TL;DR
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
A report based on interviews, conference observations and data access describes AI coding tools reshaping how engineers work as 2026 begins. Engineers are increasingly coordinating multiple agents, but the report also identifies concerns about code quality, reliability and reviews; the scale of those effects across the industry remains uncertain.
AI coding agents are changing how software engineers work, with some developers now directing several agent sessions at once rather than writing code line by line, according to a 2026 industry report from The Pragmatic Engineer. The report describes a rapid shift in software development practices, while also warning that code quality and reliability have come under pressure.
The report draws on a keynote delivered at the LDX3 engineering leadership conference in New York, attended by more than 2,000 engineering leaders and senior technical staff, as well as visits to OpenAI and Anthropic, conversations with startups and unpublished data from GitHub, Factory AI and Linear. Its author says the aim was to capture changes in AI labs, venture-backed startups and large technology companies, not to provide a census of the entire industry.
One prominent change is the move toward parallel work with multiple coding agents. Claude Code creator Boris Cherny described using five terminal tabs and running five to 10 Claude sessions on the web alongside local sessions. Cockroach Labs co-founder Peter Mattis said his cognitive capacity for concurrent agent sessions was often around five to 10, while software engineer Dima Zaytsev told the report he rotates among several local worktrees as agents work. These are individual accounts, not a measured industry-wide average.
The report also lists practices it says are changing: less code written by hand and a diminished role for the traditional integrated development environment. At the same time, it identifies problems including weaker assumptions about generated code, code reviews that can become performative, and declining quality and reliability. The source does not provide quantified rates for those problems. It argues that some fundamentals remain, including the importance of teams and planning, and says non-engineers are not broadly shipping code themselves.
How Agentic Coding Changes Engineering
The shift matters because coding agents affect more than the speed of producing code. When engineers supervise multiple tasks at once, their work can move from typing and editing toward delegating, checking and integrating outputs. That can change team workflows, development tools and the skills companies need from engineers.
The report’s cautions point to a trade-off: more generated code does not automatically mean better or more dependable software. If reviews become less substantive or teams rely on output they have not adequately checked, quality and reliability risks may grow. The report presents these as observed concerns, but does not establish their prevalence or quantify their impact.
Products Worth Considering
Jackery HomePower 3600 Plus Portable Power Station, 3584Wh, 3600W
Bose SoundLink Flex Bluetooth Speaker (2nd Gen), Portable Waterproof Speaker, Black
Anker SOLIX C200 DC Portable Power Station, 192Wh, 200W
As an affiliate, we earn on qualifying purchases.
From Earlier Tech Shifts to AI
Technology workers have adapted to major changes before, including the spread of the internet, smartphones, cloud computing and new programming languages and frameworks. The report argues that AI differs in the scale and speed of its effects, particularly since coding models improved late in 2025. That comparison is the author’s assessment, not a statistical measurement of technological change.
Martin Fowler, a software industry veteran, said AI’s impact was larger than previous shifts he had experienced. The report uses his view to frame the current period as a broad transformation, while noting that development practices will change unevenly and some established elements, such as teamwork and planning, remain relevant.
“Nothing has hit with the magnitude of AI. This is a whole size difference from anything that we’ve faced before.”
— Martin Fowler, software industry veteran, speaking at The Pragmatic Summit
How Widespread Are These Practices?
The report offers examples from prominent engineers and companies, but does not establish how common multi-agent workflows are across the wider workforce. The source material also does not give the methods, sample sizes or numerical results for the unpublished GitHub, Factory AI and Linear data it mentions.
The extent of any decline in software quality or reliability is also unclear: the report identifies these as problems but supplies no rates or comparison period in the material provided. It remains uncertain how quickly coding agents will spread beyond early adopters, how companies will adjust review practices, and whether reported productivity gains translate into better outcomes for customers.
The Next Phase of Agent-Based Development
The report expects cloud-based coding agents and agent harnesses—the systems used to coordinate and manage agents—to become more prominent. It also anticipates new infrastructure built for AI-assisted development and a possible change in how engineers interact with code. These are forecasts from the report, not confirmed outcomes.
The next useful evidence will include broader data on adoption, software quality and reliability, alongside accounts from teams using these tools in routine production work. Until then, the examples show how some developers are changing their workflows, but not how the whole industry will settle into the new model.
Key Questions
What is changing in software engineering in 2026?
The report says some engineers are writing less code directly and using multiple AI coding agents in parallel. It also describes changes to development tools and workflows, though it does not quantify adoption across the industry.
Are engineers no longer writing code by hand?
The report describes a trend toward less hand-written code, but it does not show that engineers have stopped writing code altogether. Its examples come from individual practitioners and do not represent every team.
What concerns does the report raise about AI-generated code?
It points to concerns about code quality, reliability and the substance of code reviews. The source material does not include figures measuring how often these issues occur or how large their effects are.
Does the report show that AI has improved productivity across tech?
No industry-wide productivity result is provided in the material. The report gives examples of engineers handling concurrent agent sessions, but those accounts do not establish a sector-wide productivity gain.
Source: rss
