Suddently, and then gradually

#AI Thresholds #Ethan

https://www.oneusefulthing.org/p/gradually-then-suddenly-upon-the

This thought-provoking article by Ethan Mollick delves into how technological advancements, akin to Hemingway's description of bankruptcy, often progress gradually before suddenly taking over markets once they cross certain capability thresholds. This phenomenon is particularly evident in AI technology, which is rapidly advancing and finding applications across various industries. However, despite its prowess in some areas, AI still has notable flaws in others. Recent improvements, though, are making AI newly viable for tasks that were previously challenging or impossible.

Capability Thresholds: Mollick emphasizes the concept of capability thresholds—those pivotal moments when technology reaches a level of performance that transforms its utility. These thresholds are not always easily measurable and may only be apparent to experienced users. For instance, the evolution of AI's ability to transcribe data accurately or generate high-quality images and videos has shown marked improvements, crossing thresholds that make these tasks feasible and efficient.

AI's Rapid Progress: One practical suggestion from Mollick is for organizations to maintain an "impossibility list." This list would track tasks that AI currently cannot perform but is on the verge of achieving. Regularly testing new AI models against this list can help identify when these technologies cross new thresholds, potentially unlocking new capabilities and applications.

Personal Engagement with AI: The pace of AI improvement is not just about gradual advancements; it's about those moments when AI capabilities leap forward, often surpassing human performance in specific domains. This rapid progression is facilitated by a culture of innovation and a willingness to experiment with beta-stage applications.

Reflecting on my own experience, my engagement with AI happened quite suddenly a few months ago. Since then, I have gradually incorporated AI into my daily life, both professionally and personally. I have been testing various applications and use cases, witnessing firsthand how AI can enhance productivity and creativity.

Conclusion: The future impact of AI will be shaped more by these key capability thresholds than by steady, incremental improvements. As innovators continue to push the boundaries and users become more open to experimenting with new technologies, AI will likely keep evolving at an astonishing pace, transforming the way we live and work.