Brian Carpizo published a piece called "Agile Is Dead. AI Killed It. Welcome Back, Waterfall." He is right about the diagnosis and wrong about the destination. That distinction is the whole answer.
The diagnosis is correct
Agile was a rational response to two constraints that were both real in the 1990s. Humans were bad at comprehensive planning, so we could not write a useful 200-page spec at the start of a project. Building was slow and expensive, so we could not afford to be wrong at scale. Given those two constraints, "plan less, course-correct more" was the right move. Both constraints are gone now, not evolving, gone. The planning constraint broke when AI got good enough to act as a tireless system-architecture partner. I can sit down with Claude at 9 a.m. and have a full architecture document by lunch: data model, API contracts, dependency map, eval rubric. The build-speed constraint broke when prototypes got fast. I have sat with a customer at 10 a.m. and shown them a working prototype of their request at 11. Take those two away and the ceremony economy, story points, planning poker, sprint reviews, the Scrum Master, the Product Owner as buffer, becomes overhead with no remaining justification.
Why "welcome back, waterfall" is the wrong frame
The waterfall disasters of the 1990s were not a planning problem, they were a planning ability problem. Teams tried to write comprehensive specs and got them wrong because humans, working alone in a slow-feedback environment, cannot hold a complex system in their heads coherently for months. AI removes that exact constraint. What replaces Agile is the spec-first loop: write the spec WITH AI in hours, treat the document as a context payload rather than a contract, hand it to AI tools as the primary input for code, tests, infra, and customer briefings, build the first prototype the same week, put it in front of real users by week two, then iterate with the spec as a living artifact. Waterfall had no working software for months. The spec-first loop has working software in days, plus the architectural coherence Agile teams gave up on because the planning step was too painful.
What dies and what survives
What dies is the ceremony economy and the planning aversion. Estimation theater goes, because story points were always a polite fiction and AI removes the constraint that made imprecision tolerable. The PM-as-translator role goes, because a senior engineer with Claude can produce a stakeholder-ready spec in an afternoon. The "we'll figure it out next sprint" habit goes, because AI is unforgiving: give it a half-formed thought and it gives you a half-formed system at scale, fast. What survives is the customer feedback loop, the empowered team, and the bias toward working software. That is not a small reorganization. It is a different operating model.
For the PM, the version that survives and gets more leverage is the PM as specification author and outcome owner: the person who writes the system-level spec with AI, decides what to build, articulates it crisply enough for AI to build, and owns the metric that says whether it worked.
Try one thing this week. Pick the next non-trivial feature on your backlog. Before sprint planning, sit down with Claude and write the actual spec, architecture, data model, API contracts, user journey, the metric you will watch. Time-box it to half a day. Then take that document to engineering instead of the usual story breakdown. You will learn either that the team ships 3x faster with a real spec, or that your team's actual bottleneck was never in story-point estimation. Either result is valuable.