The Reddit Story: 6 New Theories About AI

The Reddit Story: 6 New Theories About AI

Over the course of its 17-year history, Reddit has grown to become a top 10 site in the world, with over half a billion monthly active users. But what makes Reddit so successful? How did it become the "front page of the internet"? In this article, we will explore the Reddit story and connect it with six new theories about AI.

Theory 1: Fine-tuned models win battles, foundational models win wars

In the world of AI, there are two types of models: foundational models and fine-tuned models. Fine-tuned models are specifically tailored for narrow use cases, making them more cost-effective for certain tasks. On the other hand, foundational models are designed to perform broad tasks well. While fine-tuned models may have an edge in the short term, foundational models have the potential for step-changes in performance. The key is to deploy fine-tuned models for narrow use cases to lower the cost of prompt completion.

Theory 2: Long-term model differentiation comes from data-generating use cases

Data loops, where AI providers build feedback mechanisms into their products to retrain the model, play a crucial role in long-term model differentiation. Startups that can capture the model to output to retrain model loop will be able to build a specialized winner. This requires owning both the model provider and the endpoint solution. By utilizing this strategy, companies can continuously improve their models and stay ahead of the competition.

Theory 3: Open source makes AI startups into consulting shops, not SaaS companies

Open source has disrupted the AI market by putting downward pricing pressure on model providers that sell access to their models via API. When competing with free, companies often have to compromise on pricing. To overcome this challenge, some companies have taken equity stakes in promising startups to circumnavigate pricing pressure. Additionally, open source allows companies to integrate generative AI into their products, providing a white glove service while protecting their intellectual property.

Theory 4: Most endpoints compete on GTM, not AI

When it comes to selling AI services, the competitive advantage lies in the go-to-market (GTM) strategy rather than the AI technology itself. For startups competing on the basis of software as a service (SaaS), distribution and product capabilities are key. Major software providers are expected to integrate generative AI into their products in the near future, further emphasizing the importance of GTM strategy.

Theory 5: AI will not disrupt the creator economy, it will only amplify existing power law dynamics

AI tools can help creators produce better content faster, allowing them to build a critical mass of fans. However, the distribution of content will still follow existing power law dynamics, where a minority of creators capture the majority of revenue. AI will amplify this dynamic, making it even more important for creators to utilize AI tools effectively in order to stand out in the crowded digital media landscape.

Theory 6: Invisible AI will be the most valuable deployment of AI

Invisible AI refers to companies that are powered by AI but never explicitly mention it. These companies utilize AI to enable new modalities of digital interactions, breaking traditional computing models. By combining search capabilities with generative AI, companies can create innovative products that provide a seamless user experience.

In conclusion, the Reddit story and the theories about AI discussed in this article provide valuable insights into the future of AI. To stay ahead in the AI race, companies should focus on fine-tuning models, leveraging data-generating use cases, embracing open source, prioritizing GTM strategy, recognizing the power law dynamics of the creator economy, and exploring the potential of invisible AI.

Actionable Advice:

1. Invest in fine-tuning models for narrow use cases to lower costs and improve performance.
2. Build feedback mechanisms into your AI products to continuously retrain and improve your models.
3. Focus on GTM strategy and distribution capabilities to gain a competitive advantage in the AI market.

By understanding these theories and applying the actionable advice, companies can navigate the evolving AI landscape and unlock new opportunities for growth and success.