Prominent AI talks, explainers, interviews, courses, and video channels worth following.
A practical walkthrough of the transformer and language-model mechanics behind modern AI systems.
Useful for understanding what is actually inside a GPT-style model before comparing frontier releases.A visual explanation of attention, the core mechanism behind transformer models.
Good for readers who want to understand why context, tokens, and attention dominate AI model design.A high-signal talk on how GPT-style systems are trained, tuned, evaluated, and used.
Helpful context for model releases, post-training, evals, and why product behavior differs from base model capability.A clear foundation for neural networks, activation, layers, and learned representations.
Useful for non-specialists following AI news who need the conceptual base before reading model and agent stories.Accessible summaries of papers in graphics, simulation, machine learning, and generative AI.
Good for spotting research ideas that are becoming broadly understandable.Fast-moving model releases, benchmark debates, and AI-industry explainers.
Useful as a public-sentiment signal, especially when a technical story crosses into mainstream AI discourse.Technical interviews with builders across models, agents, AI engineering, and open-source tooling.
Often surfaces practitioner vocabulary and implementation details before they become mainstream posts.Researcher interviews and long-form discussions around AI research, safety, language models, and science.
Good for deeper context when a paper, lab, or researcher starts shaping the AI conversation.Long-form interviews with AI researchers, founders, scientists, and technologists.
High-reach interviews can push technical AI debates into a broader audience.Consumer technology, AI products, company moves, regulation, and internet-culture spillover.
Useful for tracking which AI topics are entering mainstream technology conversation.Weekly AI research, policy, safety, industry, and infrastructure notes.
A compact editorial source for connecting research movement with governance and industry impact.AI news and applied learning from DeepLearning.AI.
Good for short summaries of developments that matter to learners and practitioners.Talks, lectures, and explainers can reveal which concepts are becoming common knowledge before those ideas show up in formal product pages or polished reports.
Channels with primary experts, clear technical framing, and repeatable educational value.Founder and researcher interviews are useful when they explain tradeoffs, constraints, timelines, and disagreements that press releases leave out.
Specific claims, citations, demos, roadmap hints, and named limitations.Pagish treats media momentum as a signal, not proof. A popular video can identify a trend to investigate, but publication still needs source-backed evidence.
Recurring topics across channels, comment velocity, derivative explainers, and follow-up discussion.3Blue1Brown — YouTube is listed in the AI source profile. Reference watchlist URL only; not fetched by runtime unless a structured feed or official API connector is added.
sourceAndrej Karpathy — YouTubeAndrej Karpathy — YouTube is listed in the AI source profile. Reference watchlist URL only; not fetched by runtime unless a structured feed or official API connector is added.
sourceDeepLearning.AIDeepLearning.AI is listed in the AI source profile. Reference watchlist URL only; not fetched by runtime unless a structured feed or official API connector is added.
sourceinterviewing.iointerviewing.io is listed in the Internet Culture source profile. Reference watchlist URL only; not fetched by runtime unless a structured feed or official API connector is added.
sourceStanford AI LabStanford AI Lab is listed in the AI source profile. Reference watchlist URL only; not fetched by runtime unless a structured feed or official API connector is added.
sourceStanford AI NewsStanford AI News is listed in the AI source profile. Reference watchlist URL only; not fetched by runtime unless a structured feed or official API connector is added.
sourceStanford CS229 — Machine LearningStanford CS229 — Machine Learning is listed in the AI source profile. Reference watchlist URL only; not fetched by runtime unless a structured feed or official API connector is added.
sourceStanford CS231n — CNNs for Visual RecognitionStanford CS231n — CNNs for Visual Recognition is listed in the AI source profile. Reference watchlist URL only; not fetched by runtime unless a structured feed or official API connector is added.
