AI may not live up to the hype

AI may not live up to the hype

https://www.bloomberg.com/news/articles/2024-10-02/ai-can-only-do-5-of-jobs-says-mit-economist-who-fears-crash

I’ve been seeing an increase in similar articles around the economic ‘hype curve’ of AL. In this article Daron Acemoglu, a professor at the Massachusetts Institute of Technology calculates only a small percentage of all jobs—a mere 5%—are vulnerable to be taken over, or at least heavily aided, by AI over the next decade.

This aligns with my own experience. Over the past two years, I’ve been collaborating with companies on internal AI initiatives, focusing primarily on augmentation and automation. While there have been improvements in certain processes, the efficiency gains are far from the levels hyped by AI proponents.

This does not detract from the potential of AI. Many of my clients’ roadmaps include generative AI implementation. However, similar to early trials with augmentative AI, the main challenges are not the AI technology itself, but the quality of the data used and the efforts required to change the habits and behaviours of colleagues using these features. The use of virtual assistants and chatbots within organizations remains low, partly due to insufficient coaching on how to ask the right prompts or questions of AI models (I also appreciate the user experience of many chatbots is still poor).

It’s been the same issue for over 20 years, whether it’s enhanced search engines, shiny productivity and collaboration tools, or the early introduction of chatbots. Until the content and data are right and there’s a commitment to changing people’s behaviors, these initiatives will struggle to reach their full potential.