Key Takeaways
- • Menlo Ventures partner Deedy Das identifies an emerging two-tier divide among software engineers: "lazy" engineers who heavily rely on AI coding tools for productivity, and experienced "craftsmen" engineers who absorb the burden of reviewing, understanding, and fixing AI-generated code.
- • A growing identity crisis is affecting the engineering workforce as the traditional role of writing code shifts toward code review and validation management, forcing developers to reassess their professional value proposition in an AI-assisted environment.
- • Code review and maintenance have become the operational bottleneck as AI tools make code generation easier, mirroring the rise of "botsitting"—the practice of supervising AI outputs, fixing errors, and validating machine-generated work.
- • Companies pursuing aggressive "tokenmaxxing" strategies to maximize AI tool usage are inadvertently creating workplace burnout and frustration among senior technical staff, who report bearing disproportionate cognitive and operational loads.
- • The shift represents a structural challenge for software engineering teams: as AI commoditizes code generation, the industry lacks clarity on how to redistribute workload, incentivize quality oversight, or retain experienced engineers experiencing role compression and professional dissatisfaction.
Summary
Deedy Das, a partner at venture capital firm Menlo Ventures specializing in AI and enterprise software investments, has raised concerns about an emerging occupational crisis among software engineers as companies aggressively deploy AI coding tools. According to Das, the industry is experiencing what he characterizes as "an identity crisis bordering on depression" among technical staff—a consequence of how organizations are restructuring engineering workflows around AI-assisted code generation.
Das identifies a bifurcation within engineering teams between two distinct worker categories. The first group, which he terms "lazy" engineers, leverages AI tools extensively to write code, answer technical questions, prepare updates, and complete assigned tasks with minimal independent engagement. The second group—experienced "craftsmen"—bears disproportionate responsibility for understanding, reviewing, and remediating the expanding volume of machine-generated code. This structural imbalance has created what Das describes as unsustainable workload distribution: "The craftsmen are tired. Very tired. The entire burden of review falls on the craftsman. The burden of understanding."
The underlying mechanism driving this crisis is the operational shift in how engineering value is defined and distributed. As AI tools reduce friction in code generation, code review and maintenance have emerged as the new bottleneck in software development pipelines. Engineers accustomed to viewing their professional identity through the lens of creative problem-solving and code authorship now find themselves repositioned as quality assurance and validation managers—a role compression that challenges their sense of professional purpose and craft mastery. Das captured this sentiment in his observation that "the craft they loved is dead."
This dynamic reflects a broader phenomenon Business Insider identifies as "AI sprawl": organizations deploying multiple AI tools simultaneously, creating redundant workflows, and generating increasing volumes of output without demonstrable gains in operational effectiveness. The rise of "botsitting"—the practice of supervising AI systems, correcting errors, and validating outputs—exemplifies how companies are inadvertently converting technical talent into oversight labor rather than leveraging AI as a genuine force multiplier.
The implications extend beyond individual burnout to talent retention and strategic decision-making within technology organizations. The aggressive pursuit of "tokenmaxxing"—maximizing AI tool utilization for productivity metrics—appears to be creating misaligned incentive structures that penalize quality-focused senior engineers while rewarding volume-based junior adoption patterns, potentially destabilizing team cohesion and long-term code quality.
Das identifies a bifurcation within engineering teams between two distinct worker categories. The first group, which he terms "lazy" engineers, leverages AI tools extensively to write code, answer technical questions, prepare updates, and complete assigned tasks with minimal independent engagement. The second group—experienced "craftsmen"—bears disproportionate responsibility for understanding, reviewing, and remediating the expanding volume of machine-generated code. This structural imbalance has created what Das describes as unsustainable workload distribution: "The craftsmen are tired. Very tired. The entire burden of review falls on the craftsman. The burden of understanding."
The underlying mechanism driving this crisis is the operational shift in how engineering value is defined and distributed. As AI tools reduce friction in code generation, code review and maintenance have emerged as the new bottleneck in software development pipelines. Engineers accustomed to viewing their professional identity through the lens of creative problem-solving and code authorship now find themselves repositioned as quality assurance and validation managers—a role compression that challenges their sense of professional purpose and craft mastery. Das captured this sentiment in his observation that "the craft they loved is dead."
This dynamic reflects a broader phenomenon Business Insider identifies as "AI sprawl": organizations deploying multiple AI tools simultaneously, creating redundant workflows, and generating increasing volumes of output without demonstrable gains in operational effectiveness. The rise of "botsitting"—the practice of supervising AI systems, correcting errors, and validating outputs—exemplifies how companies are inadvertently converting technical talent into oversight labor rather than leveraging AI as a genuine force multiplier.
The implications extend beyond individual burnout to talent retention and strategic decision-making within technology organizations. The aggressive pursuit of "tokenmaxxing"—maximizing AI tool utilization for productivity metrics—appears to be creating misaligned incentive structures that penalize quality-focused senior engineers while rewarding volume-based junior adoption patterns, potentially destabilizing team cohesion and long-term code quality.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.