As we enter 2025, AI continues to reshape industries and redefine the boundaries of technological innovation. The rapid evolution of AI has ushered in a new era of digital transformation, with generative AI leading the charge. Investment in this groundbreaking technology is projected by PWC to contribute a staggering US$15.7tn to the global economy in 2030, underscoring its pivotal role in shaping the future of business and society.
From revolutionising healthcare with AI-driven diagnostics to transforming financial services with intelligent algorithms, the impact of AI is far-reaching and profound.
However, this technological leap forward also brings new challenges, particularly in the realms of ethics, regulation, and data security.
In light of this, AI Magazine presents a comprehensive list of the top 10 AI trends of 2024 to understand the AI developments of this year and gain a glimpse of what could be in store for 2025.
10. AI governance and ethics
AI governance and ethics have become paramount concerns as artificial intelligence technologies continue to reshape our world. The UNESCO Recommendation on the Ethics of Artificial Intelligence, adopted by 193 countries in 2021, stands as a cornerstone in this field. This comprehensive framework emphasises four core values: respect for human rights, fostering just societies, environmental well-being, and diversity and inclusiveness. It interprets AI broadly, recognising the rapid pace of technological change and the need for future-proof policies.
The recommendation outlines eleven key areas for policy actions, moving beyond high-level principles to practical strategies. These include sustainability, fairness, technical safety, accountability, explainability, and data stewardship. To support the implementation of these principles, UNESCO has established the Global AI Ethics and Governance Observatory. This platform serves as a vital resource for policymakers, regulators, academics, and civil society, offering country profiles, expert insights, and good practices in AI ethics and governance. The observatory's AI Ethics and Governance Lab gathers contributions, research, and toolkits to address pressing challenges in responsible AI innovation, standards, and institutional capacities.
9. AI talent demand
The demand for AI talent has skyrocketed in recent years, with job postings incorporating AI skills as a requirement surging by 130% over the last two years. This trend is particularly evident in the field of generative AI, where roles have seen an unprecedented 1,300% increase in demand. The surge reflects the critical importance of AI expertise in today's rapidly evolving tech landscape, with machine learning and data science roles growing by 80% and 250%, respectively.
This dramatic shift in the job market has led to a significant skills shortage, with the International Data Corporation predicting that over 90% of companies will face IT skills shortages by 2026. In response, organisations are increasingly looking beyond traditional talent pools, with a 31% increase in demand for specialised skills across borders. This global approach to talent acquisition is reshaping recruitment strategies, as companies seek to build teams with the specialised skills needed to develop and implement AI-driven solutions. The trend towards skills-based hiring is further emphasised by the fact that 80% of companies are now prioritising becoming skills-based organisations, with a particular focus on AI and related technologies.
7. Open source AI
Open source AI has emerged as a powerful force in the realm of technology, fostering collaboration and innovation on a global scale. This approach to AI development allows researchers, developers and organisations to freely access, modify and distribute AI systems, including datasets, code and model parameters. The transparency inherent in open source AI has led to significant advancements in various fields, from natural language processing to computer vision.
One of the most notable examples of open source AI is OpenCV, the Open Source Computer Vision Library. Launched in 1999 as part of an Intel research project, OpenCV has since become a cornerstone in the computer vision community. Its extensive library of algorithms supports a wide array of applications, including facial recognition, object detection and motion tracking. The longevity and versatility of OpenCV underscore the enduring impact of open source AI initiatives. Another prominent player in the open source AI landscape is OpenNN, a neural network library that excels in predictive analytics. Developed by Artelnics, OpenNN has found applications in customer intelligence and industry-specific analytics, demonstrating the practical value of open source AI in business contexts.
7. Agentic AI
Agentic AI represents a significant leap forward in artificial intelligence, offering a level of autonomy and decision-making capability that sets it apart from traditional AI systems. This advanced form of AI is designed to operate independently, making decisions and adapting to changing environments with minimal human intervention. Unlike its predecessors, agentic AI can analyse complex situations, formulate strategies and execute multi-step tasks to achieve predefined goals. At its core, agentic AI employs sophisticated algorithms that enable it to understand and interpret natural language instructions, develop action plans and adjust its approach based on real-time feedback.
This adaptability allows it to optimise workflows, enhance customer service experiences and streamline operations across various industries. For instance, in financial services, agentic AI excels at fraud detection by analysing vast amounts of data and autonomously initiating resolution workflows when issues are identified. Its ability to learn from interactions and improve over time positions agentic AI as a powerful tool for enterprises seeking to boost efficiency and maintain a competitive edge in an increasingly dynamic marketplace.
6. AI in scientific research
AI is rapidly transforming scientific research across numerous disciplines, offering unprecedented capabilities in data analysis, hypothesis generation, and experimental design. In genomics, AI algorithms can swiftly analyse genetic sequences to identify disease markers, whilst in drug discovery, machine learning models predict the efficacy of new compounds with remarkable accuracy. These advancements have led to significant time and cost reductions in research processes, with genomics seeing a 60% acceleration in research and a 50% reduction in costs.
The integration of AI in scientific research extends beyond data analysis. Natural Language Processing (NLP) systems are revolutionising how researchers manage the ever-growing volume of scientific literature. These AI-powered tools can summarise vast collections of research papers, extracting key findings and methodologies, thus enabling scientists to stay abreast of developments in their field without the need to read through entire papers. Moreover, AI is fostering interdisciplinary research by integrating data and methodologies from different scientific fields, leading to innovative solutions for complex problems such as climate change and its socioeconomic impacts.
5. Customisable AI
Customisable AI models have emerged as a powerful tool for businesses seeking to harness the full potential of AI. Unlike generic solutions, these bespoke models are tailored to address specific challenges and optimise performance in targeted areas. By aligning closely with a company's unique data sets and objectives, customised AI models deliver more accurate insights and enhanced operational efficiency.
The shift towards customised AI solutions is driven by the recognition that off-the-shelf tools often fall short of meeting the nuanced needs of each organisation. These tailored models can be fine-tuned for specific tasks, seamlessly integrated with existing systems, and designed to process unique data sets that reflect the intricacies of each business. For instance, a retail company might leverage a custom AI model to analyse consumer behaviour and optimise inventory management, whilst a healthcare provider could develop a bespoke solution to predict patient outcomes and refine treatment plans. This level of personalisation not only ensures higher accuracy but also provides a significant competitive advantage in today's data-driven landscape.
4. Small Language Models
Small Language Models (SLMs) have emerged as a formidable force in the realm of AI, offering a compelling alternative to their larger counterparts. These compact AI models, with parameter counts ranging from a few million to a few billion, are designed to process, understand and generate natural language content with remarkable efficiency. SLMs employ advanced techniques such as model compression, knowledge distillation and transfer learning to condense the broad capabilities of larger models into a more focused, domain-specific toolset.
One of the key advantages of SLMs is their ability to operate in resource-constrained environments, making them ideal for edge devices and mobile applications. For instance, Granite 3.0 1B-A400M and Granite 3.0 3B-A800M boast total parameter counts of 1 billion and 3 billion respectively, yet their active parameter counts at inference are a mere 400 million and 800 million.
This reduction in parameters translates to decreased processing times and lower latency, allowing SLMs to respond swiftly to user queries. Moreover, their compact nature makes them more environmentally sustainable and cost-effective, as they consume less energy and require fewer computational resources compared to their larger counterparts.
3. AI for workplace productivity
AI is revolutionising workplace productivity, offering a plethora of tools designed to streamline tasks and enhance efficiency across various sectors. From automating mundane administrative duties to providing sophisticated data analysis, AI-powered solutions are transforming how businesses operate in the digital age.
One notable example is Notion AI, an intelligent feature integrated into the popular workspace platform. This innovative tool assists users in generating content, answering queries and autofilling tables within their Notion pages. By leveraging AI capabilities, Notion enables teams to brainstorm more effectively, draft documents with greater speed and manage projects with increased precision.
Similarly, Otter.ai has emerged as a game-changer in the realm of meeting productivity. This AI-powered assistant offers real-time transcription, speaker identification and audio-text synchronisation, allowing professionals to focus on discussions rather than note-taking. These advancements not only save time but also enhance the quality of work output, demonstrating the tangible benefits of AI in boosting workplace efficiency.
2. Gen AI democratisation
Gen AI has emerged as a powerful force in democratising innovation, placing sophisticated tools in the hands of individuals across various sectors. This technological leap forward has significantly lowered the barriers to entry for creative and analytical endeavours, enabling a broader range of users to harness AI's capabilities without extensive technical expertise.
The impact of this democratisation is particularly evident in the business world, where Gartner predicts that by 2026, more than 80% of enterprises will be using Gen AI APIs and models in production environments, a stark increase from less than 5% in 2023.
This shift signifies a move towards more inclusive and efficient business practices, with AI tools being integrated into everyday applications. For instance, sales teams can now utilise AI-powered chatbots and proposal generators to streamline their operations, enhancing productivity without the need for extensive data science teams.
However, this widespread adoption is not without its challenges. A Boston Consulting Group study revealed that over half of executives surveyed are actively discouraging the use of generative AI due to a lack of understanding, whilst 37% are in a state of experimentation without proper policies or controls in place.
1. Multimodal AI
Multimodal AI represents a significant leap forward in AI, combining various data types such as text, images, audio, and video to create a more comprehensive understanding of the world. Unlike traditional unimodal systems that rely on a single data type, multimodal AI integrates diverse inputs, enabling it to perform complex tasks with greater accuracy and contextual awareness.
At its core, multimodal AI comprises three key components: the input module, fusion module, and output module. The input module consists of multiple unimodal neural networks, each handling a different data type. The fusion module then processes and aligns the information from each modality, employing techniques such as early fusion to concatenate raw data. Finally, the output module delivers results that are often richer and more nuanced than those produced by unimodal systems
Explore the latest edition of AI Magazine and be part of the conversation at our global conference series, Tech & AI LIVE.
Discover all our upcoming events and secure your tickets today.
AI Magazine is a BizClik brand
