The Programming Podcast

Did Google Just KILL Cursor with Gemini 3 Pro?!?

Nov 19, 2025
Google's Gemini 3 Pro and Antigravity are shaking up the AI landscape. They might challenge OpenAI's dominance with groundbreaking benchmarks like Humanity's Last Exam and the Vending Bench. Antigravity isn't just a tool; it's revolutionizing how developers work, turning them into 'agent managers' and automating much of the coding process. The discussion also touches on the brewing AI War among tech giants and the need for developers to embrace these technological shifts to stay relevant. This episode is filled with strategic insights and future predictions.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
INSIGHT

Gemini 3 Pro Leap In Reasoning

  • Gemini 3 Pro massively improved reasoning benchmarks, jumping from ~21% to ~45.8% with tools enabled.
  • This suggests a real step-change in long-horizon and tool-augmented problem solving by LLMs.
INSIGHT

Benchmarks Show Business Value

  • The Vending Bench shows agentic long-term planning delivering real dollar value improvements.
  • Gemini 3 Pro earned $5,478 versus GPT 5.1's ~$1,400, indicating superior multi-step strategy.
INSIGHT

Gemini's Stronger Contextual Intuition

  • Gemini distinguishes audience and intent more strongly than competitors in text tasks.
  • That contextual intuition makes it useful for UX and copy decisions without explicit prompting.
Get the Snipd Podcast app to discover more snips from this episode
Get the app