On a recent afternoon inside a quiet part of Silicon Valley, a group of engineers apparently observed something that seemed both remarkable and slightly unnerving. A coding method that resembled a chat window rather than a conventional development environment started producing whole software modules in a matter of minutes. Code lines appeared quickly. compiled functions. Tests were conducted automatically.
In the time it took to sip a cup of coffee, what had previously required weeks of meticulous engineering labor came to pass. The response was a mixture of fascination and discomfort for many of the engineers there. Claude Code, an artificial intelligence system created by Anthropic, is the culprit, and its developer, Boris Cherny, issued a warning that has been reverberating across the tech sector ever since.
| Category | Information |
|---|---|
| Technology | Claude Code |
| Creator | Boris Cherny |
| Company | Anthropic |
| Release | 2025 |
| Type | Agentic AI coding system |
| Key Capability | Autonomous coding and task execution |
| Notable Claim | “The title software engineer will start to go away.” |
| Adoption | Widely used by engineers in Silicon Valley |
| Related Product | Cowork AI assistant |
| Reference Website | https://www.anthropic.com |
In a recent podcast interview, Cherny stated, “I think by the end of the year, everyone is going to be a product manager, and everyone codes.” The term “software engineer” will begin to disappear. The prediction reverberated around the programming community like a soft thunderclap.
For many years, one of the most stable professions in the contemporary economy was software engineering. Attracted by the prospect of six-figure wages, job security, and a consistent need for technical skill, university students rushed to computer science programs. That presumption seems to be changing now.
Claude Code is more than just a coding aid that completes syntax lines and makes minor suggestions. What engineers refer to as “agentic behavior” is what makes it unique. The system may do activities on its own, such as developing code, troubleshooting issues, and reorganizing entire software structures, rather than just reacting to commands. According to a senior developer, the technology could replicate a year’s worth of labor in just an hour.
It usually leaves a lasting impression to witness anything like that. Cherny acknowledges that the shift has changed the way he operates. According to reports, he has been letting Claude Code write all of his programming for months, only intervening to check and validate the output. “I have not edited a single line by hand since November,” he added, but he emphasized that human monitoring remains crucial for safety and accuracy.
That particular element is more important than it may seem at first. Even the most potent AI systems still need human judgment, particularly when it comes to financial systems managing millions of transactions, infrastructure stability, or security flaws. Even so, the direction of movement appears to be clear.
It’s probable that a programmer’s job is changing from writing code line by line to designing systems, assessing results, and coordinating AI tools. According to Cherny, the job title itself may someday become obsolete. Companies may start searching for builders—people who can concurrently comprehend technology, business goals, design restrictions, and user behavior—instead of “software engineers.”
The printing press, which revolutionized entire professions centuries earlier, has been likened by Cherny himself to this time. Manuscripts had to be copied by hand by scribes prior to printing, which was a labor-intensive process requiring specialized knowledge. Their work underwent a significant transformation after printing was introduced.
Books proliferated. The level of literacy increased. While scribes who adhered completely to copying disappeared into history, those who adapted frequently went on to work in new fields including editing, drawing, and publishing. It’s unclear if the comparison turns out to be correct. Seldom does technology develop exactly as its creators had imagined.
However, many observers in the tech sector see resonance in the analogy. Software engineering has historically entailed repetitive activities, such as developing boilerplate code, debugging minor problems, and reorganizing old programs, much like manuscript copying centuries ago. These are exactly the kinds of jobs that machines excel at. Meanwhile, tools like Claude Code are already making their way into routine tasks.
Cowork, a similar system intended for non-programmers, was just released by Anthropic. With little guidance, the AI can do organizational activities by interacting with messaging apps, spreadsheets, papers, and email platforms. Cherny himself apparently uses the application to send automatic Slack reminders when colleagues forget to update shared spreadsheets. It’s the kind of low-key automation that, once dependable, spreads swiftly. And software engineering might not be the end of it.
According to Cherny, AI systems will someday be used for almost any task that is largely done on a computer. AI agents may increasingly be used as collaborators—or occasionally as replacements—in marketing analysis, financial modeling, product planning, and even project coordination. The forecast poses a query that permeates discussions in Silicon Valley today.
What is specifically human if digital tasks may be completed by machines?
Cherny’s response tends to emphasize adaptability and curiosity. He predicts that generalists—people who are at ease bridging disciplinary boundaries—will be the most successful in the next stage of the technology economy. Instead than concentrating only on developing code, he believes that the best engineers will comprehend design, infrastructure, and business strategy.
To put it another way, technical proficiency might not be sufficient anymore. As this change takes place, more people are realizing that the coming years may be difficult for certain segments of the workforce. Anthropic has reportedly employed economics and policy experts to research the wider effects of their technology because it recognizes the potential for disruption. However, the business has also clarified something. The goal is not to impede progress.
The intense competition among IT firms vying to define the next era of computing is driving the rapid advancement of AI systems. The fact that the changes are already occurring may ultimately be more important than whether society is prepared for them.
