The landscape of AI in media creation

The landscape of AI in media creation has undergone significant advancements by 2024, particularly in the realms of generative AI, agent-based models, large language models (LLMs), and AI-powered production technologies in film and television.

Generative AI in Enterprise Software: Generative AI is experiencing a substantial market surge, with predictions indicating that generative AI chip sales could reach over $50 billion, and software revenue might see a $10 billion increase in 2024. This technology is set to reshape multiple industries, with nearly all enterprise software companies integrating Generative AI into some of their offerings. These integrations are expected to enhance productivity, creativity, and enterprise engagement with ecosystems. High-powered Generative AI chips will accelerate analysis, and companies training Generative AI models on their own datasets will derive specific insights and more significant value​​.

Agent-Based Models and Generative Multimedia: In 2024, niche applications like agent-based models and generative multimedia are evolving from experimental to more practical and applied stages. Agent-based models, though not expected to fully take off in 2024, will show more convincing use cases, particularly for tedious processes like insurance claims. In the field of video and audio, generative models are finding niches where their shortcomings, such as lack of photorealism, are less noticeable. Skilled creators are using AI video in interesting ways, and generative music models are likely to be featured in major productions like games​​.

Advancements and Limitations of Large Language Models (LLMs): The capabilities and limitations of LLMs are becoming clearer. While these models have shown remarkable capabilities, 2024 is witnessing a shift in understanding their limitations and “emergent” capabilities. The industry is beginning to see diminishing returns on increasing parameter counts, leading to a consideration of more efficient models. There's a growing inclination towards using a mixture of smaller, more specific models (possibly multimodal), which could be almost as effective as a single large model but much easier to update​​.

AI in Film and Television Production: AI-powered production technologies are poised to restructure the film and television industries, offering dramatic efficiency gains and creative possibilities. Within the next 3-5 years, AI support is expected to be integrated into all digital workflows, especially enhancing virtual production. This integration will unlock resources and creative capacity for individuals and productions. However, it also implies that some jobs may start disappearing or undergoing significant changes. Additionally, streaming TV business models are shifting from subscription models towards targeted advertising​​​​​​.

These advancements suggest a transformative period in media creation, where AI not only augments but also challenges traditional processes and business models. The implications for creative workflows, job roles, and monetization strategies are profound, marking a pivotal moment in the intersection of technology and media.