Mirendil co-founder and CEO Behnam Neyshabur said the company's long-term goal is to create AI that can independently advance its expertise while working on complex scientific problems. "You can have a self-improving AI where you can point a problem at it and it keeps getting better with time," Neyshabur said.
The partnership provides Mirendil with a combination of Google's Tensor Processing Units, Nvidia graphics processors, and managed training infrastructure. The company said its software is designed to assign different workloads to the hardware best suited for each task, allowing it to use multiple accelerator types within a single training environment.
Co-founder Harsh Mehta said that approach helps reduce costs while improving efficiency.c"These models are really good at working with different workloads and chips, and assigning the right workloads to the right chips," Mehta said.
He added that Google's infrastructure allows Mirendil to "mix and match workloads with the right kind of accelerators, and then lower the cost not just for us, but also for our customers using our systems."
Mirendil's founders previously worked at Anthropic, and the startup is pursuing research centered on recursive self-improvement, where AI systems iteratively enhance their own performance instead of relying solely on traditional model training cycles. The company said it wants to automate research workflows that could otherwise require teams of scientists.
"How can we have an AI system that keeps doing research, keeps improving its own knowledge and performance when it comes to Alzheimer's disease?" Neyshabur said.
Google also highlighted the partnership as an example of its AI infrastructure strategy. Amin Vahdat, Google's Senior Vice President and Chief Technologist of AI and Infrastructure, said the agreement reflects the company's approach to supporting advanced AI workloads with a combination of hardware and software.
The partnership comes as cloud providers increasingly supply large-scale computing resources to AI startups developing frontier models. For Mirendil, the agreement provides the infrastructure needed to continue its research into self-improving AI systems, while Google Cloud expands its portfolio of customers building advanced AI technologies on its platform.
This analysis is based on reporting from the tech buzz.
Image courtesy of Mirendil.
This article was generated with AI assistance and reviewed for accuracy and quality.
