At Kalshi, a U.S.-regulated prediction-market platform, an internal AI named Harrison is automating tasks traditionally performed by analysts. This AI agent tracks news, monitors competitors, drafts contract language, and resolves market closures, marking a shift in how AI is integrated into business operations.
According to Gartner, agentic workflows, like those performed by Harrison, consume 5-30 times more tokens per task compared to typical generative AI queries. Goldman Sachs forecasts that the monthly token count for agentic AI applications could reach approximately 120 quadrillion by 2030, indicating a growing demand for computational resources.
The rise of agentic AI is transforming the inference compute landscape, with companies like Nvidia and AMD positioned to benefit significantly. As AI demand shifts from sporadic use to a persistent need, the infrastructure—covering accelerators, storage, networking, and cooling—is vital in supporting this transition.
5 Stocks Our Experts Predict Could Double In the Next Year
By submitting your email, you'll also get a free pivot & flow membership. A free daily market overview. You can unsubscribe at any time.
