Artificial intelligence (AI) is rapidly transforming the world, marking a turning point in the history of technology. With 2025 just around the corner, new trends and challenges are emerging that promise to redefine the limits of AI and its impact on our society .
Francisco Herrera Triguero, Professor of Computer Science and Artificial Intelligence at the University of Granada, detailed in an opinion piece in The Conversation the most notable advances to come , such as the implementation of hybrid human-algorithm models, such as “doctor centaur” or “professor centaur”, which embody the delicate balance between machine autonomy and human control .
The explosion of AI-based science
AI has emerged as a crucial tool for addressing some of the greatest scientific challenges of our time.
Fields such as health, astronomy, neuroscience and climate change have witnessed how AI accelerates discoveries and generates innovative solutions.
An iconic example is AlphaFold, a tool that won the Nobel Prize in 2024 for having determined the three-dimensional structure of 200 million known proteins. This achievement represents a milestone in molecular biology and medicine , paving the way for the design of new drugs and treatments. Its open access will facilitate the widespread adoption of this technology by 2025, accelerating progress in biomedicine .
On the other hand, ClimateNet, which uses neural networks to analyse massive climate data, promises to be an indispensable tool for predicting extreme weather events with greater accuracy . In a context of global warming, the use of ClimateNet in 2025 could make a difference in planning and mitigating natural disasters.
Medical diagnoses and justice: high-risk scenarios
The fields of medicine and justice pose the greatest challenges for the application of AI due to their direct impact on people’s lives and rights . This is where the concept of the “centaur doctor” emerges, a hybrid model that combines the analytical capacity of machines with human intuition.This approach ensures that critical decision-making, such as medical diagnoses or court rulings, remains in human hands, with AI acting as a support that increases accuracy and reduces errors.
The development of these systems requires ensuring user trust through transparency and the protection of individual rights. AI must be a tool and not a substitute for human judgment .
Towards autonomous AI in everyday tasks
Big tech companies like OpenAI, Meta, Google, and Anthropic are working hard to develop autonomous AI agents that can make decisions on low-stakes tasks . These tools promise to simplify our daily lives, from buying a transport ticket to managing calendars or answering emails.
In this race, initiatives such as AgentGPT from OpenAI, Gemini 2.0 from Google and the Haiku and Sonnet versions from Anthropic stand out. This last model, capable of managing a computer autonomously, opens up new possibilities for automating personal and professional tasks, transforming the way we interact with technology .
Tiny language models: compact and accessible AI
2025 will also mark the expansion of Small Language Models (SLMs), compact and efficient systems that can be installed on mobile devices . These models promise to personalize the user experience, from smarter assistants to manage tasks to learning and diagnostic tools accessible to disadvantaged communities.
Their low cost and respect for privacy make them ideal for adoption in a wide range of applications , from education to healthcare.
Towards a new era of human-machine collaboration
In 2025, AI will continue to challenge us to think about how to integrate technology into our daily lives without losing sight of human control and ethics.
The path to trustworthy and responsible AI is full of opportunities and dilemmas , but with proper governance, we could be on the cusp of a new era of human-machine collaboration.
News reference:
https://theconversation.com/lo-que-nos-espera-de-la-inteligencia-artificial-en-2025-244353
