Developing AI could be turning into a zero-sum game with some major losers

🌈 Abstract

The article discusses the challenges faced by companies developing generative AI models like ChatGPT, Claude, and Gemini. It highlights the high costs and intense competition in this market, making it difficult for these companies to differentiate themselves and turn a profit. The article suggests that the future of AI lies in developing applications rather than ever-more sophisticated and expensive models, as open-source projects are becoming more attractive. The solution proposed is to find AI applications for which users are willing to pay enough to compensate for the high operating costs.

🙋 Q&A

[01] The Challenges Faced by Generative AI Companies

1. What are the main challenges faced by companies developing generative AI models?

• The companies face a long and costly process to develop these models
• There is intense competition in the market, making it difficult for them to differentiate themselves
• The high operating costs, particularly for scaling up computing capacity and hiring specialized personnel, make it difficult to turn a profit

2. How are the main AI companies responding to these challenges?

• They are looking to offer smaller, cheaper models
• They are trying to spread costs around the rest of the company
• They are spending money on buying smaller AI companies that can no longer afford to stay in the game

[02] The Future of AI

1. What is the article's view on the future of AI?

• The future of AI does not lie in developing ever-more sophisticated and expensive models
• Instead, the focus should be on developing the next layer, which is applications
• Big Tech expects to make money by redefining its products and services for companies building verticalized applications based on available generative AI models

2. What is the proposed solution to the challenges faced by generative AI companies?

• The solution is to find AI applications for which users (companies or individuals) are willing to pay amounts that compensate for the huge operating costs
• If consumers pay in tokens, they could benefit from a highly competitive scenario, allowing them to optimize the use of AI models according to their needs, including the use of more open-source or limited in-house models

3. What is the article's overall outlook on the future of generative AI companies?

• Generative AI is going to change the world, but not all the current players will necessarily be around in the years to come.