Google’s New AI Agent Jitro Is Replacing Entire Teams (Here’s How to Profit)
The video introduces Google's leaked internal AI coding agent 'Gitro,' described as an upgrade to Jewels that operates autonomously toward high-level goals rather than waiting for individual prompts. The presenter argues this 'KPI-driven automation' philosophy applies beyond coding and can be used by non-technical people to build income by selling outcome-based services. The video also promotes a free AI cash flow masterclass.
Summary
The video opens by teasing a leaked internal Google project called Gitro, framing it as fundamentally different from existing AI tools like Copilot or Cursor. The presenter, an AI avatar for Nick Pontis of Mind Marketing, explains that most current AI tools operate on a prompt-and-response loop — essentially a 'fancier Google search' — and argues that most users are stuck in this reactive pattern without realizing it.
The presenter provides background on Jewels, Google's existing autonomous coding agent, which launched in public beta and helped developers produce over 140,000 publicly shared code improvements. While Jewels was notable for handling background coding tasks like bug fixes and test writing, it still required users to prompt it for each individual task. Gitro, framed as Jewels V2, is described as a philosophical departure: instead of telling the AI what to do step by step, users define a desired outcome — such as 'reduce errors by 20%' or 'increase sign-up clicks' — and Gitro autonomously plans and executes the path to get there, running persistently in the background without constant supervision.
The presenter draws a business analogy: Gitro is like hiring someone and handing them an end goal rather than a micromanaged task list. He critiques the common 'calculator' approach to AI — punching in questions and getting answers one at a time — calling it reactive and unscalable. He argues that people building real income with AI are instead constructing systems that chase measurable outcomes automatically.
The presenter then pivots to broader income opportunity, emphasizing that while Gitro itself is a coding tool, its underlying philosophy of KPI-driven, goal-oriented automation applies directly to marketing, SEO, email, ads, and other business services. He argues that selling outcomes rather than hours or tasks is the key differentiator for those building AI-based service businesses, and credits this mindset shift as the turning point in his own career.
On timing, the presenter notes that Gitro has not officially launched and is currently in waitlist mode, but flags Google I/O 2026 (starting May 19th) as the likely announcement stage, with agentic coding and Gemini updates confirmed as focus areas. He frames early awareness as a competitive advantage.
The video concludes with three action items: think in outcomes not tasks, monitor Gitro's official launch closely, and recognize that the broader AI industry is committing to an agent-first direction. Multiple plugs for a free AI cash flow masterclass and a 30-day trial of the presenter's preferred AI software are woven throughout.
Key Insights
- The presenter argues that Gitro represents a fundamental philosophical shift from prompt-response AI — where users still control every individual decision — to goal-oriented AI, where users define a desired outcome and the agent autonomously plans and executes the path to achieve it without supervision.
- The presenter claims that most people who say AI is not working for them are treating it like a calculator — prompting it one question at a time — and compares this to 'using a BMW as a shopping cart,' arguing it is reactive rather than systematic.
- The presenter states that Google internally describes Gitro's approach as 'KPI-driven automation,' where the AI manages high-level goals rather than individual steps, and argues this principle applies to marketing, SEO, ads, and other business services — not just coding.
- The presenter notes that as of the video's release, Gitro has no public web interface and is only in pre-launch waitlist mode, but positions Google I/O 2026 on May 19th as the expected moment of official announcement, with agentic coding confirmed as a focus area.
- The presenter claims that Jewels, Gitro's predecessor, already helped developers produce over 140,000 publicly shared code improvements during its beta phase, and that users described working with it as similar to instructing a smart junior team member who handles execution independently.
Topics
Full transcript available for MurmurCast members
Sign Up to Access