The Intersection of AI and Software: From LLMs to Aggregating Demand

The Intersection of AI and Software: From LLMs to Aggregating Demand

In the world of technology, two major trends have emerged that are reshaping industries and transforming our daily lives. These trends include the advancements in Artificial Intelligence (AI) and the disruptive power of software. Both AI and software have revolutionized the way we approach tasks, solve problems, and interact with the world around us. In this article, we will explore the common points between these two domains and how they intersect to create new possibilities and opportunities.

One notable development in the field of AI is the concept of Low Rank Adaptation (LoRA). LoRA is a method used to adapt large, pre-trained models to specific tasks or domains without significant retraining. The idea behind LoRA is to have a smaller module that contains enough domain-specific information, which can be appended to the larger model. This allows for quick adaptability without altering the large model's size or the need for extensive retraining.

LoRA leverages the mathematical concept of low rank approximation to create a smaller, adaptable module. By containing specific characteristics or information, this module can be integrated into larger models to customize them towards a particular task. This approach not only saves computational resources but also allows for the injection of domain-specific knowledge into a larger model, granting it the ability to understand and process information within a specific field without significant alteration to the core model.

On the other hand, the rise of software has brought about a paradigm shift in how industries operate and how we consume goods and services. As technology continues to advance, the world is moving from a state of scarcity to a state of abundance. Software has played a crucial role in increasing access to scarce resources by enabling us to do more with less. This has allowed orders of magnitude more people to enjoy something for a fraction of the cost.

The concept of aggregation theory, as proposed by Ben Thompson, further highlights the transformative power of software. In a pre-internet world, profits were captured by controlling supply. However, in a post-internet world, profits are captured by aggregating demand. Software has enabled the unbundling and rebuilding of industries, allowing for the creation of new forms of value and the emergence of new scarcities.

For example, in the music industry, songs have been unbundled from CDs and then rebundled into playlists. Similarly, articles have been unbundled from newspapers and then rebundled into social media feeds. The role of curation becomes crucial in a world of abundance, where attention and loyalty are the new scarce resources.