Ask HN: Disillusioned after AI?
Advances in generative AI and large language models are leaving many developers feeling disillusioned, worried that their work will be commoditized, junior roles automated away, and any successful idea quickly absorbed by big tech. Others counter that AI is more of a powerful assistant than a replacement, opening up new possibilities for internal tools, creative workflows, and faster learning, with human taste, originality, and problem selection remaining key differentiators. Underneath the optimism and anxiety runs a broader concern about concentrated power, the erosion of human craft, and whether current AI hype represents genuine progress or just another overblown tech cycle.
Emotional responses to AI
- Many posters describe disillusionment, sadness, or cultural fatigue: AI hype feels fake, demos cringey, and big tech’s dominance demoralizing.
- Others see this as a “rough patch” or age-related perspective shift; some suggest soul‑searching or simply taking a break.
- A sizeable minority are thrilled, describing the current moment as the most exciting time in decades of software work.
Impact on developers, jobs, and “building”
- Some argue AI is like low‑code/Wix: it shifts work rather than destroying it. Routine website work is already gone; remaining frontend roles are more complex and often full‑stack.
- Fears: junior roles and “entry-level” learning opportunities may disappear; products become commoditized when anyone can “wish” something into existence; only platform owners profit.
- Counterpoint: taste, design sense, and domain insight remain scarce. AI can generate generic output; differentiated products still require human judgment and refinement.
AI as tool vs threat
- Many use LLMs as “super senior devs,” rubber ducks, or tutors: debugging, learning Rust, exploring design patterns, moving career switchers faster.
- Others emphasize invisible/internal uses: classification, data extraction, newsletters, mis‑classification detection, etc. These are seen as high‑value, non‑flashy applications.
- Some developers “go back to basics” (raycasters, TUIs, algorithms) for joy, treating AI as background noise.
Centralization, data, and democratization
- Strong concern that AI’s compute and data hunger re‑centralizes tech power in a few firms, making smaller players “second tier forever.”
- Worries about training on private user data and about AI‑generated content polluting future training corpora; hope that regulation and data scarcity might rebalance incentives.
Creativity, culture, and art
- One camp says current gen‑AI only remixes training data, lacks true creativity, and can’t originate genuinely new styles or movements.
- Others argue all art builds on predecessors and that AI‑assisted blends plus human curation can yield new styles; whether they become “movements” will only be clear in hindsight.
- Several note growing preference for clearly human, imperfect work amid a flood of slick, automated content.
AGI, risk, and timelines
- Some dismiss AGI fear as hype, stressing that intelligence is domain‑specific and current systems are brittle.
- Others predict rapid progress with multimodal, tool‑using models, seeing “do what humans do, better/cheaper” as plausible within years.
- Opinions diverge sharply on whether this is an apocalyptic bubble, a durable revolution, or just another overhyped wave.