The Future of AI: Contrarian Theses and the Evolution of Search Engines

The Future of AI: Contrarian Theses and the Evolution of Search Engines

Introduction:
Artificial Intelligence (AI) has become a disruptive force in various industries, with its potential to revolutionize workflows and drive innovation. However, there are several contrarian theses surrounding AI that challenge popular notions. Additionally, the evolution of search engines and their limitations highlight the need for a paradigm shift. In this article, we will explore these contrarian theses and discuss the future of AI and search engines.

Contrarian AI Theses:

1. Horizontal LLMs will lose:
   The prevailing belief among AI experts is that Language Models (LLMs) will not lead us to Artificial General Intelligence (AGI). Instead, the next breakthroughs in AI are expected to come from new technologies rather than just more data. Furthermore, it is argued that verticalizing LLMs, using multiple versions for specific applications, may be more effective than relying on a single horizontal LLM. This challenges the notion that LLMs alone can dominate the field of AI.

2. AI will impact specific companies, not entire markets:
   Unlike previous technological waves, AI is not expected to impact entire markets. Instead, AI applies intelligence and learning to different steps in the corporate value chain. Consequently, companies with value chains that are more amenable to taking advantage of AI will benefit the most. It is suggested that companies with similar steps in their value chain may have more in common in terms of benefiting from AI, rather than being defined by the products they make or customers they serve. Thus, AI's impact will be focused on specific companies rather than entire industries.

3. AI will erode competitive advantage:
   The widely-held belief that certain companies possess sustainable competitive advantages is challenged by the rise of AI. Google's acknowledgment that they have no moat against AI exemplifies this perspective. AI has the potential to level the playing field by automating and optimizing various processes, eliminating traditional sources of competitive advantage. This thesis suggests that companies need to adapt and embrace AI to stay ahead in the rapidly changing business landscape.

The Future of Search Engines:

1. Generative AI as the next Google search engine:
   The evolution of search engines has been relatively stagnant, with the current design being rooted in the late 1990s. The content we consume today, such as social networks, data streams, video content, and apps, has vastly changed. It is argued that search engine design needs to shift from using a large database for searching to utilizing that database as training data for a neural network. By generating results with a neural network, search engines can adapt to the changing content landscape and provide more relevant and personalized search results.

2. Shrinking the internet with generative AI:
   The use of trained models in generative AI offers the possibility of drastically reducing the size of the internet. Currently, the internet comprises several petabytes of data, but a generative AI model trained on this data can potentially be reduced to a portable 100-gigabyte model. This would revolutionize the way we access information, making it more feasible to store and carry vast amounts of data on personal devices. However, significant technological advancements are necessary before such a transition can occur.

3. Hybrid solutions and distribution:
   While replacing Google entirely may require substantial technological advancements, a hybrid solution combining generative AI and traditional search engines could be a viable interim option. Google already incorporates generative AI to generate query results for certain questions. This shift in distribution would also impact user accessibility, as installing the entire Google archive on personal devices would no longer be feasible. Instead, ML models could be used to run search queries, transforming the way we interact with search engines.

Conclusion:
The future of AI and search engines is filled with exciting possibilities and challenges. To navigate this evolving landscape successfully, early-stage investors and businesses must consider the contrarian theses surrounding AI and embrace the potential of generative AI in search engine design. Adapting workflows, leveraging AI's capabilities, and investing in innovative technologies will be crucial for maintaining a competitive edge in the AI-driven era.

Actionable Advice:

1. Embrace new AI technologies: Stay updated with the latest advancements in AI and explore emerging technologies that challenge existing paradigms. Investing in innovative AI solutions can provide a competitive advantage in the long run.

2. Identify value chains amenable to AI: Instead of focusing solely on industries, analyze companies' value chains to identify those that are best positioned to leverage AI. Investing in companies with value chains that align with AI capabilities may yield better returns.

3. Prepare for the future of search engines: Recognize the limitations of current search engine designs and explore the potential of generative AI. Investing in research and development of hybrid search engine solutions can position businesses at the forefront of the evolving search landscape.

By considering these contrarian theses and embracing the transformative potential of AI and generative AI in search engines, early-stage investors can make informed decisions and seize opportunities in the rapidly evolving AI industry.