THOUGHT OF THE DAY
Agentic AI is creating a new competition for commercial discoverability. Omnicom’s warning that agents will increasingly research and act on consumers’ behalf shifts the focus from persuading people directly to making brand information accurate, consistent, and machine-readable. This is a distinct layer of the agentic-AI opportunity: brands that agents cannot reliably interpret risk losing visibility before a consumer makes a choice.
Token use is providing a harder signal of AI demand. Bernstein cited a 50% quarter-over-quarter increase in enterprise token consumption at major AI labs, a catalyst for renewed interest in Datadog’s AI workload monitoring. If usage persists, observability and cost controls could become more valuable as AI moves deeper into production; the open question is how much of this lab usage converts into durable, broad-based enterprise revenue.
COMPUTE & SEMICONDUCTORS
Intel’s reported foundry commitment from Tesla for its 14A process offers a potential anchor customer for the manufacturing roadmap. It improves the commercial credibility of Intel’s foundry push, but does not resolve execution risks around process delivery, yields, timing, or profitability. The news provides no volume or schedule detail.
ROBOTICS & PHYSICAL AI
BlackBerry launched QNX Launchpad, offering startups and smaller businesses access to safety-certified operating-system and hypervisor software for robotics and other mission-critical systems. The move could bring QNX into more developers’ platforms earlier, but free access is an ecosystem-building signal, not evidence yet of scaled deployments or material revenue.
ADOPTION & MONETIZATION
Bernstein’s estimate of 50% quarter-over-quarter growth in enterprise token consumption at major AI labs is a tangible usage signal. Datadog benefits if AI workloads increase demand for monitoring performance, latency, and cost, though usage growth at labs does not by itself establish broad enterprise monetization.
POSITIONING IDEAS
Bullish
- Datadog (DDOG) — Higher reported token consumption at major AI labs supports demand for tools that monitor AI workload performance and cost. The upside case depends on that activity translating into sustained customer usage and revenue.