AI as Normal Technology
normaltech.ai · Group blog · on HN since 2025
Princeton researchers Arvind Narayanan and Sayash Kapoor write about AI as a normal technology, covering agent reliability, jobs and policy.
What HN saysHN commenters cite it for the view that AI layoffs are partly AI-washing and that reliable AI agents are harder to build than claimed.
15HN people shared it
6linked it in comments
14stories submitted
16story points
3 months agolast shared
2025people per year2026
Most-shared posts
- 1Did Google's AI agents build an operating system for $916?3 people · 5 points · first shared 2026 · discussion on HN
- 2Towards a science of AI agent reliability3 people · 2 points · first shared 2026 · discussion on HN
- 3Fact Checking Moravec's Paradox2 people · 2 points · first shared 2026 · discussion on HN
- 4What will be left for us to work on?2 people · 1 points · first shared 2026 · discussion on HN
- 5Ai as normal technology2 people · first shared 2025
- 6Up the Stack: How AI's Escape from the Commodity Trap Risks Enterprise Lock-In1 people · 1 points · first shared 2026 · discussion on HN
- 7Do AI Risks Require Extraordinary Government Intervention?1 people · 1 points · first shared 2026 · discussion on HN
- 8AI agents can't yet do open-ended AI research1 people · 1 points · first shared 2026 · discussion on HN
- 9Why AI hasn't replaced software engineers, and won't1 people · 1 points · first shared 2026 · discussion on HN
- 10A guide to understanding AI as normal technology1 people · 1 points · first shared 2025 · discussion on HN
- 11Does the UK's liver transplant matching algorithm exclude younger patients?1 people · 1 points · first shared 2026 · discussion on HN
- 12Faq about the book and our writing1 people · first shared 2025
- 13Archive1 people · first shared 2026
- 14Starting reading the ai snake oil1 people · first shared 2025
- 15Homepage1 people · first shared 2025
What HN people wrote when they linked it
internet_points · 2026-06-12
https://www.normaltech.ai/i/201537309/the-stories-of-ai-driven-mass-la… suggests AI is used as an excuse rather than being a real reason.
tuatoru · 2025-09-19
This all sounds like it has been covered in detail by the "AI as a Normal Technology"[1][2] guys (formerly AI Snake Oil - they decided they preferred to engage rather than just be snarky).
Invention vs innovation vs diffusion - this is all well-known stuff.
It's a completely different episteme than the one IABIED guys have ("If Anyone Builds It, Everyone Dies").
I don't think there can be any meaningful dialogue between the two camps.
1. Substack: https://www.normaltech.ai/ book: https://www.normaltech.ai/p/starting-reading-the-ai-snake-oil
2. "Normal technology" like fire, metals, agriculture, writing, and electricity are normal technologies.
aanet · 2026-07-19
From his keynote at ICML 2026
More here https://www.normaltech.ai/p/what-will-be-left-for-us-to-work
vivzkestrel · 2026-07-14
- i really dont get why some of the guys go out of their way to hide the archive page https://www.normaltech.ai/archive
- i would assume it is reasonable that anyone comes and see what other posts a person has written except you cant find that page anywhere linked
cootsnuck · 2026-06-09
Yup, precisely. Turns out getting AI to be reliable at doing useful things is harder than we've all been led to believe by the dominant narratives.
https://www.normaltech.ai/p/new-paper-towards-a-science-of-ai
randomwalker · 2025-09-19
Yes, we're aware! Fortunately our book is not a broad indictment of AI :) And none of our claims are premised on tasks people can do remaining out of reach for AI. More here: https://www.normaltech.ai/p/faq-about-the-book-and-our-writing
Our more recent essay (and ongoing book project) "AI as Normal Technology" is about our vision of AI impacts over a longer timescale than "AI Snake Oil" looks at https://www.normaltech.ai/p/ai-as-normal-technology
I would categorize our views as techno-optimist, but people understand that term in many different ways, so you be the judge.