AI in 2023: The Application Layer Has Arrived - The Science of Forgetting: Why We're Already Losing Our Pandemic Memories

AI in 2023: The Application Layer Has Arrived - The Science of Forgetting: Why We're Already Losing Our Pandemic Memories

In the rapidly advancing world of artificial intelligence (AI), the application layer has finally arrived. AI is no longer just a concept or a distant possibility; it is now being integrated into various industries and sectors, transforming the way we work and live. When we look at the use cases of AI, they can be broadly grouped into two categories: creativity and productivity.

One of the most notable developments in AI is the emergence of AI-powered code generation tools like Copilot, which can generate up to 40% of the code in projects where it's installed. This percentage is expected to increase to 80% within the next five years. While we are still in the early stages of AI applications, it is difficult to predict what the killer apps will be. However, companies will need to find ways to build competitive moats to stay ahead of the competition. This could be achieved through network effects or iterative loops of user engagement and product refinement.

Interestingly, it is predicted that many of the best AI startups will be Software-as-a-Service (SaaS) companies. The go-to business models for AI applications are likely to follow the same patterns that have powered the tech industry over the past generation. Marketplaces, although more capital intensive to scale, often have powerful network effects that provide strong moats. On the other hand, SaaS models are highly desirable, but AI SaaS companies will need exceptional products to stand out in the crowded enterprise SaaS market.

There is also a need to address the concerns surrounding the capabilities of large language models. While some argue for banning these models, others like Wharton professor Ethan Mollick suggest that we need to adjust to these tools instead. As AI becomes more prevalent, it is essential for future generations to understand how to navigate this AI-driven world. Children growing up today will be immersed in a world teeming with AI, and it is crucial for them to develop the necessary skills and knowledge to thrive in this environment.

Moving on to a different topic, the science of forgetting sheds light on why we are already losing our pandemic memories. According to cognitive psychology professor Norman Brown, the default state of our memory is forgetting. Our brains have three interrelated phases for memory: encoding, consolidation, and retrieval. When we encounter new information, our brains encode it with changes in neurons, creating a physical memory trace known as an engram. However, much of this information is lost unless it is consolidated during sleep, making the memories more stable and long-term.