AI will change the world but not in the way you think

AI and Software Development

• Some see only incremental change for developers (better autocomplete, docs), likening AI to earlier outsourcing fears that never fully materialized beyond low-skill work.
• Others report dramatic productivity gains: faster prototyping, unblocking “someday” projects, lower activation energy, especially for people struggling with motivation or mental health.
• General consensus: AI augments good engineers rather than replaces them, but may raise expectations (“you have AI now, why aren’t you 10x?”).

Bullet Points, Fluff, and Business Communication

• Many agree that verbose, platitude-filled emails are already annoying; AI will make this kind of “lossy expansion” cheap and ubiquitous.
• A popular vision: future workflows where senders write terse bullet points, AI inflates them into polite prose, and recipients use AI to summarize back to bullet points—a “ridiculous communication protocol.”
• Some welcome a shift to terse bullet-point communication; others argue “fluff” carries tone, empathy, social signaling, and narrative, which can’t always be reduced without loss.

Speed, Accuracy, and the “Autocomplete Moment”

• One view: LLMs haven’t had their “Google autocomplete moment” yet—speed and integration into typing are the missing pieces.
• Others say speed is fine; the problem is hallucinations and forgetfulness that would be intolerable in a human coworker.
• Disagreement over whether “mistakes like humans” is an acceptable framing, since professional work is organized around minimizing errors.

Boilerplate, Refactoring, and Code Quality

• LLMs excel at generating boilerplate; some celebrate this as a big win.
• Critics fear juniors will lose the architectural intuition that “needing lots of boilerplate” is a design smell and refactoring signal.
• Counterpoint: if LLMs can cope with messy code, refactoring might matter less for machines (though others insist humans will still eventually need to read and maintain it).

Human Context, Education, and Culture

• Several commenters push back on the idea that people “naturally think in bullet points” or that reading long books/essays is of dubious value; they see deep reading and long-form writing as core cognitive skills under threat.
• Cultural differences in communication style (e.g., American vs German directness) shape how much “fluff” is expected or resented.

Commercial and Workplace Impacts

• Some see AI’s main current commercial use as “enshittification” and feature-bloat, but also predict simple bespoke apps generated by prompts could undercut bloated tools.
• Concerns raised about AI in hiring (LLM-written feedback on take-homes) and about people auto-denylisting obviously AI-generated messages because they erase individual voice and subtext.