136 - Venice AI: Private and Uncensored

Gabriel Custodiet speaks with Erik Voorhees of Venice AI. They discuss how Venice takes a different approach to the Big Tech AI options: namely, it does not have access to your prompts, and it does not censor them. Along the way they discuss how surveillance and censorship manifests in the paternalistic AI industry, the future of open-source AI models, the IP question of AI, and why regulating AI might have unintended side effects.   GUEST → https://x.com/ErikVoorhees → https://venice.ai/ → https://x.com/askvenice → https://moneyandstate.com/blog/the-separation-of-mind-and-state (explanation of project) WATCHMAN PRIVACY → https://watchmanprivacy.com (Including privacy consulting) → https://twitter.com/watchmanprivacy → https://escapethetechnocracy.com/   CRYPTO DONATIONS →8829DiYwJ344peEM7SzUspMtgUWKAjGJRHmu4Q6R8kEWMpafiXPPNBkeRBhNPK6sw27urqqMYTWWXZrsX6BLRrj7HiooPAy (Monero) →https://btcpay0.voltageapp.io/apps/3JDQDSj2rp56KDffH5sSZL19J1Lh/pos (BTC)   Music by Karl Casey @ White Bat Audio   Timeline 00:00 – Introduction 3:27 – Privacy-invasion of Big Tech AI companies 6:50 – How does Venice AI give the user privacy? 9:55 – How long will non-account service last? 10:25 – What does Venice (the city/concept) have to do with anything? 11:40 – How do open-source models work in AI? 16:50 – Phone version of AI (not app store version) 18:05 – Censorship within AI 24:42 – Uncensoring AI makes it faster and better 25:55 – Other uncensored AI companies 27:10 – Making AI “safe” 30:30 – IP question of AI 34:45 – Who owns the creation of AI tools? 35:10 – Any red lines where AI should be “regulated”? 37:20 – Does Venice have any access to our prompts? 38:15 – Venice AI vs self-hosted models 39:40 – Would Venice ever train their own AI model? 40:25 – Final thoughts