2025 will mark the beginning of a self-optimization phase, cooperation between AI agents and a focus on energy efficiency. This is the main conclusion that they underline from Nutanix, in their predictions on the future of artificial intelligence (AI) in the business field.
According to Induprakas Keri, senior vice president and general manager of Hybrid Multicloud in Nutanix, next year he will witness the emergence of a new type of software that will learn and improve autonomously, without the need for active programming, in addition, “it will improve the user experience and productivity without active programming ».
To take advantage of these advantages, companies will need “employees with adequate skills, processes and technologies,” said Dutta, director of AI in Nutanix. Well, new reasoning models, including open source, will provide additional capacities systems. Technologies such as “memory in memory”, which uses RAM as storage, will popularize to support these innovations.
AI in 2025
Energy efficiency will be a key challenge, since inference, rather than the training of AI models, will consume large amounts of energy. Dutta points out that this is mainly due, “to the scaling of the computer skills necessary for inference. Therefore, companies will have to rethink their expense on infrastructure and energy. ” To mitigate these costs, companies must use AI to identify inefficiencies and automate intensive labor processes.
The AI will enter 2025 in a new phase: that of the selfmeter, the collaboration between agents and the search for energy efficiency
To ensure that investments and operational costs do not endanger the success of AI projects, companies must use this technology to, “identify inefficiencies and automate intensive processes of their labor. Undoubtedly, those responsible for you must finance the AI with the AI aid itself, ”adds Tobi Knaup, General Manager Cloud Native of Nutanix Nutanix.
Following this, from the company, Rajiv Ramasswami, its president and CEO affirms that they aim to improve the productivity of their developers by 25% through the use of generative. This technology will be used to generate code for unit tests and other functions, thus optimizing internal processes.
