Gemini 3, GPT-5.1, Anti-Gravity & Yann LeCun’s Exit: Are We Near AGI or Just in a Bubble?

About this episode

Youtube Channel: https://www.youtube.com/@GenerativeAIMeetup

Mark's Travel Vlog: https://www.youtube.com/@kumajourney11

Mark's Personal Youtube Channel: https://www.youtube.com/@markkuczmarski896

Attend a live event: https://genaimeetup.com/

Shashank Linked In: https://www.linkedin.com/in/shashu10/ 

 

In this episode of the Generative AI Meetup Podcast, Mark (in Ohio) and Shashank (in India) finally sit down after a month of travel to unpack a very eventful stretch in AI. They dive into Google’s new Gemini 3 Pro, its standout scores on Humanity’s Last Exam and ARC-AGI, and why these reasoning benchmarks matter more than yet another near-perfect standardized test score. Mark also makes a public feature request to DeepMind: please increase Gemini’s max output tokens.

From there they get hands-on with the developer experience:

• Google’s new Anti-Gravity coding IDE (and how it compares to Cursor)

• Using GPT-5.1 Codex High in Cursor’s autonomous “plan mode”

• Why long context and long output windows are critical for deep research and book-length projects

The conversation then shifts to the bigger picture:

• LLMs as therapists, sycophancy, safety, and the danger of AI always agreeing with you

• Mark’s rant on robotics, humanoid robots, and a coming age of extreme abundance where robots handle most physical and intellectual work

• Why learning to code may become the mental equivalent of going to the gym—a “brain gym” in a world where AI can do most practical tasks

They also cover the latest AI industry drama and milestones:

• Yann LeCun leaving Meta, what that might signal about Big Tech AI labs, and how godfathers like Hinton, LeCun, and Bengio see the road to AGI

• DeepMind’s new game-playing agent and why world models in 3D environments matter for real-world robotics

• Genspark hitting unicorn status and what it means for “ChatGPT wrapper” startups

• Co-inventing a new term on air: a “narwhal” = a trillion-dollar private company

If you’re curious about where frontier models, coding agents, robotics, and AGI trajectories all intersect—plus some philosophical musing on jobs, meaning, and abundance—this episode is for you.