Challenging Assumptions: The Unpredictable Trajectory of AI Models

The conventional wisdom is that the value derived from AI will naturally increase as the AI continues to learn and improve. However, this assumption may not always hold true due to a variety of factors. Not all models are capable of learning, and not all models are entirely within our control.

The article linked below discusses an interesting case where GPT-4, a prominent AI model, has seemingly regressed in its capabilities. We will know more over time.

In my experience, when many enterprises start building business cases, they tend to estimate the benefits of new technology with an overly optimistic lens. The allure of AI models, with their promise of continual improvement, can further inflate these expectations.

When building business cases and estimating potential benefits, it’s crucial to consider the possibility of a decrease in model performance. Understanding the potential impact on your business case is key. Always remember, the trajectory of AI models may not always be upward. Do you factor such a potential reduction in AI performance in your business case?

https://www.linkedin.com/feed/update/urn:li:activity:7087248387112099840/

Share your thoughts!

Rajesh Kandaswamy


Rajesh Kandaswamy

AI Strategy Advisor | Board Member | Aspiring Founder
Former Gartner Chief of Research & Fellow