The author suggests that for generative AI's pricing models to improve, they need to be based on the customer segment, the application of the technology, and the value it drives for them. A simplification of the pricing tiers and models would help, making it easy for customers to determine the cost for a given piece of work. The author hopes for a future where pricing models are simplified and more value-based.
Key takeaways:
- The current pricing models for generative AI, such as OpenAI's GPT-4, are often unclear and not value-based, leading to confusion among users and professionals.
- These pricing models are usually usage-based, charging customers based on the number of API requests made or offering unlimited usage for a fixed monthly fee. However, these models do not directly correlate with a value-driven outcome.
- The cost of running large language models (LLMs) is high due to the need for expensive GPUs, which consume a lot of energy. This cost-based pricing model is common for AI language models, but can lead to sticker shock for buyers.
- To improve generative AI's pricing models, they should be based on the segment of the customers, their application of the technology, and the value it drives for them. Simplification of the pricing tiers and models would also help users understand the cost for their specific needs.