Guest Post: AI in 2025: A Combinatorial Explosion of Possibilities, but NOT AGI

Happy 2025 year! Next week, we will be reporting from CES – the largest tech event in the world. Our goal is to observe how AI is being implemented in real life. Usually, we discuss a lot of theoretical questions, but now it’s time to see what the market is getting its hold on and learn the use cases in practice.

As a good preface to that on-the-ground exploration, we'd like to share this article, written specially for Turing Post by one of our frequent contributors – Charles Fadel. Charles is a global education thought leader and author, futurist and inventor; founder and chairman of the Center for Curriculum Redesign; chair of the education committee at BIAC/OECD; member of the OECD AI Experts Group; co-author of Education for the Age of AI (2024).

His article discusses the rapid evolution of Generative AI (GenAI) by 2025, highlighting its transition from theoretical to engineering-focused development. It examines various technological advancements, including specialized datasets, improved training methods, multimodal models, and AI agents. Charles argues that despite challenges, these innovations will lead to unpredictable, transformative capabilities, but not AGI. Let's dive in!

Introduction   

Waymo, the robotaxi service, provides an interesting analogy: although it functions “up to Level 4” (while requiring expensive behind-the-scenes human intervention), it is now being deployed by the tens of thousands. It does not have to be perfect to be useful.

GenAI is now in the same situation: its problems are well-known (biases, lack of world model, hallucinations/confabulations, “jagged” capabilities, etc.) but it is rapidly transitioning from a science phase (and the hype of “scaling at all costs”) to an engineering phase where a significant number of development vectors are at play, and reviewed herein.

The combinatorial possibilities offered by all such developments preclude any hope of forecasting what capabilities might arise within 2025 (“Combinatorial” because they will interact with each other in unforeseeable ways). This list of 23 development vectors not only underscores the combinatorial complexity at play but also pinpoints the areas with the highest transformative potential (bold in italic). Whether you’re an engineer, researcher, or enthusiast, these insights offer a roadmap to understanding GenAI's pivotal transition into its engineering phase.

The views expressed by the author do not necessarily reflect the editorial stance of this publication.