What AI in 2024 Means for Us in 2025

Bottom Line Up Front

AI breakthroughs in 2024—cheaper models, multimodal capabilities, and broader access—are reshaping what’s possible in 2025. For electric utilities and IT organizations, this means new opportunities in efficiency, customer service, and sustainability, but also new responsibilities around governance, training, and data quality. Success will depend not just on adopting AI, but on deploying it thoughtfully, with the right infrastructure, collaboration, and readiness across the organization.

Introduction

Last week, I shared a post titled “AI in 2024: A Year in Review,” highlighting the groundbreaking advancements in AI technology. This week, we are narrowing the focus to explore what these changes mean specifically for people working in the utility sector and the IT organizations that support it.

Artificial Intelligence in 2024 has been nothing short of transformative, marking a turning point in accessibility, innovation, and practical application. Groundbreaking advancements in AI models, pricing, and hardware requirements have paved the way for democratization across industries.

The original article highlights the remarkable strides in AI technology, including cheaper and more efficient models, expanded functionalities like multimodal AI, and the increasing role of open-source platforms. These shifts have broad implications for industries reliant on data and innovation, such as utilities.

Three Key Highlights

GPT-4-class models running on laptops: High-performing models are now deployable without expensive infrastructure. Multimodal AI becoming mainstream: Integration across text, audio, video, and images expands use cases dramatically. LLM token cost reduction: A steep drop in costs opens doors to wider adoption and experimentation.

What This Means

The developments in AI signify a rapid shift toward greater inclusivity and efficiency in technology. Breakthroughs like extended context lengths and real-time multimodal interactions mean that AI is no longer confined to tech giants with deep pockets. Instead, small to mid-sized companies and industries like utilities can now leverage these tools effectively.

The broader availability of AI also raises the bar for innovation, as organizations can experiment with cutting-edge capabilities without being hindered by high costs. However, it also necessitates thoughtful deployment to avoid pitfalls like low-quality content and ethical concerns. Additionally, the quality of data used for AI applications becomes increasingly important. Poor-quality data can lead to unreliable outputs, hindering the effectiveness of even the most advanced models. Utilities must focus on ensuring their datasets are accurate, complete, and up-to-date.

What This Means for the Electric Utility Industry

AI’s evolution presents both challenges and opportunities for electric utilities. Real-time decision-making powered by enhanced inference models and multimodal AI can revolutionize operations. Operational Efficiency: AI-driven insights can streamline operations from grid management and maintenance to IT to HR. Customer Experience: Enhanced multimodal AI tools improve customer service through personalized and responsive interactions. Sustainability: As models become cheaper and more energy-efficient, this aligns with our commitments to the environment. Data Reliability: Ensuring high-quality data sources is critical to harnessing AI’s full potential and avoiding flawed decisions. Everyone will play a part in this critical corner stone.

What This Means for SCE’s IT Organization

The IT faces a dual challenge: integrating these advancements while maintaining operational stability. On one hand, we must support the deployment of scalable, efficient AI platforms that produce the largest business value, while addressing uneven adoption rates in our organizations. This means focusing on platforms over projects and isolated solutions.

Data quality will play a pivotal role here. IT must implement and enforce strong data governance practices to maintain the integrity and relevance of the information used by AI platforms. Regular audits and validations of AI Models, Prompts, and Datasets will help in reducing the risk of unreliable outputs and ensuring the effectiveness of AI-driven initiatives.

IT organizations will need to:

• Train employees to utilize AI Platforms effectively.
• Develop standards for evaluating, testing, and accepting AI outputs and platforms.
• Align AI strategies with overarching business goals.
• Implement stringent data quality and governance measures.

What I Should Do to Get Ready for This Change

• Educate Yourself: Stay updated on AI advancements, particularly those impacting the electric utility industry.
• Experiment with Tools: Explore the AI tools that have been deployed and test their applicability in your workflows.
• Enhance Skills: Upskill in areas like prompt engineering and multimodal AI applications.
• Collaborate Across Teams: Work closely with operations and strategy teams to align the work you want to do with AI initiatives.
• Advocate for AI Governance: Be aware of the current AI policies and regulations while supporting and promoting these policies to ensure ethical and effective AI deployment.
• Prioritize Data Quality: Ensure datasets used for AI applications are accurate, clean, and well-maintained by ensuring that all data is complete, accurate, and aligned to corporate data models and their data stewards.

Conclusion

The rapid progression of AI in 2024 signals a new era of opportunity and responsibility. For electric utilities, this transformation brings the promise of efficiency, sustainability, and enhanced customer experiences. However, thoughtful implementation and ongoing learning are essential to maximize these benefits. At the same time, maintaining high data quality is critical to ensuring that AI delivers reliable and actionable insights.

Call to Action

How are you preparing yourself and your organization for the AI revolution? Share your strategies and insights in the comments below!