Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Anthropic recently told its growth team to hire more product managers, not fewer. The reason, as reported in industry coverage, was that Claude Code had quietly morphed its engineering organization into a team that supplied about three times the actual headcount, and the hurdle had moved from the integrated development environment (IDE) to the people deciding what to build.
This detail is easy to miss in the noise of any AI performance claim. This is also the structural change that the rest of the industry is going through. The software bottleneck is no longer writing. He decides what to write. And engineers who treat this as someone else’s problem are about to hit a plateau.
For most of the last decade, that decision was made by someone else. Software engineering was a craft you learned slowly, then practiced in a long, predictable sequence: dive deep into the technology, write the code, ask Stack Overflow when you’re stuck, escalate to a senior engineer when Stack Overflow fails, submit the ticket. The product manager owned the funnel. The engineer owned the building. Both sides treated this division as physics.
Then the funnel collapsed in five steps.
A brief history of how an engineer’s day is compressed
The Stack Overflow era (2014 to the end of 2022): The way engineers thought lived in one place. But new monthly questions on Stack Overflow have already decreased by approximately 77% since November 2022, which was no accident when ChatGPT was launched. Opting out is not a referendum on the site. It is a referendum on the work process it represents.
The Era of Browser Tabs (Late 2022 to 2024): The first generation of ChatGPT was outside of an IDE. The engineers ran the same cycle they’d always run, only with a faster oracle: Write a prompt in a browser, put the answer back into VS Code, repeat. Work was still single-flow and engineer-led. The leverage was real, but local.
Native era of IDE (2024 to 2025): Cursor and Claude Code moved the model inside the editor and gave it access to the full repository. The senior engineer’s escalation path has largely broken down. For years, the prevailing wisdom among veteran engineers was that Bash had the longest shelf life of any tool in the stack. By 2026, for a significant portion of working developers, the first command typed into a new terminal is claude.
The spec-driven era (2025 to 2026): Larger context windows have transformed single-session work into something that previously required tickets, design documents, and sprints. Amazon’s Kiro IDE team reportedly compressed feature builds from two weeks to two days using the same spec-driven workflow they were shipping. An AWS engineering team described an 18-month re-architecture, originally designed for 30 engineers, completed by 6 people in 76 days. The bottleneck is no longer how long it takes to write the code. It started with how clearly the team could describe what the right thing looked like.
The Era of Routines (2026): In April, Anthropic shipped Claude Code Routines: Scheduled, persistent agents that run on a cadence, on a webhook, or at night while the laptop is closed. Cron is back. The hooks are back. The engineer’s job is now part of the orchestration: Spin up a swarm before bed, review a stack of pull requests in the morning. Third-party wrappers like OpenClaw, which was briefly suspended by Anthropic in April before a partial recovery, have made the same point on the open source side.
The bottleneck moved; most teams are not
Engineering has roughly tripled. Product management hasn’t budged. The traditional 1:8 ratio of PMs to engineers, already strained, now approaches an efficient 1:20 because each engineer delivers more per day. For example, LinkedIn replaced its associate product manager song with a "Product creator" program that trains generalists in product, design and engineering. Anthropic is hiring more PMs, not less. The pattern is consistent across companies that have actually deployed agent workflows in production: The system produces built features faster than it produces decisions about what needs to be built.
For engineers, it’s the most important career signal of the decade and the easiest to miss while productivity stories dominate the feed.
First principles matter more, not less
The instinct to declare fundamentals obsolete in the age of agents gets the trend just wrong.
When a memory leak stops production at 3am and the cause turns out to be a subtle ownership bug pushed 4 years ago, no agent currently in the wild closes that loop end-to-end. Operating systems, networks, concurrency, and query plans still decide who can resolve a real incident. They also decide who can spot the moments when an agent’s output looks right on the surface and is quietly, expensively, wrong underneath. An agent who has written 70% of the code in a modern repo cannot reliably tell anyone where his assumptions about thread safety, memory ownership, or transaction isolation have deviated at runtime. The engineer who can read the difference and know that’s the engineer the rest of the team needs in the room, and that engineer is built on fundamentals, not tipping skills.
The bottom line is that the basics are now a leverage skill, not a hygiene skill. In 2014, knowing how TCP relay works, the debug ticket was closed faster. In 2026, the same knowledge prevents an entire agent-driven release pipeline from sending a regression at scale. The blast radius of the engineer who knows what’s going on below has increased, not decreased.
Reviewing is the new writing
Engineers in 2026 are generating code at a rate that exceeds what any of them can read carefully. The team that delivers fast and survives is the team whose engineers treat AI-generated code review with at least as much rigor as they once reserved for writing it. Stack Overflow’s 2025 Developer Survey put 84% of AI tool developers with 46% saying they don’t trust the output, a sharp increase from 31% the previous year. This gap, heavy usage combined with low trust, is exactly where review skills are now most important. Coders who push a lot and review little accumulate debt that will become due during the first real incident, and the engineer who can pay it back is the one who has combined their volume with a deep knowledge of the first principles of the systems involved.
The new differentiator is the product funnel
Both are necessary. Neither is enough. The engineer who matters in 2026 is the one who has stopped waiting for the funnel to arrive in the form of a Jira ticket.
This means doing things that the role has historically not allowed.
Talk to customers. See how they actually use the product. Read the support queue. Participate in the sales call. The signal that the product team receives through three layers of aggregation can now be obtained first-hand by an engineer in an afternoon.
Generate ideas, not just predictions. A product manager who previously generated ideas for 8 engineers cannot generate ideas for 20 with the same precision. The engineer who shows up with a scoped validated opportunity is no longer doing the PM’s job. The engineer does the work that the new ratio requires.
Work in the opposite direction from the client. Amazon is writing the press release for the first time in two decades. Discipline spreads well in teams of one and in swarms of agents. Both produce a large amount of working software in the wrong direction, without a clear statement of what "the customer wins" means before any code is written.
Stop hiding behind bandwidth. The honest answer to "Do you have the capacity for this idea?" it used to be « No ». With routines, hooks, and a cooperative stack of agents, the honest answer is closer to "What is the idea worth?" It’s a different conversation and much harder to have without a true customer perspective.
What will the next decade reward
The five phase story above is not really a tool story. It is a story about what part of the job a person was meant to do. The part that is still human and will remain human for the foreseeable future has moved up the funnel: from writing, to review, to decision making, to choosing a customer to serve and a problem to solve.
The 2026 version of the great engineer is not the one who writes the most code. He’s the one who knows what to build, can prove it’s worth building, and has the fleet of agents plus the review discipline to send it without the system collapsing at its own pace.
Engineers who internalize this will spend the next decade doing the most interesting work ever created by software. Engineers waiting for a ticket will spend it watching the ticket being written by the agent next to them.
Ishan Gupta is a software engineer at Amazon.
Infrastructure,DataDecisionMakers
#Claude #Code #turned #engineer #companies #product #thinkers
