Goldman Sachs TMT trader Peter Callahan says the post-earnings AI narrative
within software is shifting: fears that generative AI would erode traditional
software moats are yielding to a view that data infrastructure and
developer-tools vendors are moving from an AI headwind to a tailwind.
Cloudflare, Palantir, Datadog, Twilio and Atlassian have drawn greater investor
attention; traditional SaaS application vendors still need to prove clear AI
benefit. Callahan notes AI commercialisation is extending from model training
into inference, agents and automation. Cloudflare disclosed non‑human traffic
now exceeds human traffic and said machine‑generated requests should grow
rapidly if current trends persist. That suggests AI can be demand‑accretive for
platforms that carry data, APIs, network traffic, security and developer tools,
since more agents and higher call frequency can create direct revenue
opportunities. Software is therefore bifurcating: AI upside increasingly depends
on whether a company sits at the application layer or the infrastructure layer
and on its ability to monetise AI traffic.
