Skip to main content
Latent

Latent

Premier AI engineering newsletter and technical podcast

About Latent

Latent Space is a leading AI engineering newsletter and podcast with over 177,000 subscribers, hosted by swyx and Alessio. It delivers in-depth technical content on how top AI labs build agents, models, infrastructure, and AI for science. The platform features interviews with AI luminaries including Greg Brockman, Andrej Karpathy, George Hotz, Simon Willison, and Soumith Chintala. Content includes long-form podcast episodes, weekday AI news roundups through AINews, and comprehensive technical articles. The community also offers Discord channels, meetups, and opportunities for guest writing. Episodes typically run 60-90 minutes and cover cutting-edge topics like world models, frontier research, and AI engineering practices.

Our Review

Latent Space stands out as one of the most technically rigorous AI resources available, targeting serious AI engineers and researchers rather than casual enthusiasts. The quality of guests is exceptional, featuring key figures from OpenAI, major AI labs, and influential builders in the space. Episodes like 'Extreme Harness Engineering for Token Billionaires' and interviews with Marc Andreessen demonstrate their access to exclusive insights. The podcast balances technical depth with entertainment value, making complex topics digestible without dumbing them down. The weekday AINews roundups provide consistent value for staying current. However, the content can be quite dense and time-intensive, with episodes exceeding an hour regularly. The newsletter format works well for those who prefer reading transcripts. The community features (Discord, meetups) add networking value beyond just content consumption. For AI engineers and technical practitioners, this is arguably essential listening, though beginners might find the technical level challenging.

Pros & Cons

Pros

Exceptional guest lineup from top AI labs and frontier researchers
Deep technical content that doesn't oversimplify complex topics
Multiple formats (podcast, newsletter, video) with 177K+ engaged subscribers
Active community with Discord and meetup opportunities
Regular weekday AI news updates keep subscribers current

Cons

Content can be very technical and inaccessible for beginners
Long episode lengths (60-90+ minutes) require significant time investment
Not an interactive AI tool but rather educational content
Limited hands-on tutorials or practical implementation guides

Best For

AI engineers and ML practitioners seeking technical depthResearchers wanting insights from frontier AI labsTechnical leaders tracking AI industry developmentsDevelopers building production AI systemsAnyone serious about understanding AI engineering practices