Magic AI's technical strategy involves handling large context windows, with their Long-term Memory Network (LTM Net) architecture and corresponding LTM-1 model boasting a 5 million context window. This large context capacity allows for more nuanced code comprehension. Despite operating in stealth mode, the startup claims to have thousands of GPUs deployed for training its next-gen models. With the new funding, Magic AI is focusing on talent recruitment and retention, seeking individuals who align with its vision of integrity and innovation.
Key takeaways:
- Magic AI, a San Francisco-based startup, has raised $117 million in Series B funding to develop its advanced AI system for automating software development. The total funding to date is over $145 million.
- The startup is working on an AI software engineer that can assist with complex coding tasks and act more as a coworker than a simple tool. The founders believe this could pave the way for more expansive artificial general intelligence.
- Magic AI's technical strategy involves handling exceptionally large context windows. They have developed the Long-term Memory Network (LTM Net) architecture and corresponding LTM-1 model with a 5 million context window.
- Despite operating in stealth mode, Magic AI claims to have thousands of GPUs deployed for training its next generation models. With the new funding, the startup is focusing on talent recruitment and retention.