Meta's new Superintelligence lab is discussing training a closed model

https://www.nytimes.com/2025/07/14/technology/meta-superintelligence-lab-ai.html

1•Voloskaya•1m ago•1 comments

Weight of an hourglass—Theory and experiment in quantitative comparison

https://pubs.aip.org/aapt/ajp/article/85/2/98/1058300/Weight-of-an-hourglass-Theory-and-experiment-in

1•voxadam•2m ago•0 comments

A summer of security: empowering cyber defenders with AI

https://blog.google/technology/safety-security/cybersecurity-updates-summer-2025/

1•alexcos•3m ago•0 comments

The CEO of Nvidia Admits What Everybody Is Afraid of About AI

https://gizmodo.com/the-ceo-of-nvidia-admits-what-everybody-is-afraid-of-about-ai-2000628701

2•freedomben•5m ago•0 comments

Nvidia's CEO says it gained US approval to sell H20 AI chips to China

https://www.aljazeera.com/economy/2025/7/15/nvidias-ceo-says-it-gained-us-approval-to-sell-h20-ai-chips-to-china

1•andsoitis•5m ago•0 comments

A Handshake in Orbit 50 Years Ago Transformed the Space Race

https://www.nytimes.com/2025/07/14/science/apollo-soyuz-test-project-1975-anniversary.html

1•benbreen•6m ago•0 comments

Show HN: ProjectD – Google Drive-based, AES-encrypted C2 in C/C++

https://github.com/BernKing/ProjectD

1•bernking•6m ago•0 comments

Featured notebooks on selected topics in NotebookLM

https://blog.google/technology/google-labs/notebooklm-featured-notebooks/

1•janandonly•7m ago•0 comments

Show HN: Write better AI system prompts

https://www.tramann-projects.com/MagicPromptWriter/index.php

1•Marten42•7m ago•0 comments

Fedora's i686 support gets a reprieve

https://lwn.net/Articles/1026917/

1•voxadam•8m ago•0 comments

Cellulose instead of crude oil: Researchers develop sustainable foams

https://techxplore.com/news/2025-07-cellulose-crude-oil-sustainable-foams.html

1•PaulHoule•8m ago•0 comments

Are personas even doing something when prompting?

https://blog.fjrevoredo.com/are-personas-even-doing-something-when-prompting/

1•holyknight•10m ago•0 comments

Packages Proxies

https://blog.garambrogne.net/packages-proxies-en.html

1•todsacerdoti•10m ago•0 comments

Hyperreal Specifications for Continuous Sparse Data Computations [video]

https://www.youtube.com/watch?v=4-4jS574tLc

1•surprisetalk•11m ago•0 comments

Anthropic Economic Futures

https://www.anthropic.com/economic-futures

1•surprisetalk•11m ago•0 comments

Don't Eat Honey

https://benthams.substack.com/p/dont-eat-honey

2•surprisetalk•11m ago•0 comments

A Unifying Framework for Robust and Efficient Inference with Unstructured Data

https://arxiv.org/abs/2505.00282

1•surprisetalk•11m ago•0 comments

Building Grafana dashboards with AI, CLI and a bit of pragmatism

https://quesma.com/blog-detail/building-grafana-dashboards-ai-cli

4•stared•12m ago•0 comments

NASA retiree and crewmates splash down in Pacific after private ISS mission

https://www.theguardian.com/us-news/2025/jul/14/international-space-station-astronauts-home

1•andsoitis•13m ago•0 comments

LLMs fail to demonstrate internal world model, according to Harvard/MIT study

https://arxiv.org/abs/2507.06952

3•upghost•14m ago•1 comments

OpenZFS Bug Ported to Zig

https://andrewkelley.me/post/openzfs-bug-ported-zig.html

1•Bogdanp•14m ago•0 comments

Profiling Crosswords' Rendering Pipeline

https://blogs.gnome.org/anonymoux47/2025/07/14/profiling-crosswords-rendering-pipeline/

1•JNRowe•14m ago•0 comments

Medieval medicine was smarter than you think – and similar to TikTok trends

https://www.binghamton.edu/news/story/5669/medieval-medicine-was-smarter-than-you-think-and-weirdly-similar-to-tiktok-trends

2•geox•16m ago•0 comments

The Machine Stops (1909) [pdf]

https://www.cs.ucdavis.edu/~koehl/Teaching/ECS188/PDF_files/Machine_stops.pdf

1•kunzhi•17m ago•0 comments

A bad ad for a device that photographed CRTs

https://buttondown.com/suchbadtechads/archive/video-printer-1983/

2•rfarley04•19m ago•0 comments

Show HN: Opper AI – Task-Completion API for LLMs

https://opper.ai/

8•felix089•20m ago•1 comments

Book Review: Arguments About Aborigines

https://www.astralcodexten.com/p/book-review-arguments-about-aborigines

1•feross•22m ago•0 comments

O3 and Grok 4 Accidentally Vindicated Neurosymbolic AI

https://garymarcus.substack.com/p/how-o3-and-grok-4-accidentally-vindicated

2•Ratelman•23m ago•1 comments

How I do it

https://daniel.haxx.se/blog/2025/07/13/how-i-do-it/

2•gpi•29m ago•0 comments

Plasma 6.4 review – A worrying trend

https://www.dedoimedo.com/computers/plasma-6-4-review.html

7•coffeeaddict1•29m ago•1 comments

Meta's new Superintelligence lab is discussing training a closed model

Weight of an hourglass—Theory and experiment in quantitative comparison

A summer of security: empowering cyber defenders with AI

The CEO of Nvidia Admits What Everybody Is Afraid of About AI

Nvidia's CEO says it gained US approval to sell H20 AI chips to China

A Handshake in Orbit 50 Years Ago Transformed the Space Race

Show HN: ProjectD – Google Drive-based, AES-encrypted C2 in C/C++

Featured notebooks on selected topics in NotebookLM

Show HN: Write better AI system prompts

Fedora's i686 support gets a reprieve

Cellulose instead of crude oil: Researchers develop sustainable foams

Are personas even doing something when prompting?

Packages Proxies

Hyperreal Specifications for Continuous Sparse Data Computations [video]

Anthropic Economic Futures

Don't Eat Honey

A Unifying Framework for Robust and Efficient Inference with Unstructured Data

Building Grafana dashboards with AI, CLI and a bit of pragmatism

NASA retiree and crewmates splash down in Pacific after private ISS mission

LLMs fail to demonstrate internal world model, according to Harvard/MIT study

OpenZFS Bug Ported to Zig

Profiling Crosswords' Rendering Pipeline

Medieval medicine was smarter than you think – and similar to TikTok trends

The Machine Stops (1909) [pdf]

A bad ad for a device that photographed CRTs

Show HN: Opper AI – Task-Completion API for LLMs

Book Review: Arguments About Aborigines

O3 and Grok 4 Accidentally Vindicated Neurosymbolic AI

How I do it

Plasma 6.4 review – A worrying trend

EU-sponsored report says GenAI's 'fair use' defense does not compute

Meta's new Superintelligence lab is discussing training a closed model

Weight of an hourglass—Theory and experiment in quantitative comparison

A summer of security: empowering cyber defenders with AI

The CEO of Nvidia Admits What Everybody Is Afraid of About AI

Nvidia's CEO says it gained US approval to sell H20 AI chips to China

A Handshake in Orbit 50 Years Ago Transformed the Space Race

Show HN: ProjectD – Google Drive-based, AES-encrypted C2 in C/C++

Featured notebooks on selected topics in NotebookLM

Show HN: Write better AI system prompts

Fedora's i686 support gets a reprieve

Cellulose instead of crude oil: Researchers develop sustainable foams

Are personas even doing something when prompting?

Packages Proxies

Hyperreal Specifications for Continuous Sparse Data Computations [video]

Anthropic Economic Futures

Don't Eat Honey

A Unifying Framework for Robust and Efficient Inference with Unstructured Data

Building Grafana dashboards with AI, CLI and a bit of pragmatism

NASA retiree and crewmates splash down in Pacific after private ISS mission

LLMs fail to demonstrate internal world model, according to Harvard/MIT study

OpenZFS Bug Ported to Zig

Profiling Crosswords' Rendering Pipeline

Medieval medicine was smarter than you think – and similar to TikTok trends

The Machine Stops (1909) [pdf]

A bad ad for a device that photographed CRTs

Show HN: Opper AI – Task-Completion API for LLMs

Book Review: Arguments About Aborigines

O3 and Grok 4 Accidentally Vindicated Neurosymbolic AI

How I do it

Plasma 6.4 review – A worrying trend