Tech Disruptors

“Not everything is an AI problem,” says IBM Software SVP and Chief Commercial Officer Rob Thomas. In this episode of the Tech Disruptors podcast, Thomas and Bloomberg Intelligence senior technology analyst Anurag Rana discuss IBM’s hybrid cloud and AI strategy, from Red Hat/OpenShift and containers for multicloud portability to why the mainframe remains the platform for real-time, high-availability transactions. Thomas outlines IBM’s “new enterprise stack” and a layered, multimodel approach where orchestration and proprietary enterprise data matter more than any single frontier model — especially as sovereignty requirements rise. They also explore how system-of-record companies are best positioned to succeed in the AI era.

AI demand is scaling and infrastructure complexity is rising. Vultr CEO JJ Kardwell returns to the Bloomberg Intelligence Tech Disruptors podcast with an update on the market’s AI cloud demand. He spoke to BI tech analyst Woo Jin Ho about production AI workloads, GPU utilization and lifecycle economics, global data-center strategy, supply-chain constraints and capital discipline, as well as outlining how privately held Vultr is positioning for durable growth in the industry’s next phase.

“If anyone’s going to disrupt Khan Academy, it should be us,” founder and CEO Sal Khan tells Bloomberg Intelligence Senior Tech Analyst Anurag Rana, discussing how AI can deliver personalized learning at scale if embedded in classrooms with teacher oversight, guardrails for minors and rigorous model evaluation. Khan explains Khanmigo’s early GPT-4 roots, why Khan Academy is going multi-model to match use cases like Writing Coach and how district packaging helps cover compute costs while enabling monitoring and accountability. He also lays out a vision of the 2030 classroom where AI reduces teacher planning and grading burdens, supports small-group instruction and enables richer assessment, while warning workforce disruption may arrive faster than society is prepared for.

Lightmatter CEO Nick Harris joins Bloomberg Intelligence’s Kunjan Sobhani and Jake Silverman to explain why lasers — not just chips — are the missing piece to making co-packaged optics practical at hyperscale. Harris unpacks Guide, Lightmatter’s VLSP light engine, and Passage, the company’s photonic interconnect platform, walking through real-world reliability, density and power trade-offs, and how new EDA and foundry partnerships (Synopsys, Cadence, GUC) move photonics into standard semiconductor workflows. The conversation covers near-package vs co-package optics, deployment timing, who the early buyers will be, and the milestones to watch as photonics shifts from lab demos to production racks.

Russia invasion of Ukraine “has completely reframed perceptions of drone utility and value. Where once high-end, exquisite systems dominated the procurement logic, today there’s a much clearer recognition that quantity and replaceability are just as critical,” says AV Chief Growth Officer Church Hutton. On this episode of the Tech Disruptors podcast, Hutton tells Bloomberg Intelligence senior defense weapons analyst Wayne Sanders how the landscape of drone technologies and the scalability of platforms must balance between rapid production and programs of record, while maintaining a technological advantage over adversary weapon systems. Critical to this process is reducing “cost per effect” so that we’re no longer shooting down $10,000 drones with million-dollar interceptors. AV and other defense tech companies continue to have this as their mission statement.