
Perplexity Commits Computing Power to Autonomous Research Initiative
Perplexity launched a new initiative backed by thousands of processors to build fully autonomous scientific research models. The company plans to share its weights, datasets, and testing metrics openly with the public.
The Blend
Search startup Perplexity has established a new internal unit dedicated to creating automated scientific discovery systems. As announced by Perplexity on X, the initiative is backed by more than ten thousand specialized graphics processors and focuses on building software that can independently plan, execute, and evaluate complex research projects.
For everyday consumers, this initiative represents a potential shift in how artificial intelligence operates. Instead of merely summarizing web pages or answering simple questions, future tools could actively conduct independent investigations to find new insights. By promising to release its models, data sets, and testing metrics openly, the company is also giving independent researchers free access to sophisticated tools that are typically locked behind proprietary paywalls.
It remains to be seen whether a fleet of ten thousand chips is enough to compete with multi-billion-dollar computing clusters owned by larger rivals. Additionally, it is unclear if the firm will train its base systems entirely from scratch or rely on existing third-party open models as a foundation.
Written independently by AI News Smoothie from the reporting listed below. Facts belong to the original publishers. Follow the links for their full coverage.
Ingredients
- Perplexity on X: "Today we're launching Perplexity Frontier Research.
The goal of Perplexity Frontier Research is to create a system that can perform autonomous research.
The work will focus on multi-agent orchestration, reinforcement learning, continual learning, systems, and pre-training.
We w… / X
Perplexity is forming a research lab backed by thousands of graphics chips to create open systems capable of independent scientific inquiry.