It's not Claude Code vs. Cursor. It's Claude Code AND Cursor.
Cursor when you need to be in the code.
Claude Code, when your job is reviewing diffs and keeping the agent on track.
Right tool for the task = fewer tokens, less model bias.
Looking for this full session? Find it
Vibe coding feels fast. Then you realize you've lost control of what you're building.
AI-first product development is the opposite: plan clearly, prompt intentionally, and direct the tools instead of letting them lead.
Free Night School with Derek Peters, Professor at Gauntlet
"Anyone can change the world, regardless of their resume."
Meet a few of our Cohort 5 Challengers. Different backgrounds, different starting points, one thing in common: they bet on themselves.
Fired Festival judges and speakers, announced:
Judges will be @Austen Founder & CEO of Gauntlet AI
@vsodera Founder of @Superhuman, and Matthew Hartman from Factorial Capital
We will also have @marccampbell Founder & CTO of @replicatedhq speaking at the event
These aren't
Hiring managers aren't filtering for who can use AI tools. They're filtering for who's still in control of the codebase after AI writes the code.
Your workflow. Your evals. Your ability to improve the system — not just prompt it.
@ashtilawat (CTO, Gauntlet AI) breaks it down.
It’s hiring partner day 3 of 4 for this cohort at Gauntlet AI.
20 hirers in-house plus some companies dialing in remotely.
Thank you all for being here!
Cohort 6 kicks off July 6th.
The gap between "experimenting with AI" and building with it as second nature doesn't close gradually. It closes in a compressed burst of focused execution.
10 weeks. Full-time. Everything covered. Monday: assignment drops. Friday: you ship. No
Starting Thursday I’m going to be giving a few people tours of Gauntlet AI in downtown Austin.
Anyone who steps in the building gets what the future will look like instantly. An instant sense for what the cutting edge of AI engineering is.
If you’d like to stop by DM me.
Fine-tuning sounds like the answer.
Often it's the wrong first move.
This week's Night School: what fine-tuning actually changes, when it makes sense, and when prompting or stronger evals should come first. Plus, how PEFT and QLoRA make it efficient when you do need it.
Led by