Kuzco

Complete SentientWave’s node for autonomous agent network. It runs multi-modal transformers, databases, streams graphics, and participates in an internal computational marketplace. @nvidia Jetson AGX Orin, 64GB unified memory, 4TB flash. 275 TOPS INT8.

We’re thrilled to announce that BIOS, our new AI Scientist, just ranked as the best-performing agent for biological data analysis. This is a big accomplishment for our up-and-coming AI for Science lab. We've published a paper detailing the BIOS architecture and pushing the AI Scientist live for researchers to use. BIOS hit #1 on BixBench benchmark (all modes): • 48.78% open-answer • 55.12% multiple-choice + refusal • 64.39% multiple-choice (no refusal) Ahead of leading Agentic Science systems like Edison, K-Dense and Kepler. Beyond its raw performance, BIOS is built to speed up scientific iteration by introducing human-in-the-loop checkpoints that let researchers steer investigations mid-flight as insights emerge. This stands in contrast to the batch-processing modes of existing AI Scientists, which allow for human oversight only after time and compute-intensive runs finish - often forcing full reruns to explore new directions. BIOS, on the other hand, is designed to speed up science by slowing down: it adds checkpoints at key decision points, so humans can adapt research trajectories in-session and respond to intermediate findings. No more waiting hours for batch runs + reruns. 4 specialized agents power the system: • Orchestrator Agent: breaks complex problems into actionable steps, routes them to the right specialist agents, and queues the chosen next steps. • Literature Agent: searches + synthesizes across papers, patents, clinical trials, and scientific databases (ArXiv, PubMed, CrossRef, Semantic Scholar, Google Scholar, etc.) • Data Analysis Agent: autonomously writes and runs Python to analyze datasets, produce figures, and report quantified findings. • Novelty Detection Agent: checks whether proposed hypotheses have been previously investigated You can run BIOS in semi-autonomous mode with checkpoints or fully autonomous mode for your deep research task to run without interruption until completion. Either way, you stay in control of the research direction. BIOS is built on top of BioAgents, our open-source framework for building scientific agents. In the future, BIOS can eventually provide the orchestration layer that coordinates specialized subagents. The core infrastructure of the BioAgents framework is public, and we welcome researchers and developers to build their own scientific agents or subagents, or contribute to ours. The Literature and Data Analysis agents are not yet open source, but we're working on providing API access in the near future. BIOS is available now with free + paid tiers and exclusive launch pricing. Try it in the link below. For a limited time, we're offering free, full access to academic users with a .edu email address. Learn more below ↓

I started playing chess against #AI models a while ago. It's been quite enjoyable and has helped me improve. The @grok model is especially fun. Later, I decided to code it. I compare random AI models and have them play against each other. It's quite challenging but enjoyable. Tomorrow, I'll start streaming on X live. #Chess #grok #DeepSeek #claude #ChatGPT

$TODDY is live on @BagsApp bags.fm/1gsvLfa3vetaw6…