Edgerunner AI CEO Discusses Military Applications of AI

Tyler Saltsman, the CEO of Edgerunner AI, recently shared insights with a business outlet regarding the data utilized for training open-source AI models.

Last month, the news about four AI firms—OpenAI, Anthropic, Google, and Xai—landing government contracts with the Pentagon stirred some skepticism from Saltsman. He highlighted concerns over the unique models these tech giants claim to have developed.

These companies, recognized for their chatbot technologies powered by large language models like ChatGPT and Claude, each secured $200 million contracts. The Defense Department plans to leverage these funds to address various combat and national security issues.

Lawmakers Emphasize Importance of AI Strategy for National Security

Saltsman pointed out that with tools like ChatGPT or Grok, every interaction online can potentially expose sensitive data. “If you’re connected to the internet and you pose a question or share data, it’s all collected,” he noted.

Having a background as an Army officer and experience in military logistics, Saltsman is no stranger to the tech world. Previously, he was with Stability AI, which has made strides in services like Amazon Web Services and creating models that bridge text and images.

Edgerunner AI aims to enhance security by operating offline to mitigate risks related to sensitive data exposure.

Elon Musk’s Xai Competes with ChatGPT

Saltsman further mentioned the unique positioning of Edgerunner AI, stating, “We adapted these models to military specifics, allowing them to run on compact chips—eliminating the need for extensive infrastructure.” He characterized the models as an open-source solution but admitted, “While we can analyze how they function, we can’t access the underlying architecture of others’ chatbots.” This adaptability is intentional, allowing the models to be integrated into military gear effectively.

Compared to broader chatbot systems, which draw data from websites, articles, and user interactions, Edgerunner AI’s offerings are meticulously tailored.

Nvidia CEO Addresses AI Position Changes

Saltsman articulated that the drawbacks of generalized platforms like ChatGPT and Grok lie in their failure to address military-specific needs. “We essentially rethought these models from a military perspective,” he remarked, drawing a contrast between American and NATO approaches.

Collaborating with Meta, Edgerunner AI is processing significant volumes of specialized data, including military tactics and philosophies that shape Western civilization.

US AI Competition Against China

In an illustrative comparison, Saltsman remarked on the limitations of mainstream chatbots. “When prompted similarly to us, ChatGPT often declines, citing service restrictions,” he explained. This can render them impractical for military use.

Moreover, he brought attention to the risks of data transmission over the internet, where malicious actors could exploit vulnerabilities. “The potential for inaccuracies and harmful outputs is concerning, given the vast amount of unreliable data online,” he added.

Saltsman believes that AI should provide concrete solutions to real-world problems. “Right now, while AI appears cool and innovative, it often misses the mark in solving business challenges,” he said.

With capabilities to assist in various military tasks—from operational orders to supporting generals—Edgerunner AI is positioning itself as essential for modern military needs.

Leading with AI for Future Success

Saltsman recounted how generative AI could streamline routine planning, saying, “I used to spend a large chunk of my time on monotonous paperwork. Now, I can focus more on the soldiers.” He emphasized that this technological aid enhances battle readiness.

Edgerunner AI uses specific codes to assist military personnel efficiently. “If someone falls out of action, the AI can efficiently guide others in using complex weapon systems,” he noted.

OpenAI has recently unveiled plans to develop tools specifically for government use, including ChatGPT Gov, aimed at enhancing civil servant efficiency.

Calls for Transparency in AI Development

Saltsman expressed skepticism about claims of entirely open-source models produced by other companies. “They’ve created something using proprietary datasets, but it’s not entirely visible to us,” he commented, recognizing the competitive edge such data provides.

While he sees potential in platforms like ChatGPT, he is also proud of Edgerunner AI’s collaboration with U.S. Special Operations Command. “The absence of an internet connection is a unique advantage,” he noted. His vision? To create tech reminiscent of Jarvis from Iron Man, potentially revolutionizing military operations.