As we dive deeper into 2024, the relentless pace of AI innovation shows no signs of slowing. With each passing day, new breakthroughs and advancements are reshaping the technological landscape. But what can we expect from AI by the year’s end?

IBM Master Inventor Martin Keen discusses nine AI trends that are set to emerge throughout the year:

1. Reality Check: The era of hyperbolic expectations surrounding generative AI is giving way to a more nuanced understanding of its capabilities. Integrating AI tools into existing workflows is becoming the norm, enhancing and complementing traditional methods rather than replacing them entirely.
2. Multimodal AI: The convergence of natural language processing and computer vision is unlocking new possibilities for AI applications. Models like OpenAI’s GPT-4v and Google Gemini are leading the charge, enabling seamless interaction between text and visual data.
3. Smaller Models: While massive AI models have dominated headlines, the trend is shifting towards smaller, more efficient models. These models offer comparable performance with significantly fewer parameters, reducing both training and inference costs.
4. GPU and Cloud Costs: The demand for GPUs to power AI infrastructure is driving up cloud costs. As organizations seek more cost-effective solutions, the focus is shifting toward optimizing model size and resource usage.
5. Model Optimization: Techniques like quantization and low-rank adaptation are gaining traction, allowing for more efficient use of AI resources. Expect to see further innovations in model optimization throughout the year.
6. Custom Local Models: Open-source models enable organizations to develop custom AI solutions tailored to their specific needs. Keeping AI training and inference local helps protect proprietary data and sensitive information.
7. Virtual Agents: Beyond traditional chatbots, virtual agents automate tasks and streamline workflows. From making reservations to completing checklists, these agents are becoming indispensable tools in the workplace.
8. Regulation: With the rise of AI comes increased scrutiny and regulation. The European Union’s Artificial Intelligence Act and ongoing debates over copyright issues are just the beginning of what promises to be a complex regulatory landscape.
9. Shadow AI: Employees’ unofficial use of AI in the workplace poses challenges around security, privacy, and compliance. As AI capabilities continue to grow, so too do the risks associated with unauthorized usage.

As we navigate the ever-evolving world of AI, it’s essential to remain vigilant and adaptable. The future promises exciting possibilities and new challenges that must be met carefully and carefully considered. As we embark on this journey, let’s keep our eyes open to the opportunities and pitfalls that lie ahead.