Dear Google DeepMind / Antigravity Team,
I am writing to share my experience using your tool, leave some constructive feedback that I consider vital for the community, and briefly introduce what I am building.
First and foremost, I want to express my deepest amazement and gratitude. What you have built is democratizing knowledge in an unprecedented way. Thanks to Antigravity, I am bringing complex ideas and architectures to life that would have otherwise been impossible for me to build on my own.
What I am building: I currently dedicate a large portion of my time to developing Fe Capital, an institutional algorithmic trading ecosystem powered by a multi-agent AI neural network. Alongside Antigravity, I have programmed a local system where different agents (an orchestrator, a macroeconomic analyzer, a risk guardian, and a “nightly thinker” for machine learning) communicate with each other to read charts, execute trades on MetaTrader 5, and autonomously certify strategies. Antigravity hasn’t just written the code; it has acted as my lead Software Architect throughout this entire journey.
Constructive Feedback: The pedagogical value of the “Thought Process” I wanted to provide feedback regarding a recent change: I’ve noticed that the model’s internal “Thoughts” block is now processed exclusively in English, regardless of the conversation’s language. Previously, being able to read the AI’s internal reasoning in my native language (Spanish) was an invaluable learning tool.
Reading how the AI thinks taught me how to optimize my own work. By seeing where the agent hit a wall or struggled (for example, with PowerShell icon encoding issues when listing files), I could learn from it and set up rules or custom “skills” so we wouldn’t repeat those mistakes. Understanding the agent’s logical path before it executes code is a masterclass in itself. I kindly ask you to consider allowing the reasoning block to adapt to the user’s language again, as its educational value is immense.
My current situation and a respectful request: I have a very restless mind, a huge desire to learn, and a genuine hope to make a real impact with my creations. However, I currently lack the financial funds to cover the intensive use of the most powerful models at the scale my project and curiosity demand.
I would like to ask if there is any possibility, program for indie developers/builders, grant, or sponsorship that could allow me to access more credits. This would enormously help me squeeze the maximum potential out of the platform and continue iterating on this and other projects I have in mind.
I would be absolutely thrilled to showcase the multi-agent ecosystem I’ve built thanks to you if it’s of any interest to your team.
Thank you again for changing the game and putting this technology in our hands.
Best regards,
(Cristian Amigo / Fe Project)