Gartner's latest research report points out that up to 30% of enterprises will abandon generative AI projects being tested by the end of 2025. This is not due to technical bottlenecks, but high costs. Reports show that deploying an AI project can cost millions of dollars, or even as much as $20 million, which is an unaffordable expenditure for many businesses. This article will conduct an in-depth analysis of the key data in the Gartner report, explore the reasons behind the high cost of generative AI projects, and how companies weigh costs and benefits to ultimately achieve effective application of AI technology.
The latest research from Gartner, a well-known analysis agency, shows that by the end of 2025, at least 30% of the generative AI businesses being tested will be ruthlessly abandoned by enterprises. This astonishing data was announced at the Gartner Data and Analysis Summit held in Sydney, Australia, and really gave the AI industry a blow.
Why is this happening? It turns out that the cost is too high! Gartner found that early adopters of generative AI are struggling with rising costs. Just imagine, deploying an AI project can cost anywhere from $5 million to $20 million, that’s a lot of money!
Picture source note: The picture is generated by AI, and the picture authorization service provider Midjourney
For example, if you want to design a custom generative AI model, such as fine-tuning the Llama model with industry-specific data, the initial investment will be $5 million to $6 million, and an additional $11,000 per year thereafter. If you want to develop a model from scratch, you have to prepare $20 million! Even a simple document search function will cost you $750,000 to start.
No wonder companies are all frowning! Gartner said that no matter how ambitious it is, generative AI requires companies to have a higher tolerance, because it is an indirect investment and cannot be expected to be effective immediately.
Gartner analyst Rita Sallam said: AI was so hot last year, and now executives can’t wait to see the returns. However, companies are finding it difficult to prove and realize the value of AI.
However, it is not all hopeless. Gartner surveyed 822 business leaders and found that after adopting AI, 15.8% of companies increased revenue, 15.2% of companies saved costs, and 22.6% of companies improved productivity. Although the number is not high, it at least proves that AI is still somewhat useful.
Therefore, business owners, if you want to play with AI, you have to do some calculations! To analyze the business value and total cost, you must not only look at the direct return, but also consider the future impact.
Highlights:
? Gartner predicts: 30% of generative AI projects will be abandoned by the end of 2025, mainly due to high costs.
? The cost of AI projects is staggering: custom models can cost $5-6 million, and development from scratch can cost as much as $20 million.
? The effects of AI applications are mixed: about 15% of companies have increased revenue or saved costs, and 22.6% have improved productivity.
All in all, generative AI has broad application prospects, but high costs and uncertain returns are still huge challenges facing enterprises. Enterprises need to carefully assess their own needs, plan their budgets reasonably, and pay close attention to technology development trends in order to move forward steadily in the AI wave and avoid blindly following the trend that leads to a waste of resources.