As AI becomes increasingly embedded in our daily lives, industry leaders and experts are forecasting a transformative 2025.
Smaller, purpose-driven models
Grant Shipley, Senior Director of AI at Red Hat, predicts a shift away from valuing AI models by their sizeable parameter counts. "2025 will be the year when we stop using the number of parameters that models have as a metric to indicate the value of a model," he said. Instead, AI will focus on specific applications. Developers will move towards chaining together smaller models in a manner akin to microservices in software development. This modular, task-based approach is likely to facilitate more efficient and bespoke applications suited to particular needs.
Open-source leading the way
Bill Higgins, VP of Watsonx Platform Engineering and Open Innovation at IBM, expects open-source AI models will grow in popularity in 2025. "Despite mounting pressure, many enterprises are still struggling to show measurable returns on their AI investments—and the high licensing fees of proprietary models is a major factor. In 2025, open-source AI solutions will emerge as a dominant force in closing this gap," he explains. Alongside the affordability of open-source AI models comes transparency and increased customisation potential, making them ideal for multi-cloud environments.
Augmenting human expertise
Jonathan Siddharth, CEO of Turing, sees 2025 as the year when AI systems will learn from human expertise at scale. "The key advancement will come from teaching AI not just what to do, but how to approach problems with the logical reasoning that coding naturally cultivates," he says. Competitiveness, particularly in industries like finance and healthcare, will hinge on mastering this integration of human expertise with AI.
Behavioural psychology will catch up
Understanding the interplay between human behaviour and AI systems is at the forefront of predictions for Niklas Mortensen, Chief Design Officer at Designit. "Agentic AI marks a more flexible and creative era for AI in 2025," concludes Dominic Wellington, Enterprise Architect at SnapLogic. However, such systems require robust data integration, as siloed information risks undermining their reliability.
From hype to reality
Jason Schern, Field CTO of Cognite, predicts that 2025 will be remembered as the year when truly transformative, validated generative AI solutions emerge. "Through the fog of AI for AI’s sake noise, singular examples of truly transformative embedding of Gen AI into actual workflows will stand out," predicts Schern.
Deepfakes and crisis of trust
Sophisticated generative AI threatens the authenticity of images, videos, and information, warns Siggi Stefnisson, Cyber Safety CTO at Gen. "Even experts may not be able to tell what’s authentic," he warns. Combating this crisis requires robust digital credentials for verifying authenticity and promoting trust in increasingly blurred digital realities.
2025: Foundational shifts in the AI landscape
As multiple predictions converge, it’s clear that foundational shifts are on the horizon. The experts that contributed to this year’s industry predictions highlight smarter applications, stronger integration with human expertise, closer alignment with sustainability goals, and heightened security. However, many also foresee significant ethical challenges. 2025 represents a crucial year: a transition from the initial excitement of AI proliferation to mature and measured adoption that promises value and a more nuanced understanding of its impact.
Conclusion
The predictions outlined above demonstrate the significant changes that AI is likely to undergo in 2025. As the technology becomes increasingly embedded in our daily lives, it’s essential to acknowledge the challenges and opportunities that lie ahead. By understanding the shifts in AI’s landscape, we can harness its potential to drive real value and make a positive impact.
Frequently Asked Questions
Q: What changes can we expect in AI in 2025?
A: Smaller, purpose-driven models, open-source AI solutions, and a shift towards more strategic AI investments are some of the key predictions for 2025.
Q: What are the ethical implications of AI in 2025?
A: The potential for deepfakes and the crisis of trust are significant concerns, and it’s crucial to develop robust digital credentials for verifying authenticity and promoting trust in digital realities.
Q: How will AI be integrated with human expertise in 2025?
A: AI will learn from human expertise at scale, and competitiveness will hinge on mastering this integration, particularly in industries like finance and healthcare.
