Get ready for a groundbreaking AI revolution! Simular, a startup with a unique vision, has secured a $21.5 million Series A investment led by Felicis, with NVentures and other prominent backers. But here's the twist: Simular's AI agent isn't just about browsing; it aims to take control of your entire computer!
Simular's co-founder, Ang Li, reveals their ambitious goal: to create an AI that can physically move the mouse and click, mimicking human actions. But this isn't just a futuristic concept; it's already in motion. Simular's 1.0 version for Mac OS is out, and they're collaborating with Microsoft to bring a Windows version to life. And this is where it gets controversial—Simular is one of five companies chosen for Microsoft's Windows 365 for Agents program, sparking debates about the future of AI-human interactions.
The founders' expertise adds credibility to their mission. Li, a continuous learning scientist, and Jiachen Yang, a reinforcement learning specialist, met at Google's DeepMind. Their work, while academically published, had a practical focus on improving Google products. But the road to an agentic future is riddled with challenges, especially the issue of AI hallucinations.
AI agents, tasked with complex sequences, can fail if they hallucinate at any step. Simular's innovative solution? Let the AI explore and learn, with human oversight. Once a successful path is found, it's locked in as deterministic code. This approach ensures creativity and reliability, a delicate balance in AI development.
Simular's secret weapon is its 'neuro symbolic computer use agents' technology, a non-LLM-based approach. By writing deterministic code, they empower users to trust and understand the AI's actions. Early adopters, from car dealerships to HOAs, are already automating tasks. With a total funding of $27 million, Simular is poised to redefine the relationship between humans and AI. But will this vision truly bring AI agents to every worker's desk? The debate is open, and the future of work hangs in the balance.
