The grief, loneliness, and burnout sweeping through the tech industry right now | Molly Graham

What It Covers

Molly Graham, creator of the 13-year-old "give away your Legos" career framework, revisits her advice in the context of AI-driven workplace transformation. She examines what still holds true about embracing change, what has fundamentally shifted when delegating to AI versus humans, and how grief, loneliness, and burnout are reshaping how tech workers experience their careers in 2025.

Key Questions Answered

• •Burnout acceleration: Burnout in tech jumped from 44% to 55% in a single year according to Lenny's annual survey, a 10-point spike driven by three compounding pressures: the emotional exhaustion of constant narrative shifts, the expectation to produce more for the same pay, and the cognitive overhead of overseeing AI outputs. Leaders need to name this explicitly rather than defaulting to enthusiasm-only messaging about AI's potential.
• •AI delegation is not the same as human delegation: The original Legos framework assumed giving something away meant fully releasing mental ownership. Delegating to AI does not work this way — humans retain oversight, quality accountability, and final responsibility. This means the psychological burden stays with the person, even as AI handles execution. The mental tax of managing multiple AI agents is additive, not subtractive, which partially explains rising burnout rates.
• •Treat AI as a junior intern, not a superintelligence: The dominant narrative frames AI as smarter than any human employee, which causes people to skip the coaching, context-setting, and iteration that produce quality output. Treating AI like a junior intern — requiring onboarding, correction, and multiple drafts before anything ships — produces better results and prevents the AI-generated slop that is now a measurable drag on organizational efficiency.
• •Reframe job fear using the Manoush Zomorodi model: Journalist Manoush Zomorodi built a 30-year career through repeated industry disruption by asking not "will my job disappear?" but "what will this job look like in six years?" Applied to tech: engineering, design, and product management will not disappear, they will transform. Lenny's survey data confirms engineering demand is currently higher than ever, even as the daily work of writing code has fundamentally changed.
• •Some Legos should not be given away: For 13 years, Graham's advice was to give away everything without exception. She now adds a caveat for AI: work requiring judgment, trust-based relationships, taste, and vision should stay with humans. CEOs copy-pasting AI-generated strategy memos are role-modeling that accountability is optional. The specific categories to retain include anything where the person cannot yet define what "good" looks like, since quality cannot be delegated without a quality standard.
• •Grief requires acknowledgment before adaptation: Engineers report missing the flow state of writing code; designers report loneliness from reduced human collaboration; product leaders describe feeling isolated working primarily with AI tools. Graham recommends leaders explicitly name these losses rather than bypassing them with optimism. Holding space for grief — including what she calls a "funeral" for what work used to be — is a prerequisite for genuine adaptation, not a detour from it.
• •Manager quality is the single most actionable lever for employee happiness: Lenny's survey identifies direct manager quality as the strongest correlate of workplace happiness, more controllable than AI tooling or company size. Smaller teams and companies report significantly higher satisfaction. Simultaneously, many large companies are eliminating management layers for efficiency. Graham argues this is a strategic error — great managers make people feel seen, and that function becomes more critical, not less, during periods of rapid disorienting change.