Metaculus thinks there will be significant replacement in the next decade: [https://www.metaculus.com/questions/424/in-what-year-will-half-of-new-cars-sold-in-the-us-be-fully-autonomous/](https://www.metaculus.com/questions/424/in-what-year-will-half-of-new-cars-sold-in-the-us-be-fully-autonomous/) [https://www.metaculus.com/questions/11608/self-driving-taxis-available-to-metaculites/](https://www.metaculus.com/questions/11608/self-driving-taxis-available-to-metaculites/)

even a significant number of AI researchers themselves believe, that [AGI is many, many decades away](https://www.metaculus.com/questions/5121/date-of-artificial-general-intelligence/). I personally don't think, that AI will be a necessity for the first generation

There are things called prediction markets and prediction aggregators. See [https://www.metaculus.com/questions/11676/mmlu-sota-in-2023-2025/](https://www.metaculus.com/questions/11676/mmlu-sota-in-2023-2025/) for up-to-date predictions for MMLU SotA. Also, see [https://www.lesswrong.com/posts/arveXgFbJwascKtQC/forecasting-ml-benchmarks-in-2023](https://www.lesswrong.com/posts/arveXgFbJwascKtQC/forecasting-ml-benchmarks-in-2023) for a detailed prediction from a single forecaster

like [this one](https://www.metaculus.com/questions/7483/ai-artist-to-be-credited-in-shonen-jump/). >When will a computer program be credited as an artist of at least ten pages of content in Weekly Shonen Jump? A year ago, predictions were that

www.metaculus.com/questions/3479/date-weakly-general-ai-is-publicly-known/