The actual definition of AGI provided in the Wikipedia link you provide is that "an AGI could learn to accomplish any intellectual task that human beings or animals can perform".

You chose the alternate, "superior to" definition. I'm curious as to why, since we already have terminology for this.

"• Artificial Narrow Intelligence (ANI) or “weak” AI, narrowly-defined set of specific tasks

• Artificial General Intelligence (AGI) or “strong” AI, think and make decisions like us

• Artificial Super Intelligence (ASI), surpassing human intelligence"

https://medium.com/@frickingruvin/artificial-intelligence-ai-taxonomies-caa2ddc6cc7e

I'll also point out that the "incredible surge in complexity and capability of AI models in recent years" has taken place entirely in the narrow (ANI) type of AI. No one's even made a plausible start on general (AGI), let alone super (ASI).