{"version":1,"type":"rich","provider_name":"Libsyn","provider_url":"https:\/\/www.libsyn.com","height":90,"width":600,"title":"Ep. 268 AI at the Edge Turning Promise into Mission Advancement","description":"Today, we sat down with a group of experts who have experience taking AI out of the data center and putting it into the field. This enables fast, resilient decision-making in disconnected, bandwidth-constrained, contested, and high-consequence environments. Charles Ritter from Idaho National Labs provides an excellent use case. He begins by outlining the growing popularity of local LLMs and the quality gains products from Nvidia can deliver. Ritter\u2019s use case is managing a nuclear facility. When humans must make laboratory decisions, it can take years; an autonomous laboratory system can drastically reduce that time. Remote military situations have high demands that AI can meet. During the interview Dan Folliard from Astra AI details how AI can address restrictions on size, weight, and power. All participants agree that trust in AI must be a primary consideration. Folliard  details several methods to gain trust from stakeholders. Open architecture, modularity, model registries, experimentation, and small initial wins can accelerate adoption and build trust. Another way to encourage AI adoption at the edge is to use experimentation and small wins to build trust before scaling. ","author_name":"Feds At The Edge by FedInsider","author_url":"http:\/\/fedinsider.libsyn.com\/website","html":"