The Continued Trajectory of Idiocy in the Tech Industry

Original Article ↗ Hacker News Discussion ↗

Blockchain / Crypto vs Real-World Impact

  • Some argue blockchain/NFT hype was largely online and didn’t translate into significant real-world adoption.
  • Others counter with concrete examples: major banks ran serious blockchain initiatives, El Salvador’s Bitcoin experiment, Bitcoin ATMs in rural areas, and widespread exposure via financial media.
  • Perceived “real” crypto use cases: buying drugs, ransomware payments, and speculative investment; little evidence in the thread of mainstream non-criminal utility.
  • General sentiment: blockchain was heavily driven by grifters, vendors, and buzzword-chasing executives.

Nature and Impact of AI Hype

  • Many posters see a clear difference between AI/ML and blockchain: AI is viewed as a long-running academic field that’s delivering real results (LLMs, computer vision, protein folding, robotics).
  • Others think generative AI is overhyped, especially around content creation and AGI claims, and lump current “AI” marketing in with past bubbles.
  • Comparison to earlier cycles: web, smartphones, SaaS, cloud seen as hypes that left lasting value; question is where AI will land on that spectrum.

Practical Uses and Limitations of LLMs

  • Reported useful applications:
    • Translation, documentation Q&A, RAG over large wikis, semantic search.
    • Programming help, CLI examples, debugging, general “how do I…?” questions.
    • Accessibility (speech interfaces, help for blind users), radiology support, research, robotics/vision tasks.
  • Some users say AI assistants have significantly changed their workflows and largely replaced basic web search.
  • Others report frequent hallucinations and time-wasting failures (wrong queries, made‑up frameworks/APIs), leading them to revert to traditional search or manual work.
  • Several stress the need to distinguish LLMs from broader ML used in self-driving and robotics.

Ethical and Social Concerns Around AI

  • One camp sees “zero ethical concerns” in training data.
  • Another lists issues: scraped books (Books3), social media and YouTube data without consent, GitHub code regardless of license; calls these ethically problematic.
  • There is also discomfort with forced, opt‑out deployment of AI features (OS-level assistants, “AI Overviews,” AI buttons in products).

Patterns of Tech Hype and Grift

  • Recurrent theme: tech cycles are driven by grifters, VCs, marketing, and credulous management; useful innovation and bullshit coexist.
  • Some posters think critics are simply threatened by new tech; others say skepticism is rational given enshittification and past bubbles.

Other Topics

  • Brief debate over software patents as defense vs patent trolls.
  • VR/AR cited as another hype cycle; views differ on its long-term value.
  • Some mention potential non-crypto uses for blockchains (e.g., direct democracy, data integrity), but acknowledge these are drowned out by grift.