Iran conflict: ai's defining moment, says palantir cto
The ongoing conflict in Iran is rapidly solidifying its place in history, not for its geopolitical maneuvers, but as the first major war fundamentally shaped by artificial intelligence, according to Palantir Technologies’ CTO, Shyam Sankar. His assertion, delivered at the Hill and Valley Forum in Washington, D.C., suggests a significant shift in modern warfare – one that’s already underway.
The bureaucratic reality behind 'targeted strikes'
Sankar, speaking amidst an event shadowed by the escalating tensions, likened the impact of AI on targeting precision to streamlining complex processes like supply chain management, which previously consumed months but now can be expedited. He challenged the public’s perception of “targeted strikes,” labeling them as a “very bureaucratic and large-scale doctrinal process,” a process now significantly accelerated through the deployment of AI-powered tools.
Palantir’s Maven platform, a digital mission control system that manages data to improve decision-making, is a critical component of the U.S. military’s efforts in the region. Recently elevated to an official defense program, Maven’s status guarantees continuous funding and development, a clear signal of the Pentagon's commitment to AI integration. However, the use of AI in target selection hasn't been without scrutiny, particularly after Anthropic PBC’s relationship with the Pentagon faltered over concerns regarding fully autonomous weapons systems.
The Pentagon’s recent designation of Anthropic as a risk to the supply chain, stemming from the company’s insistence on security measures for military AI applications, introduces a potential complication for Palantir. The company currently utilizes Anthropic’s Claude product within its Maven platform, and may now be forced to seek alternative suppliers.
A darwinian approach to military tech
Sankar’s vision for the U.S. military doesn't stop at current deployments. Drawing a parallel to the sheer volume of aircraft designs during World War II – 154 different fuselages built, with only a handful proving truly impactful – he advocates for a more experimental approach. “We should be deploying 154 different variants of the LUCAS now,