Co-Founder at Zentropi (Trustworthy AI). Formerly Meta Civic Integrity Founder, Google X and Google Civic Innovation Lead, and Groq CPO.https://bsky.app/profile/samidh.bsky.socialhttps://bsky.app/profile/samidh.bsky.social/post/3musrarhuz227The divide in America used to be rural vs urban. Then red states vs blue states. But now I think the real emerging gap is between W-2 Americans and 1099-DIV Americans.06 Sep 2026 00:44 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3musrarhuz227https://bsky.app/profile/samidh.bsky.social/post/3mu3wao5mxc2uICYMI [contains quote post or other embedded content]27 Aug 2026 22:42 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mu3wao5mxc2uhttps://bsky.app/profile/samidh.bsky.social/post/3mu3bgq4alc2hAs usual, @masnick.com is spot on here: https://www.techdirt.com/2026/08/26/meta-just-paid-nearly-17-billion-to-make-sure-it-gets-to-write-the-kid-safety-rules-for-every-other-social-media-platform/ Nothing says "regulatory capture" more than making a payment to the government contingent upon them enforcing certain rules of your choice on your competitors.27 Aug 2026 16:30 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mu3bgq4alc2hhttps://bsky.app/profile/samidh.bsky.social/post/3mtzbzifxb22pThe $12.7B in damages Meta will be paying to the states for harming children is less than... 1. How much it paid ScaleAI for Alexandr Wang ($14.3B) 2. Its one-day market cap increase upon launch of Muse Spark ($94.9B) 3. How much it burned on the Metaverse/VR in 2025 ($19.1B)26 Aug 2026 21:35 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mtzbzifxb22phttps://bsky.app/profile/samidh.bsky.social/post/3mqmuy6357k2dBanning all teens from social media hasn't ever seemed like a great idea to me. A much healthier approach is for platforms to build smoother on-ramps for teens that are developmentally appropriate for each child's age. 🧵 [1/n]...14 Jul 2026 18:54 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mqmuy6357k2dhttps://bsky.app/profile/samidh.bsky.social/post/3mnxmzmqbec2jThe real question is whether these classifiers can find all the dogs that Dave has at home. [contains quote post or other embedded content]10 Jun 2026 21:15 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mnxmzmqbec2jhttps://bsky.app/profile/samidh.bsky.social/post/3mnxlok7uok2yI distinctly remember being at Meta in the wake of the Christchurch massacre, when horrific videos were circulating across Facebook without end. The technologies we had for being able to accurately classify videos just didn't exist. [1/n]10 Jun 2026 20:51 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mnxlok7uok2yhttps://bsky.app/profile/samidh.bsky.social/post/3mnvhhj4or22mToday, in a single day, @dwillner.bsky.social and I had meetings with people in England, Turkey, NYC, SF, Argentina, and Australia. Very inspiring to see four continents of folks all using Zentropi and united in the earnest work of building a better internet.10 Jun 2026 00:30 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mnvhhj4or22mhttps://bsky.app/profile/samidh.bsky.social/post/3mnfk7bhqqk2iCan your LLM actually follow your content policies, or will it just revert to its own static training? Today we're introducing a new benchmark that we call policy steerability that tries to measure this concept: https://blog.zentropi.ai/introducing-the-policy-steerability-benchmark/03 Jun 2026 16:37 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mnfk7bhqqk2ihttps://bsky.app/profile/samidh.bsky.social/post/3mmtyvdlo622yVery cool! @julietshen.bsky.social's independent tests show that our new model CoPE-B cooks :-) Direct link to her results: https://github.com/julietshen/cope-evaluation/blob/main/RESULTS.md [contains quote post or other embedded content]27 May 2026 17:12 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mmtyvdlo622yhttps://bsky.app/profile/samidh.bsky.social/post/3mmtwcgsjzc2kExactly right. Speech shouldn't be ruled by the platform hegemons. Your rules should rule. [contains quote post or other embedded content]27 May 2026 16:25 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mmtwcgsjzc2khttps://bsky.app/profile/samidh.bsky.social/post/3mmtpczp7fk2uSuper pumped to release CoPE-B, our latest policy-adaptive content classification model. It delivers frontier-level accuracy in a self-hostable package that's orders of magnitude cheaper to run-- opening up new possibilities in trustworthy platform design. Details: https://blog.zentropi.ai/meet-cope-b-frontier-quality-content-classification-you-can-self-host/27 May 2026 14:20 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mmtpczp7fk2uhttps://bsky.app/profile/samidh.bsky.social/post/3mllogu6a522oCheck out how the @oversightboard.bsky.social used Zentropi to better understand how child marriage-related content manifests on Meta's platforms. Fantastic example of how advanced content labeling technologies can strengthen both our online and offline world. https://blog.zentropi.ai/how-the-oversight-board-uses-zentropi-to-study-policy-impact-at-scale/11 May 2026 16:18 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mllogu6a522ohttps://bsky.app/profile/samidh.bsky.social/post/3mjhvha4ygk2xIt has been incredible partnering with character.ai since the very start of zentropi.ai. We're excited to share some details of that partnership with this case study. Anyone creating AI-powered systems might find it interesting! https://blog.zentropi.ai/how-zentropi-partners-with-character-ai/14 Apr 2026 17:23 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mjhvha4ygk2xhttps://bsky.app/profile/samidh.bsky.social/post/3mhdxbjecls2mOne of the things we've been thinking about a lot at Zentropi is: what happens when AI agents need to make judgment calls about content — not humans reviewing a queue, but agents acting autonomously?18 Mar 2026 16:54 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mhdxbjecls2mhttps://bsky.app/profile/samidh.bsky.social/post/3mgqfa2tdws2jThere's a major gap in content safety tooling: classifiers typically only score complete text. When you're working with generative AI, "complete text" means the user already saw it. That's too late. So we built a streaming classifier that we're releasing today! Here's what we did and why. 🧵...10 Mar 2026 22:11 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mgqfa2tdws2jhttps://bsky.app/profile/samidh.bsky.social/post/3mfabhegvoc2rZentropi is now integrated into Coop, @roost.tools's open source moderation platform. You can write a content policy in plain English on Zentropi, plug it into Coop as a signal, and have a moderation pipeline running in minutes.19 Feb 2026 18:55 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mfabhegvoc2rhttps://bsky.app/profile/samidh.bsky.social/post/3mddwadrhcs2gI can has cats. [contains quote post or other embedded content]26 Jan 2026 18:55 +0000at://did:plc:zfekat7e2s7iftys552ro7b4/app.bsky.feed.post/3mddwadrhcs2g