As humans need to learn most things from scratch after birth, including language, it’s worth considering whether AI chatbots represent a generational leap in biological evolution. Once a language model is trained, it can be replicated indefinitely, leading to higher-kind knowledge that’s inheritable.What implications does this have for our understanding of evolution and the acquisition of knowledge?
I used to believe that human language skills set us apart from other animals and computers, and with the increase in model size, large language models are gaining new abilities like mathematical ability, logic, and humor. Is language truly the basis for human-like intelligence? If animals learned language (by whatever means, even genetic modification), could they also develop these other skills?
When an AI agent achieves superhuman general intelligence, what will be the impact on society? Will people treat it like a “god” and rely on it for everything, leading to a halt in education? A recent trend is people making chatGPT write their homework. Alternatively, will people continue to learn despite the availability of better and instant answers provided by machines - as a kind of “rebellion”?
In the age of chatGPT-like technology, can we trust new information anymore? The ease and affordability of rewording Wikipedia articles to support a particular viewpoint and the use of chatGPT to write books raise concerns. Have we entered a post-information era where we can’t trust anything produced after 2022?
Discussion and theories are appreciated.
Politely asking Genghis Khan to stop killing people
Marimo (also known as moss ball) is a rare growth form of Aegagropila linnaei (a species of filamentous green algae) in which the algae grow into large green balls with a velvety appearance.
The species can be found in a number of lakes and rivers in Japan and Northern Europe. Colonies of marimo balls are known to form in Japan and Iceland, but their population has been declining.
