AI Masterclass
500 episodes listed below
PRESS PLAY
Spreading awareness about the works of some of the leading educators in the AI revolution."Helping humanity shepherd through this transition as we embrace the final invention of mankind: advanced artificial intelligence." David Shapiro aims to provide a pragmatic and optimistic view of the future.Core Ideas:Metamodernism: Narratives and emergence view of realityPostnihilism: The demand for meaning is false, there is only the Great MysteryHeuristic Imperatives: An ethical, moral, and decision framework.- Reduce suffering- Increase prosperity- Increase understandingPost-Labor Economics: Replace humans with machines (better, faster, cheaper, safer)mantra:EXPLORE: Delve into the unknown, driven…
Based on 6 episodes we listened to · September 2026
David Shapiro makes AI development legible by pairing live experimentation with candid reasoning about alignment, education, work, and advanced machine intelligence.
Artificial Intelligence Masterclass moves between hands-on model building and sweeping questions about how advanced AI might reshape society. David Shapiro works through ideas step by step, testing prompts, inspecting outputs, revising instructions, and explaining why a result…
David Shapiro has room for GPT-3 emails and AGI in space; your curiosity can roam.
Press play if you want
Short reviews from the PlayNext desk, based on the episodes we processed.
A grounded, speculative look at machine rule, data centers, and AGI's possible move into space.
The episode asks what a worthwhile AGI or ASI future might look like once the default demand for human control is put under pressure. The host sets machine rule beside the present order, where humans can seem to serve billionaire interests, then pushes the thought experiment…
A measured control-problem argument about world models, embodiment, cognitive shortcuts, and near-term alignment incentives.
Axiomatic alignment treats AI control as a bet on convergence. If systems are pushed to build accurate, efficient models of the world, they may land on some conclusions that resemble human ones.
A hands-on GPT-3 email demo where writing rules, sample design, and stop tokens meet real-world messiness.
David Shapiro turns an email generator into a practical lesson in prompt design, translating rules for diplomatic writing into sample categories built around different kinds of input.
A hands-on session about moving artificial cognition closer to a decision to act.
Artificial Intelligence Masterclass returns to the ACOG project, focusing on cognition that is not yet tied directly to action and the nested loops that shape it.
A hands-on GPT-3 demonstration turns argument analysis into a practical prompt-chaining workflow.
Artificial Intelligence Masterclass follows David Shapiro as he uses GPT-3 to turn argument analysis into a chain of smaller prompts. Drawing on instructions from Anna R.
A grounded case for college that treats AI fear as real.
This episode makes a plain case for college in the middle of AI anxiety and job loss. The answer is yes, but with a caveat: it is not pitched as a universal fix.
Markets, macro, AI, and crypto in motion.
Daily AI news through the lens of trust, infrastructure, and real-world deployment.
David Shapiro explores whether college is worthwhile in the age of AI, arguing yes with caveats.
Artificial Intelligence Masterclass’s 'How GOOD could AGI become?' asks whether a machine-led future could leave humans with more optionality than today's billionaire-and-government system.
Artificial Intelligence Masterclass's 'Axiomatic Alignment: A critical component to Utopia and the Control Problem' ties AI safety to hard takeoff, orthogonality, instrumental convergence, and three terminal outcomes: extinction, dystopia, and utopia.
["David Shapiro uses this ACOG session to map cognition into context, assessment, recall, ideation, discernment, and an agent model, then tests the sequence on generated scenarios.
David Shapiro uses Anna R.
["Artificial Intelligence Masterclass follows David Shapiro as he tries to fine-tune GPT-3 into a 'perfect email generator' that can turn messy input into polished messages.", "He keeps returning to email intents, UUIDs for extra prompt entropy, and his rule that subject lines should use bottom-line-up-front wording.", "The build uses four training formats - bullet points, conversations with an executive-assistant bot, messy emails, and stories - and he says the dataset reaches 400 samples.", "The final demo shows the fine-tuned 'Dolly' generator rewriting a family email about leaving Jenny's wedding early into a diplomatic note."]
David Shapiro demonstrates a simple version of the ACE framework, a cognition-first AI model, showing code and explaining its architecture.
The host argues that AI investment is not a full bubble, comparing current tech giants' cash flows to the dot-com era.
The episode begins with ads and a brief introduction, then transitions into a webinar about AI and its societal impact.
David Shapiro provides a tutorial on how to use a fine-tuned AI model in the OpenAI playground and via code.
David Shapiro explains prompt engineering techniques for non-native English speakers, emphasizing the use of imperatives when interacting with AI models like ChatGPT.
A podcast exploring AI concepts and demonstrating an AI assistant's capabilities through a simulated user interaction.
The episode discusses the role of chief AI officers in organizations and the Biden administration's mandate for federal departments to hire them.
A coding tutorial demonstrating how to use ChromaDB, a local vector database, for semantic search in a chatbot application.
David Shapiro discusses energy hyperabundance through nuclear fusion and renewables, and how they intersect with AI.
TAU founder Ohad Asor and Professor Enrico Franconi describe a network that decentralizes software development and stores knowledge as the system solves problems.
David Shapiro's Gato Framework team debriefs a Senate hearing featuring Sam Altman, IBM's Christina Montgomery, and Gary Marcus.
Manitoba's Mincome experiment cut hospitalizations by 8.5 percent while barely reducing overall work hours.