Stargate – A $ 500 billion joint venture between Oracle, Openai and SoftBank. The objective is to build data centers and infrastructure necessary for the development of AI. Unrelated to this announcement, many conversations and publications exist on the AI generation, AI agents and agency workflows. Many SaaS organizations launch agents workflows within their product suite. There are many predictions on the role of agents, some suggest that agents will soon represent individuals in discussions, leading to scenarios such as: “Ask your agent to speak to my agent, and we can discuss after the convergence of our agents. AI assistants are present on each application and on each platform. There are many models to use on LLM, many open source and many new ones that we can consider every week/days. There is so much enthusiasm and AI in the news, on the covers of notable magazines and the advertisements of major technology suppliers.
However, for most of the organizations I work with, adoption is not generalized. There are productivity “microrer” mainly motivated by the gains generated by repeatable specific tasks. But 10 hours per employee and per month do not change the situation for anyone. This does not mean that specific organizational roles are not significantly affected. Meta CEO, Zuckerberg, told the Podcastor Joe Rogan how the company sought to replace “intermediate level engineers” with AI. Organizations also realize that it is expensive. Are yields proportional to investments in all commitments? Finally, hallucinations remain a real risk. Even Apple recently canceled its news aggregator on AI.
Does the absence of generalized adoption at this stage raises the following question: the adoption of generative AI in our daily professional and personal life is inevitable? Or is it another technology that some positions within an organization will adopt because of its significant impact, but which is unlikely to be generalized? The answer is that we expect large -scale effects of this technology. Throughout history, the combination of different technologies has been crucial for significant advances, and generative AI also follows this model.
Technological convergence refers to the integration of two or more independent technologies. This convergence can accelerate these independent technologies or sometimes come together to train new ones. Some notable examples of this phenomenon include mobile and internet phones. Mobile phones were initially designed for voice communication, but with the introduction of smartphones, they have become multifunctional devices. You can call, send SMS, navigate the Internet, take photos, take commercial transactions and more. This convergence favored the adoption of mobile phones and accelerated the development and use of Internet services. Another example, on a smaller scale, is that of smart watches, like Apple Watch. You can use the intelligent watch to follow the time, follow your activity, follow key health settings, communicate and interact with various applications on a portable device.
What are the components that support convergence for generative AI?
Cloud“The data movement to the cloud is continuous and scalable, reinforced by technological progress and powerful IT capacities. This has happened for some time, and it is appropriate that we are in a situation where most of the organizations are in the maturity stages of their journey.
Data“We are capturing more data, which is also more easily accessible. We can consume data in many forms, structured and not structured. In addition, the fact that generative AI can be used to interact with all these data in their current forms is an essential motivation for using this technology.
Digital literacy—The organizations have an increasingly good digital culture. In addition, these fundamental models have democratized AI, from data scientists seated in an ivory tower conceiving algorithms that the rest of the organization does not understand until they put the power of these models in the hands of all . Getting involved with Chatgpt does not require any particular knowledge of the functioning of AI.
Multimodal—The that started as large models of language (LLM) which include, generate and manipulate natural language has now transformed into image, audio and video processing. These are therefore technically large multimodal models which are more natural in the way we make decisions.
Based on my conversations with managers from different organizations, here is my illuminated hypothesis on the impact. The generative AI will fully automate 20 % of daily tasks, leaving more time for creativity by eliminating the trivial. After all, we will adopt everything that suits us. This will further improve efficiency and productivity in 60 % of additional tasks. The impact could be considerable here – 25 to 50 %. Meanwhile, the 20 % of remaining tasks will evolve and will always require human surveillance, combining technology and human touch. All this to say that the question of a large adoption is perhaps premature. We must understand that this is a trip and start by where adoption will bring the greatest value. Write code, for example. And as we identify other sources of value, we will make sure to design effective and economically viable solutions that work.
Even if challenges remain, the future prospects for generative AI are promising. We are ready to change our way of creating, communicating and committing. However, we must balance this innovation with a responsible approach that will allow us to exploit the potential of the generative AI, by creating effective and fair tools for all users.
