# Let’s Slow Down by Yoshua Bengio > "Let’s slow down [AI development]. Let’s make sure that we develop better guardrails, let’s make sure that we discuss these questions internationally just like we’ve done for nuclear power and nuclear weapons. Let’s make sure we better understand these very large systems, that we improve on their robustness and the process by which we can audit them and verify that they are safe for the public." > **— Yoshua Bengio**, *March 2023, Mila intervention on AI safety* ## Sources and Context - **Exact original source passage:** “Let’s slow down. Let’s make sure that we develop better guardrails, let’s make sure that we discuss these questions internationally just like we’ve done for nuclear power and nuclear weapons. Let’s make sure we better understand these very large systems, that we improve on their robustness and the process by which we can audit them and verify that they are safe for the public.” - **Editorial repair and evidence:** Brackets add the object from the same Mila report: the requested pause concerned training AI systems more powerful than GPT-4. The complete passage retains Bengio’s reasons for slowing down. [Mila](https://mila.quebec/en/news/yoshua-bengio-joins-hundreds-of-signatories-in-open-letter-pleading-for-safer-ai-systems), published March 31, 2023. No timecode is asserted. ## Related Articles and Collections - **Collection:** [[collections/Machine Succession|Machine Succession]] - **Collection:** [[collections/Consciousness Continuity|Consciousness Continuity]] - **Article:** [[articles/AI Escape Is the Wrong Metaphor|AI Escape Is the Wrong Metaphor]] - **Article:** [[articles/The Skeuomorphic Interval|The Skeuomorphic Interval]] - **Article:** [[articles/Host-Indexed Autonomy|Host-Indexed Autonomy]] - **Wiki map:** [[wiki/Artificial Intelligence|Artificial Intelligence]] ## Related Topics - [[wiki/Alignment Problem|Alignment Problem]] - [[wiki/AI Coordination Systems|AI Coordination Systems]] - [[wiki/Control Theory|Control Theory]] - [[wiki/Machine Succession|Machine Succession]] - [[wiki/OpenAI Frontier|OpenAI Frontier]] - [[wiki/Cybernetic Control|Cybernetic Control]] ## Share on Social Media ``` "Let’s slow down [AI development]. Let’s make sure that we develop better guardrails, let’s make sure that we discuss these questions internationally just like we’ve done for nuclear power and nuclear weapons. Let’s make sure we better understand these very large systems, that we improve on their robustness and the process by which we can audit them and verify that they are safe for the public." — Yoshua Bengio, March 2023, Mila intervention on AI safety https://bryantmcgill.com/simple-reminders-yoshua-bengio-slow ```