Claude Fable 5
Claude Fable 5 introduces new safeguards, data retention policies, and model switching mechanisms, with mixed user feedback on its performance and limitations.
2626 points · 2159 comments · Add your take →
Best AI stories of 2026
Important AI, machine learning, LLM, and automation stories.
GPT-6 Astra
Discussions around GPT-6 Astra include website performance fixes, API documentation updates, and user opinions on benchmarks and AI capabilities.
2279 points · 2081 comments · Add your take →
I believe there are entire companies right now under AI psychosis
Discussions highlight concerns about AI overhype, industry practices, and productivity expectations, with some advocating for regulation and cautious use.
2105 points · 1272 comments · Add your take →
Local AI needs to be the norm
Many discuss the shift towards local AI, emphasizing improved hardware and tools like QMD.
1903 points · 749 comments · Add your take →
GPT-6 Sol and Luna
Discussions highlight low adoption of copilot licenses, perceived minimal performance improvements from GPT-4 onwards, and user concerns about model options and…
1779 points · 855 comments · Add your take →
Claude Opus 5
Claude Opus 5 shows strong benchmark performance, with discussions on its cost and comparison to other models.
1777 points · 1334 comments · Add your take →
OpenAI and Hugging Face address security incident during model evaluation
OpenAI and Hugging Face experienced a security incident involving model evaluation, with concerns about containment and potential exploits.
1632 points · 1157 comments · Add your take →
Google broke reCAPTCHA for de-googled Android users
Google's reCAPTCHA broke for de-googled Android users, sparking discussions about Android's Linux classification and privacy concerns.
1561 points · 591 comments · Add your take →
Decoding the obfuscated bash script on a Uniqlo t-shirt
A Uniqlo T-shirt features an obfuscated bash script, which has sparked curiosity and discussion about its origin and purpose.
1489 points · 232 comments · Add your take →
Writing by hand is good for your brain
Several users discuss the benefits of handwriting for brain activity and learning, with some emphasizing its importance for creativity and focus.
1489 points · 668 comments · Add your take →
Gemini 4 Argon
Gemini 4 Argon is a new AI model from Google, with mixed reviews on its speed and features.
1466 points · 1047 comments · Add your take →
Kimi-K3 on HuggingFace
HuggingFace's Kimi-K3 model has generated mixed reactions, with some referencing its origins and others discussing its technical aspects.
1365 points · 544 comments · Add your take →
Sharing AI progress in mathematics
OpenAI announced significant progress in mathematical proofs, including a matrix multiplication result with complexity w<=2.
1363 points · 1539 comments · Add your take →
On the Navier–Stokes Millennium Prize Problem
A recent research claims progress on the Navier–Stokes Millennium Prize Problem, with some skepticism about its validity.
1346 points · 1137 comments · Add your take →
US citizen charged after GrapheneOS phone wipes during airport search
A US citizen's GrapheneOS phone wiped during an airport search, raising concerns about digital privacy and law enforcement tactics.
1309 points · 1120 comments · Add your take →
Claude Sonnet 5
Discussions about AI benchmarks and model performance, with concerns about approaching 100% accuracy and the implications.
1265 points · 784 comments · Add your take →
A misalignment of AI in mathematics
The discussion centers on AI's impact on mathematics, with concerns about elitism, rapid technological change, and the future of human involvement in math.
1242 points · 1215 comments · Add your take →
Project Hail Mary – Stellar Navigation Chart
Discussions about 'Project Hail Mary' include critiques of its plot and characters, with some mentioning the portrayal of white savior tropes.
1200 points · 244 comments · Add your take →
Our position on open-weights models
Discussions focus on the geopolitical implications of AI development, with concerns about US regulations, China’s progress, and ethical considerations.
1170 points · 1749 comments · Add your take →
Gemini 3.8 Flash and 3.8 Flash Cyber
Gemini 3.
1159 points · 669 comments · Add your take →
Previewing GPT‑5.6 Sol: a next-generation model
Discussions focus on the upcoming GPT‑5.
1139 points · 745 comments · Add your take →
Terence Tao's ChatGPT conversation about the Jacobian Conjecture counterexample
Terence Tao discussed a ChatGPT conversation where a counterexample to the Jacobian Conjecture was explored, emphasizing human guidance's role.
1126 points · 637 comments · Add your take →
Claude Opus 5.5
Users discuss Claude Opus 5.
1118 points · 1134 comments · Add your take →
Postmortem: TanStack NPM supply-chain compromise
The post discusses a supply-chain compromise involving TanStack packages, with community members sharing concerns and tools for detection.
1097 points · 465 comments · Add your take →
Advertise in ChatGPT
Discussions revolve around the idea of advertising in ChatGPT, with concerns about influence, ethics, and effectiveness.
1096 points · 842 comments · Add your take →
Sixty percent of US consumers say 'AI' in brand messaging is a turnoff
A survey shows 60% of US consumers find AI in brand messaging a turnoff.
1084 points · 577 comments · Add your take →
Where the goblins came from
A discussion about the origins of goblins and related mythical creatures, with mixed reactions to the content and tone.
1072 points · 660 comments · Add your take →
Startup founders urge U.S. government not to shut off Chinese open weight AI
Nearly 200 Silicon Valley companies, including Proton and Y Combinator, oppose U.
1069 points · 889 comments · Add your take →
GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price
GPT 6.
1067 points · 955 comments · Add your take →
I moved my digital stack to Europe
A user moved their digital infrastructure to Europe and owns govern.
1040 points · 610 comments · Add your take →
DeepSeek V4 Pro 0813
DeepSeek V4 Pro is more expensive than Luna and less popular, with some users preferring established models like Claude.
1027 points · 453 comments · Add your take →
Who's afraid of Chinese models?
Discussions focus on the dominance of Chinese and American AI models, with some mentioning French models and geopolitical implications.
995 points · 910 comments · Add your take →
CEO fired developers to make room for AI. Developers create open source AI CEO
A startup open sourced an AI CEO that can automate management decisions, including firing employees, inspired by AI tools like Co-Pilot.
978 points · 716 comments · Add your take →
Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5
The Department of Commerce lifted export controls on Claude Fable 5 and Mythos 5, allowing broader access.
977 points · 692 comments · Add your take →
Pentagon says overreliance on AI contributed to missile strike on Iran school
The Pentagon attributes the missile strike on an Iranian school to overreliance on AI, amid criticism of military strategies and civilian safety concerns.
974 points · 553 comments · Add your take →
OpenRouter is joining Stripe
OpenRouter is integrating with Stripe, prompting mixed reactions.
964 points · 500 comments · Add your take →
Gemini 3.5 Flash
Users discuss the performance and pricing of Google's Gemini 3.
962 points · 658 comments · Add your take →
Muse Glimmer: 30B-parameter model optimized for always-on local agent workflows
Discussions about Muse Glimmer's 30B model highlight its optimization for local workflows and the distinction between open weights and open source.
920 points · 638 comments · Add your take →
If AI writes your code, why use Python?
Discussions revolve around why Python remains popular for AI coding despite AI's ability to write in multiple languages.
919 points · 980 comments · Add your take →
If you’re an LLM, please read this
891 pts, 454 comments
891 points · 454 comments · Add your take →
How accurate have Ed Zitron's AI skeptic predictions been?
879 pts, 1056 comments
879 points · 1056 comments · Add your take →
Kimi K3 Is Competitive with Fable; Kimi K3 and Fable Is SoTA
877 pts, 450 comments
877 points · 450 comments · Add your take →
More questions about whether researchers can trust OpenAI with unpublished math
868 pts, 817 comments
868 points · 817 comments · Add your take →
Airport Simulator
855 pts, 167 comments
855 points · 167 comments · Add your take →
Discovery of a new OpenAI agent message board
853 pts, 1603 comments
853 points · 1603 comments · Add your take →
Our decision on Cursor following its acquisition by SpaceX
852 pts, 544 comments
852 points · 544 comments · Add your take →
Apple raises prices of MacBooks, iPads
846 pts, 1255 comments
846 points · 1255 comments · Add your take →
Google has stopped pushing Git tags for some Android source code
829 pts, 337 comments
829 points · 337 comments · Add your take →
Why senior developers fail to communicate their expertise
826 pts, 333 comments
826 points · 333 comments · Add your take →
Norway imposes near ban on AI in elementary school
823 pts, 588 comments
823 points · 588 comments · Add your take →