Harmonic, a start of AI co -founded by Robinhood Vlad Tenev CEO, announced Monday the launch of Beta of an iOS application and Android Chatbot, where users can access Aristotle’s AI model.
With this launch, the company aims to expand access to Aristotle, which harmoniously claims to offer “no illusions” answers to questions related to mathematical reasoning-a bold claim, given the credibility problems of today’s AI models. Harmonic focuses on creating “Mathematical Hypernum” or MSI. Startup finally wants to help users with all mathematics -based fields, including physics, statistics and computer science.
‘[Aristotle] It is the first product available to people who are reasoning and formally verifying production, “CEO and co -founder Tudor Achim said in an interview with TechCrunch.
In the end, Harmonic says she is also planning to release an API to let businesses have access to Aristotle, as well as a web application for consumers.
Harmonic says Aristotle achieved the performance of the gold metal at the Math 2025 International Olympiad (IMO) through a formal test (which means that the problems were translated into machines readable). Google and Openai also developed AI models that achieved the performance of the gold metal in this year’s IMO, but through informal tests in the natural language.
Harmonic says he does not release other points of reference for Aristotle right now.
Aristotle’s Beta launch has come just weeks since Harmonic raised $ 100 million in a B -series round, led by Kleiner Perkins at $ 875 million. Achim claims that Harmonic is “going very quickly” in his career to achieve MSI and that investors believed it was a fair assessment given the scope of implementation of the start of the start.
TechCrunch event
Francisco
|
27-29 October 2025
Many top technology companies focus on training AI models to solve mathematics problems. The AI that can make mathematics is valuable in itself, but mathematics is also considered a unique verifiable sector that requires basic reasoning skills. The systems that develop these capabilities can prove useful in other areas.
Achim says Harmonic is achieving its over-accurate solutions, having Aristotle to produce answers to open source programming language. Before Aristotle gives an answer to users, the dual control model says that the solution is correct through an algorithmic process that does not include AI. Harmonic’s chief executive notes that similar technology is used to verify results in high -stage fields, such as medical devices and aviation.
Even in a narrow field, achieving without illusion of an AI model is an incredibly difficult task. Studies have found that even AI models give a lot, and the problem does not seem to improve. The latest AI logic models of Openai remain longer than its older ones.
