The AGI hype is damaging the AI opportunity

🌈 Abstract

The article discusses the hype around AI, particularly generative AI (GenAI), and how it is damaging the perception of AI within businesses. The author highlights two main ways this hype is harmful:

1. The "blockchain" or "metaverse" type of hype, where a mythical technology is promised to be great in the future if everyone adopts it.
2. The hype that suggests current AI approaches are not good enough for real problems, leading to unhealthy skepticism and overconfidence in "low-powered" current AIs.

The author argues that AI can solve many problems today, but this will not be done by deploying a single model that magically understands a business and makes the best decisions. Instead, it will require more rigor, focus, and control. The author advises not to believe the hype and instead focus on "boringly successful" AI that is not dependent on the mythical AGI.

🙋 Q&A

[01] Hype around AI and GenAI

1. What are the two main ways the hype around AI and GenAI is damaging the perception of AI within businesses?

• The "blockchain" or "metaverse" type of hype, where a mythical technology is promised to be great in the future if everyone adopts it.
• The hype that suggests current AI approaches are not good enough for real problems, leading to unhealthy skepticism and overconfidence in "low-powered" current AIs.

2. What is the author's view on how AI can solve problems in businesses? The author argues that AI can solve many problems today, but this will not be done by deploying a single model that magically understands a business and makes the best decisions. Instead, it will require more rigor, focus, and control.

3. What does the author advise regarding the hype around AI and GenAI? The author advises not to believe the hype and instead focus on "boringly successful" AI that is not dependent on the mythical AGI.

[02] Concerns about the hype

1. What are the author's concerns about the hype being used to push regulations? The author has a concern that the hype is being used to push regulations towards a mythical existential threat, rather than the very real threats that current AIs bring.

2. What is the author's view on the "sidecar" GenAI approaches being used by some vendors? The author sees these "sidecar" GenAI approaches as half-baked, where vendors are just slapping a ChatGPT bot onto an application, often with a poor quality implementation. This leads to disappointment as business users find out the promises were extremely hollow.

3. How does the author characterize the "AGI" and visionary hype, and the capital being thrown at general purpose AI? The author has had conversations with multiple people who are getting jaded by the "AGI" and visionary hype, as well as the huge amounts of capital being thrown at general purpose AI with only a distant promise of relevance.