If I follow you, you ask for less AI safety to allow a warning shot, to get more AI safety in the end ? I see how it could work the first time, but once AI safety has been increased in the lab, you should expect less warning shots. Moreover, how can you be sure to allow a mere warning shot and not full takeover ? I think we need more honeypot setups but not less control (sandboxing etc).
If I follow you, you ask for less AI safety to allow a warning shot, to get more AI safety in the end ? I see how it could work the first time, but once AI safety has been increased in the lab, you should expect less warning shots. Moreover, how can you be sure to allow a mere warning shot and not full takeover ? I think we need more honeypot setups but not less control (sandboxing etc).