March 8, 2026
Human vibes vs robot slides
AI doesn't replace white collar work
AI vs Office Life: “Bots did my busywork” vs “Humans run the room”
TLDR: The author argues AI excels at quick tasks but can’t replace the trust and judgment behind white‑collar decisions, like strategy. Commenters split: some brag AI already does most of their work and trims headcount, while others predict stricter rules and higher standards will keep humans firmly in the loop.
Author Andrew Marble says AI is great for quick answers but not for the human stuff that actually drives white‑collar work—trust, taste, and talking it out. He points to chatbots nailing code fixes while strategy and advice still need a real person. Cue the comments cage match.
One camp is yelling, AI already ate the paperwork. A user bragged they’ve replaced almost all their non‑meeting tasks with AI, while another claimed it “replaced a couple of white collars” doing translation and content tasks for websites. The efficiency crowd adds, if fewer, smarter people plus fewer status meetings equals more output, why keep the chit‑chat.
On the other side: the “humans matter” squad. One commenter warns we’re just one big AI disaster away from a regulation tsunami—more testing, certifications, security checks—so jobs don’t vanish, the bar just gets higher. Another calls out design: UI stays human only if you bring real taste, not just button‑pushing in Figma. Marble’s core idea—LLMs (large language models) are fine for facts, not feelings—gets both applause and eye‑rolls.
The jokes wrote themselves: “AI does the work, I take the meeting,” “PowerPoint can’t hold your hand,” and “Chatbots can make slides but can’t read the room.” Read Marble’s piece here and decide: is office life about deliverables… or vibes?
Key Points
- •The article separates questions into transactional (fact-finding) and relationship-based (advisory/social) types.
- •It argues LLMs and reference sources are well-suited for transactional questions but not for relationship-based advisory work.
- •Strategy consulting is cited as a domain where trust, understanding, and relationships drive value beyond correct answers.
- •A specific technical example shows AI solving a pandas/Python issue by using pd.Series with dtype='object' to avoid None becoming NaN.
- •The article contends many business tasks rely on judgment and human factors, especially in procedural organizations lacking market feedback.