Generative AI generates tricky choices for managers
The remarkable capabilities of generative artificial intelligence (AI) are evident when you experience it firsthand. However, these remarkable abilities also pose challenges for managers. It becomes more difficult to determine how to utilize a new technology when it has the potential to impact numerous activities. Its adoption relies not only on the capabilities of machines but also on unpredictable human factors. Additionally, generative AI may have unexpected flaws.
Various studies consistently highlight the potential of large language models (LLMs), such as ChatGPT, to enhance various aspects. LLMs can save time by generating meeting summaries, analyzing data, or drafting press releases. They can improve customer service. However, they cannot assemble IKEA bookshelves, just like humans.
Furthermore, AI can contribute to innovation. Karan Girotra from Cornell University and his co-authors compared the idea-generating capabilities of the latest version of ChatGPT with those of students from prestigious American universities. While a single human can generate around five ideas in 15 minutes, equipping the human with GPt-4 increases the number to 200. Importantly, these ideas are of better quality, as indicated by purchase-intent surveys for new product ideas. Such possibilities can overwhelm managers, as having the ability to do everything often leads to doing nothing.
2023-11-27 16:04:14
Original from www.economist.com
rnrn
