“Untitled Retrospective and Learnings from AI in Context’s First Two VideosDraft” by ChanaMessinger

<p> Note: I used LLMs to draft different parts of this. I've checked almost everything, but there might be some mistakes remaining.</p><p> Apologies for posting this on Christmas Eve. I wanted to get this out the door before the end of the year. Questions welcome, and if it's easy to pull metrics to answer them, I will.</p><h4 data-internal-id="Summary">Summary</h4><p> 80,000 Hours launched a video program in 2025 focused on longform, cinematic, personality-driven content about AI risks. Our first two longform releases were:</p><ul> <li> We're Not Ready for Superintelligence&nbsp;(the "AI 2027" video): 8.9M views, ~1.4M watch hours</li><li> If you remember one AI disaster, make it this one&nbsp;(the "MechaHitler" video): 2.7M views, ~419K watch hours</li></ul><p> Both videos significantly outperformed our expectations (we'd anticipated 15-50K views for the first). The cost per engagement hour ($0.11 and $0.39 respectively, including staff time) compares favorably to other 80,000 Hours programs.</p><p> This post covers: what we spent, what we got, why we think it worked, and what we'd do differently.</p><h4 data-internal-id="The_numbers">The numbers</h4><h5 data-internal-id="Costs">Costs</h5>CategoryAI 2027MechaHitlerDirect costs~$50K~$64KStaff hours~450 hrs~450 hrs (Note, I’m assuming it's about the same as for AI 2027, I didn't re-ask people how much time they spent.)Total cost (making some assumptions about we should incorporate staff [...] <p>---</p><p><strong>Outline:</strong></p><p>(00:34) Summary</p><p>(01:33) The numbers</p><p>(01:36) Costs</p><p>(02:16) Timing</p><p>(02:40) Results</p><p>(03:46) How valuable is a video watch hour?</p><p>(04:24) Qualitative Feedback</p><p>(04:28) AI 2027</p><p>(05:51) MechaHitler</p><p>(06:12) YouTube commenters like:</p><p>(06:52) What the comments don't like:</p><p>(07:17) Qualitative Analysis</p><p>(07:21) Why we think AI 2027 did well</p><p>(09:56) Why MechaHitler did less well (but still well)</p><p>(10:50) Lessons Learned</p><p>(10:54) Overall what we think matters</p><p>(11:25) Our guess at what's less important (though we're certainly unsure, maybe if we nailed these, we'd get more success)</p><p>(12:24) How our production works</p><p>(12:43) The timeline</p><p>(13:32) Ideation</p><p>(14:06) Scripting</p><p>(14:57) Shooting</p><p>(15:31) Reshoots / Voiceover</p><p>(15:45) Editing</p><p>(16:06) Launch</p><p>(17:00) What we're still figuring out</p><p>(17:36) Closing thoughts</p> <p>---</p> <p><b>First published:</b><br/> December 24th, 2025 </p> <p><b>Source:</b><br/> <a href="https://forum.effectivealtruism.org/posts/RCRaBYSqBaMzHzTjF/untitled-retrospective-and-learnings-from-ai-in-context-s?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Source+URL+in+episode+description&utm_campaign=ai_narration" rel="noopener noreferrer" target="_blank">https://forum.effectivealtruism.org/posts/RCRaBYSqBaMzHzTjF/untitled-retrospective-and-learnings-from-ai-in-context-s</a> </p> <p>---</p> <p>Narrated by <a href="https://type3.audio/?utm_source=TYPE_III_AUDIO&utm_medium=Podcast&utm_content=Narrated+by+TYPE+III+AUDIO&utm_term=ea_forum&utm_campaign=ai_narration" rel="noopener noreferrer" target="_blank">TYPE III AUDIO</a>.</p> <p>---</p><div style="max-width: 100%";><p><strong>Images from the article:</strong></p><a href="https://res.cloudinary.com/cea/image/upload/f_auto,q_auto/v1/mirroredImages/RCRaBYSqBaMzHzTjF/oz9tu9nezie334ckphnx" target="_blank"><img src="https://res.cloudinary.com/cea/image/upload/f_auto,q_auto/v1/mirroredImages/RCRaBYSqBaMzHzTjF/oz9tu9nezie334ckphnx" alt="Line chart showing two metrics declining from October to December 2025." style="max-width: 100%;" /></a><p><em>Apple Podcasts and Spotify do not show images in the episode description. Try <a href="https://pocketcasts.com/" target="_blank" rel="noreferrer">Pocket Casts</a>, or another podcast app.</em></p></div>