# AI future guide: videos, podcasts & key perspectives | AGI

Watch AI interviews on the page, read podcast takeaways, and explore views on work, learning, business and the future.

---
Canonical page: https://agiscorecard.com/future-guide
Machine-readable verdicts: https://agiscorecard.com/data.json (CC BY 4.0)
This Markdown mirror is generated from the page; the HTML page is canonical.

## Source-linked future guide

You can read all sources and guides without JavaScript. Enable it to filter, save or export your own plan. Big ideas. Worth watching. AI interviews and creator videos, organized around the questions that matter to you. Read the takeaways. Watch the original. Videos 381 Audio 289 Reading 10 Search views or creators All perspectives Work & agents Learning & skills Business & creativity Education & family AI progress Future forecasts 381 videos · Read the view, then press play More filters Order Takeaways first Newest first Language All languages Chinese English Source All creators Interviews Creator commentary Research & learning Company channels Creator All creators a16z AI Engineer AI Explained AI for Good AI For Humans AI Jason All-In Podcast Andrej Karpathy Anthropic CNBC DeepLearning.AI Dwarkesh Podcast Fireship Getting Smart Google DeepMind Hugging Face IBM Technology Latent Space Lenny’s Podcast Lex Fridman Machine Learning Street Talk Matt Wolfe Matthew Berman No Priors NVIDIA OpenAI Sequoia Capital Stanford HAI The AI Advantage The AI Daily Brief The Cognitive Revolution The Next Wave TWIML AI Podcast Two Minute Papers Wes Roth WorkLife / TED Y Combinator Yannic Kilcher 张小珺商业访谈录 李宏毅 硅谷 101 硅谷 101 播客 Published All dates Past 30 days Past 90 days Archive Latent Space OpenAI Computer use, recovery and programmable decisions ▶ Watch here Ari Weinstein & Nikunj Handa 2026-09-30 Start here Can an agent recover when a task goes wrong? (/future-guide/agents-recover) Key takeaways Watch what happens after failure Weinstein emphasizes agents retrying and debugging after problems, beyond merely starting a task. Route, classify, then act Handa describes a decision interface for classification and routing inside applications. Read the full perspective ↗ (/future-guide/agents-recover)Original video ↗ (https://www.youtube.com/watch?v=z9OkBD2-MDU) Latent Space Claude Code Claude Code and software that adapts to you ▶ Watch here Thariq Shihipar 2026-09-29 When AI writes code, who defines what good means? (/future-guide/requirements-still-matter) Key takeaways Make the requirement clear Shihipar stresses understanding the problem and communicating requirements to coding agents. Software that keeps changing Shihipar discusses software that users can continually reshape with agents. Read the full perspective ↗ (/future-guide/requirements-still-matter)Original video ↗ (https://www.youtube.com/watch?v=IZAlq-V19U8) Lenny’s Podcast Molly Graham Career transitions and giving work to AI ▶ Watch here Molly Graham 2026-09-27 What should I keep when I give work to AI? (/future-guide/delegate-keep-judgment) Key takeaways Delegate tasks, retain judgment Graham distinguishes delegating to AI from developing people, emphasizing human judgment and direction. Make the transition together Graham describes career change as involving identity, loss and relationships as well as new opportunities. Read the full perspective ↗ (/future-guide/delegate-keep-judgment)Original video ↗ (https://www.youtube.com/watch?v=5-96FyJFiCA) Latent Space OpenRouter A market with more than one winning model ▶ Watch here Alex Atallah & Anjney Midha 2026-09-25 Will one model be best for everything? (/future-guide/many-models) Key takeaways Expect different model strengths Atallah and Midha challenge the idea of a single model winning every use case. Help useful AI reach a user The discussion highlights distribution and access as important parts of the AI market. Read the full perspective ↗ (/future-guide/many-models)Original video ↗ (https://www.youtube.com/watch?v=dCX4PE2HxMs) No matching content. Try another topic or clear the filters. Clear filters More videos Videos play here when you click. If a publisher limits playback, use the original link. Updated daily · Sources and editorial notes Sources checked · 46/46 · Last check: 2026-10-03 09:00:29 UTC Takeaways paraphrase the source. Newly discovered episodes show a short publisher introduction when available, until their viewpoints are reviewed. Topic labels are navigation aids, not fact checks. Company channels present the publisher’s own perspective. The rolling collection keeps up to two years of episodes; Shorts are excluded. Latent Space (https://www.latent.space/) Checked · Last success: 2026-10-03 09:00 UTC Dwarkesh Podcast (https://www.dwarkesh.com/) Checked · Last success: 2026-10-03 09:00 UTC Lenny’s Podcast (https://www.lennysnewsletter.com/) Checked · Last success: 2026-10-03 09:00 UTC Lex Fridman (https://lexfridman.com/podcast/) Checked · Last success: 2026-10-03 09:00 UTC No Priors (https://www.youtube.com/@NoPriorsPodcast) Checked · Last success: 2026-10-03 09:00 UTC AI Explained (https://aiexplainedopodcast.buzzsprout.com/) Checked · Last success: 2026-10-03 09:00 UTC Two Minute Papers (https://www.youtube.com/channel/UCbfYPyITQ-7l4upoX8nvctg) Checked · Last success: 2026-10-03 09:00 UTC Fireship (https://www.youtube.com/channel/UCsBjURrPoezykLs9EqgamOA) Checked · Last success: 2026-10-03 09:00 UTC 硅谷 101 (https://sv101.fireside.fm/) Checked · Last success: 2026-10-03 09:00 UTC 张小珺商业访谈录 (https://www.xiaoyuzhoufm.com/podcast/626b46ea9cbbf0451cf5a962) Checked · Last success: 2026-10-03 09:00 UTC Latent Space (https://www.youtube.com/@LatentSpacePod) Checked · Last success: 2026-10-03 09:00 UTC Dwarkesh Podcast (https://www.youtube.com/@DwarkeshPatel) Checked · Last success: 2026-10-03 09:00 UTC Lenny’s Podcast (https://www.youtube.com/@LennysPodcast) Checked · Last success: 2026-10-03 09:00 UTC Lex Fridman (https://www.youtube.com/@lexfridman) Checked · Last success: 2026-10-03 09:00 UTC No Priors (https://www.youtube.com/@NoPriorsPodcast) Checked · Last success: 2026-10-03 09:00 UTC AI Explained (https://www.youtube.com/@aiexplained-official) Checked · Last success: 2026-10-03 09:00 UTC Matt Wolfe (https://www.youtube.com/@mreflow) Checked · Last success: 2026-10-03 09:00 UTC Matthew Berman (https://www.youtube.com/@matthew_berman) Checked · Last success: 2026-10-03 09:00 UTC Y Combinator (https://www.youtube.com/@ycombinator) Checked · Last success: 2026-10-03 09:00 UTC a16z (https://www.youtube.com/@a16z) Checked · Last success: 2026-10-03 09:00 UTC Sequoia Capital (https://www.youtube.com/@sequoiacapital) Checked · Last success: 2026-10-03 09:00 UTC Stanford HAI (https://www.youtube.com/@StanfordHAI) Checked · Last success: 2026-10-03 09:00 UTC AI Engineer (https://www.youtube.com/@aiDotEngineer) Checked · Last success: 2026-10-03 09:00 UTC Google DeepMind (https://www.youtube.com/@googledeepmind) Checked · Last success: 2026-10-03 09:00 UTC OpenAI (https://www.youtube.com/@OpenAI) Checked · Last success: 2026-10-03 09:00 UTC Anthropic (https://www.youtube.com/@anthropic-ai) Checked · Last success: 2026-10-03 09:00 UTC Andrej Karpathy (https://www.youtube.com/@AndrejKarpathy) Checked · Last success: 2026-10-03 09:00 UTC DeepLearning.AI (https://www.youtube.com/@Deeplearningai) Checked · Last success: 2026-10-03 09:00 UTC Hugging Face (https://www.youtube.com/@HuggingFace) Checked · Last success: 2026-10-03 09:00 UTC Wes Roth (https://www.youtube.com/@WesRoth) Checked · Last success: 2026-10-03 09:00 UTC The AI Advantage (https://www.youtube.com/@aiadvantage) Checked · Last success: 2026-10-03 09:00 UTC AI Jason (https://www.youtube.com/@AIJasonZ) Checked · Last success: 2026-10-03 09:00 UTC All-In Podcast (https://www.youtube.com/@allin) Checked · Last success: 2026-10-03 09:00 UTC The Cognitive Revolution (https://www.youtube.com/@CognitiveRevolutionPodcast) Checked · Last success: 2026-10-03 09:00 UTC AI For Humans (https://www.youtube.com/@AIForHumansShow) Checked · Last success: 2026-10-03 09:00 UTC The Next Wave (https://www.youtube.com/@TheNextWavePod) Checked · Last success: 2026-10-03 09:00 UTC Yannic Kilcher (https://www.youtube.com/@YannicKilcher) Checked · Last success: 2026-10-03 09:00 UTC The AI Daily Brief (https://www.youtube.com/@AIDailyBrief) Checked · Last success: 2026-10-03 09:00 UTC 硅谷 101 (https://www.youtube.com/channel/UCKV2yWPB3wn0RTZh3cTD8YA) Checked · Last success: 2026-10-03 09:00 UTC 硅谷 101 播客 (https://www.youtube.com/@valley101podcast) Checked · Last success: 2026-10-03 09:00 UTC 李宏毅 (https://www.youtube.com/@HungyiLeeNTU) Checked · Last success: 2026-10-03 09:00 UTC TWIML AI Podcast (https://www.youtube.com/@twimlai) Checked · Last success: 2026-10-03 09:00 UTC Machine Learning Street Talk (https://www.youtube.com/@MachineLearningStreetTalk) Checked · Last success: 2026-10-03 09:00 UTC AI for Good (https://www.youtube.com/@AIforGood) Checked · Last success: 2026-10-03 09:00 UTC IBM Technology (https://www.youtube.com/@IBMTechnology) Checked · Last success: 2026-10-03 09:00 UTC NVIDIA (https://www.youtube.com/@NVIDIA) Checked · Last success: 2026-10-03 09:00 UTC Explore all reviewed views & evidence 30 What brought you here? One person can have more than one question. Switch whenever you need. Understand the change What is happening, and what is still a prediction? Adapt my work Which parts of my work could change first? Choose what to learn Which skill can I put into practice? Explore income ideas What useful result could someone pay for? Think about education How can we use AI without skipping learning? Test my judgment What do I believe, and what would change it? Understand work changes Start with the task, then explore the forecast. All views ↗ Browse 16 reviewed interviews 2026-10-02 · Written interview How Airbnb is reorganizing work around AI Ahmad Al-Dahle · Latent Space Show related views (2) → 2026-09-30 · Video Computer use, recovery and programmable decisions Ari Weinstein & Nikunj Handa · Latent Space Show related views (2) → 2026-09-29 · Video Claude Code and software that adapts to you Thariq Shihipar · Latent Space Show related views (2) → 2026-09-28 · Video Agent permissions and the engineering of safety Jensen Huang · CNBC Show related views (2) → 2026-09-27 · Video Career transitions and giving work to AI Molly Graham · Lenny’s Podcast Show related views (2) → 2026-09-25 · Video A market with more than one winning model Alex Atallah & Anjney Midha · Latent Space Show related views (2) → 2026-09-25 · Video Interactive worlds and creative control Anastasis Germanidis · Latent Space Show related views (2) → 2026-09-22 · Video AI for science, useful metrics and domain knowledge John Platt · Latent Space Show related views (2) → 2026-09-21 · Video Programmable judgment in everyday software Diogo Almeida · Latent Space Show related views (1) → 2026-09-17 · Video Agent swarms and the organization of AI work Noam Brown · Dwarkesh Podcast Show related views (2) → 2026-09-16 · Video Why trust and testing can limit AI adoption Rune Kvist · Latent Space Show related views (2) → 2026-09-16 · Video Education needs practice, feedback and durable skills Sal Khan · Getting Smart Show related views (2) → 2026-09-14 · Audio Can AI teach without replacing human connection? Leah Belsky · WorkLife / TED Show related views (2) → 2026-09-11 · Video A debate about recursive AI improvement John Schulman · Beren Millidge · Charlie O’Neill · Dwarkesh Podcast Show related views (2) → 2026-02-13 · Video Powerful AI and economic change Dario Amodei · Dwarkesh Podcast Show related views (1) → 2025-10-17 · Video Agents, learning and the path to AGI Andrej Karpathy · Dwarkesh Podcast Show related views (2) → Search views Published All dates Past 30 days Past 90 days Archive: over 90 days Saved only 16 interviews · 30 views · Reviewed 2026-10-03 Latent Space Airbnb Written interview (/future-guide/evaluate-your-task) Ahmad Al-Dahle · 2026-10-02 Should the best AI model depend on the task? (/future-guide/evaluate-your-task) Al-Dahle describes choosing models using application-specific evaluations of performance, latency and cost. Read the view & evidence → (/future-guide/evaluate-your-task) Save Latent Space Airbnb Written interview (/future-guide/prototype-together) Ahmad Al-Dahle · 2026-10-02 Could a shared prototype reduce handoffs? (/future-guide/prototype-together) Al-Dahle describes Airbnb teams moving from sequential handoffs to collaborating on working prototypes. Read the view & evidence → (/future-guide/prototype-together) Save Latent Space OpenAI Video · 00:06:04 (/future-guide/agents-recover) Ari Weinstein & Nikunj Handa · 2026-09-30 Can an agent recover when a task goes wrong? (/future-guide/agents-recover) Weinstein emphasizes agents retrying and debugging after problems, beyond merely starting a task. Read the view & evidence → (/future-guide/agents-recover) Save Latent Space OpenAI Video · 00:25:32 (/future-guide/small-decisions) Ari Weinstein & Nikunj Handa · 2026-09-30 Could a small decision be the useful AI feature? (/future-guide/small-decisions) Handa describes a decision interface for classification and routing inside applications. Read the view & evidence → (/future-guide/small-decisions) Save Latent Space Claude Code Video · 00:36:35 (/future-guide/mutable-software) Thariq Shihipar · 2026-09-29 Could software adapt to one person’s workflow? (/future-guide/mutable-software) Shihipar discusses software that users can continually reshape with agents. Read the view & evidence → (/future-guide/mutable-software) Save Latent Space Claude Code Video · 00:15:37 (/future-guide/requirements-still-matter) Thariq Shihipar · 2026-09-29 When AI writes code, who defines what good means? (/future-guide/requirements-still-matter) Shihipar stresses understanding the problem and communicating requirements to coding agents. Read the view & evidence → (/future-guide/requirements-still-matter) Save CNBC NVIDIA Video (/future-guide/agent-permissions) Jensen Huang · 2026-09-28 How much permission should an AI assistant have? (/future-guide/agent-permissions) Huang argues that agentic systems should operate with restrictive rights. Read the view & evidence → (/future-guide/agent-permissions) Save CNBC NVIDIA Video (/future-guide/safety-is-engineering) Jensen Huang · 2026-09-28 Can engineering measures keep pace with AI risk? (/future-guide/safety-is-engineering) Huang frames AI safety as an engineering problem that can be addressed through system design. Read the view & evidence → (/future-guide/safety-is-engineering) Save Lenny’s Podcast Molly Graham Video (/future-guide/career-change-together) Molly Graham · 2026-09-27 Why does adapting to AI need more than new tools? (/future-guide/career-change-together) Graham describes career change as involving identity, loss and relationships as well as new opportunities. Read the view & evidence → (/future-guide/career-change-together) Save Lenny’s Podcast Molly Graham Video (/future-guide/delegate-keep-judgment) Molly Graham · 2026-09-27 What should I keep when I give work to AI? (/future-guide/delegate-keep-judgment) Graham distinguishes delegating to AI from developing people, emphasizing human judgment and direction. Read the view & evidence → (/future-guide/delegate-keep-judgment) Save Latent Space Runway Video · 01:24:48 (/future-guide/creative-control) Anastasis Germanidis · 2026-09-25 If AI generates images, where does creative skill matter? (/future-guide/creative-control) Germanidis emphasizes references, steerability and iteration within creative workflows. Read the view & evidence → (/future-guide/creative-control) Save Latent Space OpenRouter Video · 00:17:27 (/future-guide/distribution-opportunity) Alex Atallah & Anjney Midha · 2026-09-25 Where is the opportunity beyond building a model? (/future-guide/distribution-opportunity) The discussion highlights distribution and access as important parts of the AI market. Read the view & evidence → (/future-guide/distribution-opportunity) Save Latent Space Runway Video · 00:39:39 (/future-guide/interactive-worlds) Anastasis Germanidis · 2026-09-25 What changes when generated video becomes interactive? (/future-guide/interactive-worlds) Germanidis discusses real-time video and world models as a direction for interactive experiences. Read the view & evidence → (/future-guide/interactive-worlds) Save Latent Space OpenRouter Video · 00:14:28 (/future-guide/many-models) Alex Atallah & Anjney Midha · 2026-09-25 Will one model be best for everything? (/future-guide/many-models) Atallah and Midha challenge the idea of a single model winning every use case. Read the view & evidence → (/future-guide/many-models) Save Latent Space John Platt Video (/future-guide/domain-knowledge) John Platt · 2026-09-22 How do I build expertise when AI can do the easy steps? (/future-guide/domain-knowledge) Platt emphasizes domain knowledge and the practice involved in developing scientific judgment. Read the view & evidence → (/future-guide/domain-knowledge) Save Latent Space John Platt Video (/future-guide/metrics-versus-progress) John Platt · 2026-09-22 Does a better score mean we solved the right problem? (/future-guide/metrics-versus-progress) Platt discusses the importance of choosing objectives that reflect useful scientific progress. Read the view & evidence → (/future-guide/metrics-versus-progress) Save Latent Space Jev Video · 01:18:28 (/future-guide/software-judgment) Diogo Almeida · 2026-09-21 What if AI improves the software you already use? (/future-guide/software-judgment) Almeida argues that AI can strengthen existing software by becoming a small, programmable part of it. Read the view & evidence → (/future-guide/software-judgment) Save Dwarkesh Podcast Noam Brown Video · 00:00:00 (/future-guide/agent-coordination) Noam Brown · 2026-09-17 Does adding more agents always improve the result? (/future-guide/agent-coordination) Brown explores parallel agent work and the challenges of coordinating many contributors. Read the view & evidence → (/future-guide/agent-coordination) Save Dwarkesh Podcast Noam Brown Video · 00:15:28 (/future-guide/ai-work-organization) Noam Brown · 2026-09-17 Will AI change how work is organized? (/future-guide/ai-work-organization) Brown considers how more capable models might organize increasingly large groups of agents. Read the view & evidence → (/future-guide/ai-work-organization) Save Latent Space AIUC Video · 00:07:58 (/future-guide/deployable-trust) Rune Kvist · 2026-09-16 Why can a capable AI still be hard to deploy? (/future-guide/deployable-trust) Kvist argues that trust and accountability can constrain adoption even when AI capabilities improve. Read the view & evidence → (/future-guide/deployable-trust) Save Getting Smart Khan Academy Video (/future-guide/durable-skills) Sal Khan · 2026-09-16 Which skills deserve practice alongside AI tools? (/future-guide/durable-skills) Khan highlights communication and collaboration as skills that education should develop alongside technical work. Read the view & evidence → (/future-guide/durable-skills) Save Getting Smart Khan Academy Video (/future-guide/feedback-before-answers) Sal Khan · 2026-09-16 Could a tutor notice the mistake before I ask? (/future-guide/feedback-before-answers) Khan discusses proactive feedback instead of relying only on learners to ask a chatbot questions. Read the view & evidence → (/future-guide/feedback-before-answers) Save Latent Space AIUC Video · 00:18:52 (/future-guide/test-failure-cases) Rune Kvist · 2026-09-16 What does a polished AI demo leave untested? (/future-guide/test-failure-cases) Kvist emphasizes testing failure cases beyond a successful demonstration. Read the view & evidence → (/future-guide/test-failure-cases) Save WorkLife / TED AI & learning Audio (/future-guide/learn-by-making) Leah Belsky · 2026-09-14 Could learning start with a problem I noticed? (/future-guide/learn-by-making) Belsky discusses learning through making solutions to concrete problems. Read the view & evidence → (/future-guide/learn-by-making) Save WorkLife / TED AI & learning Audio (/future-guide/learning-needs-people) Leah Belsky · 2026-09-14 Can an AI tutor replace the motivation people provide? (/future-guide/learning-needs-people) Belsky’s conversation examines motivation, accountability and human relationships in AI-supported education. Read the view & evidence → (/future-guide/learning-needs-people) Save Dwarkesh Podcast AI debate Video · 00:00:00 (/future-guide/improvement-bottlenecks) John Schulman · Beren Millidge · Charlie O’Neill · 2026-09-11 What could slow AI’s own research progress? (/future-guide/improvement-bottlenecks) The panel debates whether research and engineering bottlenecks could constrain recursive AI improvement. Read the view & evidence → (/future-guide/improvement-bottlenecks) Save Dwarkesh Podcast AI debate Video · 00:45:24 (/future-guide/simulation-gap) John Schulman · Beren Millidge · Charlie O’Neill · 2026-09-11 Will success in simulation transfer to real work? (/future-guide/simulation-gap) The discussion examines the difficulty of transferring simulated success to complex real-world tasks. Read the view & evidence → (/future-guide/simulation-gap) Save Dwarkesh Podcast Dario Video · 00:12:36 (/future-guide/fast-ai-progress) Dario Amodei · 2026-02-13 · Archive How should we prepare if powerful AI arrives sooner? (/future-guide/fast-ai-progress) Amodei expects very powerful AI within a few years and discusses how capability may spread through the economy. Read the view & evidence → (/future-guide/fast-ai-progress) Save Dwarkesh Podcast Karpathy Video · 00:00:00 (/future-guide/agents-take-time) Andrej Karpathy · 2025-10-17 · Archive Why might dependable AI coworkers take longer? (/future-guide/agents-take-time) Karpathy describes agents as a decade-long engineering effort, pointing to gaps including continual learning. Read the view & evidence → (/future-guide/agents-take-time) Save Dwarkesh Podcast Karpathy Video · 01:56:20 (/future-guide/learning-with-ai) Andrej Karpathy · 2025-10-17 · Archive Does getting an answer mean I have learned? (/future-guide/learning-with-ai) Karpathy contrasts answering questions with a tutor who understands a learner and chooses an appropriate challenge. Read the view & evidence → (/future-guide/learning-with-ai) Save Show more reviewed views No matching views yet. Try another question or clear your filters. Show all views ✳ JARVIS Research guide Start with a task you know. Choose one repetitive task. Compare a manual attempt and an AI-assisted attempt, including review time and errors. Make my next step ↗ (/future-guide/agents-take-time?goal=work#plan) Based on published sources and editorial guidance. This is not an AI-generated forecast or a personal capability assessment. My saved views & next steps My next steps Your saved views and plans, kept on this device. Save a view, then record one action you want to try. Export my notebook Import a backup Delete local records No automatic notifications. You can set your own review date and export a calendar entry from a plan. When you want to keep the work Your next step deserves a place to return to. Reading, video access, local saves and exports are free. Use the same plan on another device with AGI’s existing cloud workspace and version history. No paid interview collection or automated personal-report subscription is offered here. AGI cloud membership 9 USDT / 30 days Plus the order-matching decimal and network fees. 50 workspaces, 10 versions per workspace, 64 KB per workspace and 5 MB in total. No auto-renewal or extra AI allowance. Open my plan in membership The member page checks availability. Opening it does not upload your plan or start a purchase; save there explicitly. USDT payment only.