Top all-time · #3934
Jack Clark
Personal blog · jack-clark.netaipoliticswriting
Jack Clark writes Import AI, a personal blog at jack-clark.net about ai, politics and writing. Since 2016, 20 different Hacker News readers have linked it 28 times, most often “Technological Optimism and Appropriate Fear”. It is still writing: the last post was on 2026-10-26, and you can follow it by feed.
Too few sampled comments (18) to say how often HN recommends it.?
Why HN links it
Jack Clark's site, home to Import AI, a weekly AI newsletter cross-posted to the blog that covers research papers and OpenAI updates, plus short AI sci-fi stories.
Summarised by Claude from a sample of 18 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Most-linked posts?
- /6 people · 7 mentions · first linked 2017-03-12
- Technological Optimism and Appropriate Fear4 people · 4 mentions · first linked 2025-10-13
- What does 10^25 versus 10^26 mean? – AI regulations2 people · 2 mentions · first linked 2024-06-19
- /import-ai2 people · 2 mentions · first linked 2017-04-11
What Hacker News says?
I write a weekly AI newsletter called Import AI which is also cross-published to this WP blog. I try to cover a mixture of fundamental research papers and applied stuff. It also includes some OpenAI updates: https://jack-clark.net/
Comment linking /
on “Ask HN: What are some good technology blogs to follow?”Submitted AI development is going to get even faster as a story · 39 points
HN discussionIt was mentioned in Anthropic Jack Clark's "Import AI" newsletter. https://jack-clark.net/2025/02/17/import-ai-400-distillation...
Comment linking /2025/02/17/import-ai-400-distillation-scaling-laws-recursive-gpu-kernel-improvement-and-wafer-scale-computation/
on “Grok3 Launch [video]”Submitted Automating AI Research as a story · 4 points
HN discussionYour sci-fi idea reminds me of Jack's short stories: https://jack-clark.net/ (scroll to the bottom of the post for it)
Comment linking /
on “You can no longer reinstall CS 2 or 3 even if you have the original disks”Submitted Technological Optimism and Appropriate Fear as a story · 3 points
HN discussionJust the links from comments: http://hndigest.com/ http://ambassadors.thehustle.co/ http://androidweekly.net/ http://ben-evans.com/newsletter http://delanceyplace.com/ http://eepurl.com/cE6e9H http://fair.org/ http://fermatslibrary.com/ http://flowingdata.com/ http://fortune.com/tag/term-sheet/ http://freshpatents.com http://frontendfocus.co/ http://growthhackingidea.com/ http://hackaday.com http://hackernewsbooks.c…
Comment linking /import-ai
on “Ask HN: Mailing lists that HN readers ought to know about?”Submitted Technological Optimism and Appropriate Fear as a story · 2 points
HN discussion(Disclosure: I write Import AI). Link: http://jack-clark.net/import-ai
Comment linking /import-ai
on “Ask HN: Mailing lists that HN readers ought to know about?”Submitted What does 10^25 versus 10^26 mean? – AI regulations as a story · 2 points
HN discussionImport AI is good https://jack-clark.net/
Comment linking /
on “Ask HN: What hyper-focused industry newsletters do you enjoy?”Submitted AI Safety and Corporate Power – Remarks Given – United Nations Security Council as a story · 1 points
HN discussionApparently GPT-3 has learned a large number of personality types and their fine grained biases, well enough to simulate a poll with good accuracy. This means you can poll it any time to check how a population of choice would have reacted had something happened. The language model does not carry the just biases of its builders, it learns all biases equally, you just need to specify the desired bias (personality type)…
Comment linking /2022/10/11/import-ai-305-gpt3-can-simulate-real-people-ai-discovers-better-matrix-multiplication-microsoft-worries-about-next-gen-deepfakes/
on “A Danish political party led by an AI”Submitted My agents are working. Are yours? as a story · 1 points
HN discussion> How? This is absolutely impossible. It is possible to use GPT-3 to impersonate someone. You give it a profile and it assumes the personality, then speaks from that point of view replicating its biases with uncanny ability. So what they did was to create a simulated poll using GPT-3 by aligning the model to personality profiles distributed identically to the real population. The result? GPT-3 can predict how an ide…
Comment linking /2022/10/11/import-ai-305-gpt3-can-simulate-real-people-ai-discovers-better-matrix-multiplication-microsoft-worries-about-next-gen-deepfakes/
on “A Danish political party led by an AI”Submitted Technological Optimism and Appropriate Fear as a story · 1 points
HN discussionI know a good blog post about it. https://jack-clark.net/2022/10/11/import-ai-305-gpt3-can-sim...
Comment linking /2022/10/11/import-ai-305-gpt3-can-simulate-real-people-ai-discovers-better-matrix-multiplication-microsoft-worries-about-next-gen-deepfakes/
on “GPT-3 can create both sides of an Interactive Fiction transcript”Submitted Technological Optimism and Appropriate Fear as a story · 1 points
HN discussionIf you enjoyed this style also check out [1], a weekly short story based on contemporary AI. Scroll down to "Tech Tales" for the stories. [1] https://jack-clark.net/
Comment linking /
on “Short Story on AI: Forward Pass”Submitted Import AI 228, US AI Regulations and More as a story · 1 points
HN discussionImport AI: https://jack-clark.net/; TechCrunch; MacRumors
Comment linking /;
on “Ask HN: Which tech blogs or publications do you follow?”SciFi about AIs deciding elections? Here's a real paper on using GPT-3 to simulate a human poll. > Specifically, they do three social science experiments on GPT-3 and discover that GPT-3 has biases that are “fine-grained and demographically correlated, meaning that proper conditioning will cause it to accurately emulate response distributions from a wide variety of human subgroups.” https://jack-clark.net/2022/10/11…
Comment linking /2022/10/11/import-ai-305-gpt3-can-simulate-real-people-ai-discovers-better-matrix-multiplication-microsoft-worries-about-next-gen-deepfakes/
on “People tricking ChatGPT “like watching an Asimov novel come to life””Import AI: https://jack-clark.net/ Less Wrong: https://www.lesswrong.com/ The Azimuth Project: http://www.azimuthproject.org/
Comment linking /
on “What are the good-to-read academic blogs on any topic?”Some suggestions: https://aimagazine.com/ https://jack-clark.net/ https://www.ben-evans.com/newsletter https://scholar.google.com/ search 'AI'
Comment linking /
on “Ask HN: Where can I find an updated AI reading list?”https://jack-clark.net/
Comment linking /
on “DeepSeek v4”GPT-3 can be used for "Silicon Sampling" - running a virtual poll based on prompting with varied personality profiles. https://jack-clark.net/2022/10/11/import-ai-305-gpt3-can-sim...
Comment linking /2022/10/11/import-ai-305-gpt3-can-simulate-real-people-ai-discovers-better-matrix-multiplication-microsoft-worries-about-next-gen-deepfakes/
on “New GPT-3 model: text-DaVinci-003”This article has some information. https://jack-clark.net/2024/03/28/what-does-1025-versus-1026...
Comment linking /2024/03/28/what-does-1025-versus-1026-mean/
on “Ask HN: What's with floating point operations and AI regulations?”For example, the guy who said Google's LaMDA model was conscious was subtly influencing the model with loaded questions. The model picked up on the trend and played the role. Language models can infer your expectations from just a few words and then latch onto them. It's what they are trained to do, literally - take a prompt and write a continuation. They are not trained to oppose the prompt. Now, this can be very d…
Comment linking /2022/10/11/import-ai-305-gpt3-can-simulate-real-people-ai-discovers-better-matrix-multiplication-microsoft-worries-about-next-gen-deepfakes/
on “ChatGPT Political Compass”
