Why AI alignment could be hard with modern deep learning

Why would we program AI that wants to harm us? Because we might not know how to do otherwise.

Click lower right to download or find on Apple Podcasts, Spotify, Stitcher, etc. --> This is a guest post by my colleague Ajeya Cotra . Holden previously mentioned the idea that advanced AI systems (e.g. PASTA ) may develop dangerous goals that cause them to deceive or disempower humans. This might sound like a pretty out-there concern . Why would we program AI that wants to harm us? But I think it could actually be a difficult problem to avoid, especially if advanced AI is developed using deep learning (often used to develop state-of-the-art AI today). In deep learning, we don’t program a com

related reading

• Why AI alignment could be hard with modern deep learningcold-takes.com
• Alignment remains a hard, unsolved problem — LessWronglesswrong.com
• Self-Fulfilling Misalignment Data Might Be Poisoning Our AI Modelsturntrout.com
• The Artificiality of Alignmentjoinreboot.org
• Current AIs seem pretty misaligned to me — LessWronglesswrong.com
• Model Organisms of Misalignment: The Case for a New Pillar of Alignment Research — AI Alignment Forumalignmentforum.org
• What failure looks like — AI Alignment Forumalignmentforum.org
• Deep Deceptiveness — LessWronglesswrong.com
• AI Safety Seems Hard to Measurecold-takes.com
• Current AIs seem pretty misaligned to meblog.redwoodresearch.org
• New report: “Scheming AIs: Will AIs fake alignment during training in order to get power?”joecarlsmith.com
• How do we (more) safely defer to AIs? — LessWronglesswrong.com