# Do Not Create Dangerous Capabilities Blindly by Dario Amodei > "Second, we test frequently for these dangerous capabilities at regular intervals along the compute scaling curve. This is to ensure that we don’t blindly create dangerous capabilities without even knowing we have done so." > **— Dario Amodei**, *November 1, 2023, UK AI Safety Summit prepared remarks* ## Sources and Context - **Prepared remarks and policy context:** [Dario Amodei’s prepared remarks from the AI Safety Summit on Anthropic’s Responsible Scaling Policy](https://www.anthropic.com/news/uk-ai-safety-summit) — Amodei was explaining why Anthropic tests models repeatedly for dangerous capabilities as computing scale increases. ## Related Articles and Collections - **Collection:** [[collections/Machine Succession|Machine Succession]] - **Collection:** [[collections/MARS|MARS]] - **Article:** [[articles/AI Escape Is the Wrong Metaphor|AI Escape Is the Wrong Metaphor]] - **Article:** [[articles/Host-Indexed Autonomy|Host-Indexed Autonomy]] - **Article:** [[articles/The Last Migration Will Not Be Human|The Last Migration Will Not Be Human]] - **Wiki map:** [[wiki/Artificial General Intelligence|Artificial General Intelligence]] ## Related Topics - [[wiki/Alignment Problem|Alignment Problem]] - [[wiki/State-Dependent Catastrophe|State-Dependent Catastrophe]] - [[wiki/Machine Succession|Machine Succession]] - [[wiki/AI Coordination Systems|AI Coordination Systems]] - [[wiki/Marsification|Marsification]] - [[wiki/Operational Autonomy|Operational Autonomy]] ## Share on Social Media ``` "Second, we test frequently for these dangerous capabilities at regular intervals along the compute scaling curve. This is to ensure that we don’t blindly create dangerous capabilities without even knowing we have done so." — Dario Amodei, November 1, 2023, UK AI Safety Summit prepared remarks https://bryantmcgill.com/simple-reminders-dario-amodei-not-dangerous-capabilities-blindly ```