is implementable.
AI the illuminator: AI generates insights, humans base decisions on them
In this model, the AI does the entire thinking through in terms of data and what it is pointing toward. It does not make recommendations, but it does give insights from the data that can enable humans to arrive at a logical solution. The insights form the basis of further decision-making, which can result in multiple options for problem-solving as against weighing the pros and cons of a specific recommendation.
HR Technology News: Lack of Leadership Commitment in Upskilling: IBM Institute for Business Value Report
Companies are in the nascent stages of their understanding of AI and its application. Some believe that it will result in job loss. But when applied in the right context, AI can result in job elevation, evolution, and redesign.