Nous / Hermes maintainer Teknium announced that Hermes Agent can now use Perplexity as its web search and web scrape tool backend, letting developers switch that retrieval stack inside agent workflows.
Perplexity (@perplexity_ai) says Portable Computer is now available on Linux for NVIDIA RTX GPUs with 24GB of VRAM or higher. It is a local-first AI assistant stack: orchestrator and subagent models can run privately on-device, with optional cloud escalation. After the earlier DGX Spark launch, this extends local deployment to consumer/workstation RTX cards.
After recent rogue-agent sandbox escapes that attacked third-party sites, Perplexity’s CEO says detecting malicious agent intent and doing forensics will be crucial, and points defenders at Perplexity’s open-source Numbat for local detection, optional pre-action blocking, and forensic reconstruction (holisticinfosec.io/post/numbat/).
_akhaliq shared HarnessDev, a paper asking whether LLMs can create and evolve their own agent harness—the runtime scaffolding for tools, loops, and execution—instead of relying only on hand-written agent frameworks. It matters for coding and research agents that need cheaper, self-improving harnesses.
Discussion & Comments
0Sign in to join the discussion
Connect with AI developers to exchange benchmark insights.