TL;DR
Get the latest gadgets delivered free — and shop member deals
- Fast, free delivery on millions of items
- Access to Prime Big Deal Days deals on October 6–7
- Prime Video, Amazon Music and more included
A report based on a keynote to engineering leaders describes AI as a fast-moving force changing how software is built in 2026. Engineers are increasingly using multiple coding agents in parallel, while the report also flags concerns about code quality, reliability and review practices. The scale of adoption and its effects across the wider industry remain difficult to measure from the material available.
A 2026 report from The Pragmatic Engineer says AI coding tools are rapidly changing software development, with some engineers coordinating five to 10 agent sessions at once rather than writing code line by line. The account, drawn from a keynote to engineering leaders and interviews with practitioners, also identifies concerns about code quality, reliability and reviews as companies adapt to the new workflow.
The report’s author said the snapshot drew on visits to OpenAI and Anthropic, conversations with startups and technology companies, and unpublished data from GitHub, Factory AI and Linear. It focused on AI labs, venture-backed startups and large technology companies. The supplied report does not publish detailed figures from that data, so the scale of adoption cannot be independently quantified from the account.
Several engineers described dividing work among concurrent AI sessions. Boris Cherny, identified as the creator of Claude Code, said he uses five terminal tabs with separate repository checkouts and runs another five to 10 Claude sessions on the web. Peter Mattis of Cockroach Labs described a similar working limit of about five to 10 agent sessions, sometimes with multiple subagents. Linear engineer Dima Zaytsev said he rotates among local worktrees while agents work on separate tasks.
The report says expectations about code output have changed, and that code reviews can become “theatrical” when reviewers do not meaningfully scrutinize AI-generated work. It also identifies lower quality and reliability as current concerns. These are the author’s observations and interpretations, not industry-wide measurements in the material provided. The report’s counterpoint is that teams and planning still matter, and that non-engineers have not broadly taken over software delivery.
AI Changes the Developer’s Work
If engineers increasingly direct several agents rather than produce each line of code themselves, the job shifts toward task definition, coordination and verification. That could change how teams plan projects, measure individual output and train early-career developers. The report presents these as emerging implications, not settled outcomes.
The quality concerns matter because software must work reliably after it is generated. If review processes do not catch defects, faster code production may create additional maintenance and operational work. The report does not provide defect rates or a direct comparison with pre-AI development, so it cannot establish the size of that risk.
For businesses and workers, the practical question is not only whether AI can generate code, but whether teams can integrate it safely and productively. The experiences described suggest a significant change in working practice, while leaving open whether the approach will deliver consistent gains across different organizations and projects.
As an affiliate, we earn on qualifying purchases.
From Coding Tools to Agent Workflows
The report places the current shift alongside earlier changes such as the spread of the internet, mobile computing, cloud services and new programming languages. Its author argues that AI’s impact is arriving at a greater scale and speed. That comparison is an assessment, not a measured ranking of technology shifts.
It also points to coding-model improvements near the end of 2025 as a catalyst for wider use. In the report’s account, practices that began among AI-lab engineers are spreading into other parts of the industry. Multiple agents, separate worktrees and cloud-based coding sessions are examples of the workflow taking shape.
Martin Fowler, a software engineering author and practitioner, described AI as a change larger than earlier shifts he had experienced, including the internet and agile development. The report uses his remarks to convey the perceived pace of change; they do not establish how quickly adoption is occurring across the entire technology workforce.
“Nothing has hit with the magnitude of AI. This is a whole size difference from anything that we’ve faced before.”
— Martin Fowler, speaking at The Pragmatic Summit
As an affiliate, we earn on qualifying purchases.
Adoption and Quality Data Remain Limited
The report offers accounts from particular engineers and companies, but the material provided does not give representative adoption rates, productivity comparisons or measured defect data. It is not clear how common multi-agent workflows are across the industry, or whether their reported use translates into faster delivery overall.
The claims about declining quality and theatrical reviews are not accompanied here by benchmarks or a defined measurement period. The unpublished data mentioned by the author is not detailed in the supplied material. It also remains unclear how different results may be for less experienced developers, regulated industries or software requiring unusually high reliability.
As an affiliate, we earn on qualifying purchases.
Teams Test New Agent Infrastructure
The report expects cloud coding agents and supporting infrastructure to develop further, alongside changes to the tools and practices used to build software. It also suggests engineers may spend less time reading code directly, although the material does not specify how that would work or what safeguards would replace close review.
The next useful evidence will be measured results from organizations adopting these systems: how much work agents complete, what review and testing catch, and whether maintenance costs rise or fall. Until those details are available, the examples show a changing workflow among some developers, not a confirmed outcome for the industry as a whole.
As an affiliate, we earn on qualifying purchases.
Key Questions
What is the main development described in the report?
The report says AI coding agents are changing software development practices, with some engineers managing several agent sessions in parallel.
Are all engineers now using coding agents?
No industry-wide adoption figure is provided. The report draws on interviews and observations, so it does not establish how many engineers or companies use these workflows.
What risks does the report identify?
It raises concerns about code quality, reliability and superficial reviews. The supplied material does not include measured defect rates or comparisons with earlier practices.
Does the report say engineers are no longer needed?
No. It describes changes to how engineers work and says teams and planning remain important. It does not claim that AI has replaced software engineering teams.
Source: rss
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
