In this era of rapid technological advancement, artificial intelligence (AI) has permeated every aspect of our lives, from intelligent assistants to chatbots, all contributing to the convenience of our daily routines.
However, as more and more enterprises integrate AI technology into their operations, many are unaware of the hidden risks behind this trend. Recently, two analysts from Gartner shared their profound insights at a seminar in Australia, which are worth our careful consideration.
Image source: Generated by AI, authorized by Midjourney
Firstly, Mesaglio candidly stated: "Wasting money on generative AI is a piece of cake!" She also mentioned the shock and confusion when people first saw their cloud computing bills, suggesting that many companies might be about to experience a similar situation. Mesaglio warned: "AI costs could have errors ranging from 500% to 1000%!" Imagine being enthusiastic about trying AI, only to find the bill several times higher than expected, the feeling is understandable.
Why does this happen? One reason is the price hikes from suppliers, and another hidden "black hole" is the negligence of companies in managing cloud resources. Often, companies do not need AI to handle simple queries, which is frequently overlooked. Think about asking AI a question and receiving a complex answer of thousands of words when all you wanted was a one-sentence reply. This naturally leads to increased costs.
Additionally, it's important to note that processing unstructured data (like random documents, emails, etc.) with AI, although it can provide more information and better results, is also a "money-burning" path. Centralized data management might save money but faces various challenges in IT management. It's like cooking: choosing to add more spices for better taste or simplifying ingredients to control costs? It's a vexing choice.
Speaking of which, we must mention another "potential issue"—the "productivity leak" caused by successful AI applications. According to Gartner's research, employees using AI save an average of 43 minutes per day, which sounds great, right?
But the question is, how do people use this extra time? Mesaglio humorously mentioned: "I'd use that time to have a coffee." This indicates that many might choose leisure over working harder, ultimately diminishing the benefits brought by AI.
When explaining AI's productivity enhancement, the analysts used a vivid example. Imagine two lawyers, a newbie and a seasoned veteran. The newbie will quickly increase efficiency with AI's help, while the veteran might not benefit as they already possess the knowledge. This phenomenon within companies might lead to dissatisfaction, especially when the performance gap between new and old employees narrows, potentially marginalizing the veterans.
So, how should enterprises handle AI, this "double-edged sword"? Analysts suggest that when considering AI applications, enterprises should first clarify whether the pace is "steady" or "accelerated." For industries not yet affected by AI transformation, a steady approach might be more prudent; for companies aiming to establish themselves in the market with AI, they need to pick up the pace. However, enterprises should understand that every organization has its rhythm; there's no need to rush blindly. Choosing the most suitable development path is the key.
AI is a sharp tool; used properly, it can bring significant value to enterprises, but it also requires caution to avoid various troubles caused by improper use.
Key Points:
🌟 When using AI, enterprises might face astonishing cost errors of 500% to 1000%!
💰 Improper use of AI for simple issues leads to unnecessary cost waste.
⏳ Successful AI might trigger "productivity leaks," with employees not necessarily using the saved time to increase efficiency.