AI could replace data analysts and improve analysis
Categories: economy, governance
What? MIT researchers designed the Data Science Machine, an Artificial Intelligence (AI) system that automatically searches datasets for predictive patterns. In a 2015 data analysis competition against 906 human teams, the AI system beat two-thirds (615) of them, achieving a pattern-identification accuracy rate just slightly under (96%) that of the winning team. However the AI system took only hours to achieve what required months of analysis by the human teams.
So what? New rapid AI data analysis systems could dramatically transform the current data analysis labour market, replacing the majority of current data analysis jobs by taking the most time consuming "grunt work" analysis. In turn, AI could enable data scientists to engage in higher-level thinking by speeding the rate of data analysis, and since it is difficult for AI to analyze long-tail data, long-tail data analysts could become more relevant in the future. Furthermore, commercially available, user-friendly AI data analysis software could increasingly enable non-experts to participate in data analysis themselves, especially as AI makes data analysis cheaper. Lastly, by providing easier and faster ways to analyze data, AI systems could make mountains of data more understandable and useable by governments and companies. This could improve governance and corporate efficiency in service delivery and marketing.
Source: Innovations Enterprise - Does Machine Learning Spell the End of the Data Scientist?