When creating policies to deal with AI’s rapid progress, it is important to minimize dangerous AI capability gains that could cause a catastrophe. While restricting scientific research is controversial, it is worthwhile to look at past instances to see if there are lessons that can be applied to limiting or...
We interviewed five AI researchers from leading AI companies about a scenario where AI systems fully automate AI capabilities research. To ground the setting, we stipulated that each employee is replaced by 30 digital copies with the same skill set and the ability to think 30 times faster than a...
When creating policies to deal with AI’s rapid progress, it is important to minimize dangerous AI capability gains that could cause a catastrophe. While restricting scientific research is controversial, it is worthwhile to look at past instances to see if there are lessons that can be applied to limiting or prohibiting certain types of AI research.
This paper reviews previous cryptographic, nuclear, chemical, and biological research restrictions. Each has policy issues that are relevant to understanding how various forms of AI research might be disincentivized or prohibited.
Cryptographic Research
Cryptography played an essential role in World War II. The German Reich used the Enigma cipher to communicate, and Britain’s success in cracking the code gave... (read 4664 more words →)
We interviewed five AI researchers from leading AI companies about a scenario where AI systems fully automate AI capabilities research. To ground the setting, we stipulated that each employee is replaced by 30 digital copies with the same skill set and the ability to think 30 times faster than a human. This represents a 900-fold increase in the cognitive labor that AI companies direct towards advancing AI capabilities.
Our key takeaways are: