frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

Open in hackernews

An attempt at a standardized definition of consciousness

https://docs.google.com/document/d/1Tmd_3DXbnC2YovDHuMslTs681lN-goSB0NqAv9N3EK0/edit?usp=drivesdk
2•Trenthug•2h ago

Comments

Trenthug•2h ago
This paper presents a descriptive model of consciousness based on information theory The aim is to enumerate definitely described aspects of consciousness and present them in a coherent manner,talks about both phenomenal Consciousness (seen as attention)and access Consciousness,

Some aspects described about consciousness might seem akin to karl friston's view of defining brains as Bayesian inference machines, Wittgenstein's private language argument and Tononi's IIT usage of a complexity metric(amount of information) for measurement purposes with some subtleties of of it's own to add .

The part about choosing a method of measuring complexity seems to be something that can be looked over by a logician to get it refined for the models usage, some thought experiments with conjectures are there in that regard which might be worth looking into or to check their validity.

It seems there is a possibility for curation of a Complexity measuring method here,and the type that is required might be of interest to a logician/complexity theorist

Tell HN: C and C++ are absolute beasts when it comes to performance efficient

1•delduca•2m ago•0 comments

AWS Lambda now supports GitHub Actions to simplify function deployment

https://aws.amazon.com/about-aws/whats-new/2025/08/aws-lambda-github-actions-function-deployment/
1•mariuz•6m ago•0 comments

Mind Blowing Websites Hiding in Plain Sight

https://www.offscopes.com/
14•OffScopes•9m ago•1 comments

Block-Based Configuration Language

https://n16f.net/bcl/specification.html
1•gearnode•12m ago•0 comments

Technical Interviews are realigning with reality through AI

https://cendyne.dev/posts/2025-08-07-technical-interviews-are-realigning-with-reality-through-ai.html
1•furkansahin•13m ago•0 comments

GPT5 Is Horrible

https://old.reddit.com/r/ChatGPT/comments/1mkd4l3/gpt5_is_horrible/
1•druskacik•14m ago•0 comments

Robots.txt for the AI Era but Enforceable

https://aiprivacylicense.com/
1•nabanita•19m ago•1 comments

Air Force buying two Tesla Cybertrucks so it can learn to destroy them

https://www.theregister.com/2025/08/08/usaf_cybertruck_missile_tests/
2•ndsipa_pomu•24m ago•1 comments

Agentic Workflow: What's inside RAGFlow v0.20.0

https://medium.com/@infiniflowai/agentic-workflow-whats-inside-ragflow-v0-20-0-789fdd4397f6
1•vissidarte_choi•25m ago•0 comments

defer-import-eval: proposal for introducing a way to defer evaluate of a module

https://github.com/tc39/proposal-defer-import-eval
1•tilt•27m ago•0 comments

OpenAI is taking GPT-4o away from me – despite promising they wouldn't

https://community.openai.com/t/openai-is-taking-gpt-4o-away-from-me-despite-promising-they-wouldnt/1337378
2•Mzxr•31m ago•0 comments

Show HN: Streamed JSON Lines (A big JSON can't be streamed)

https://medium.com/@marius_18835/streamed-json-lines-using-laravel-crud-wizard-free-cd650e272caa
1•marius-ciclistu•35m ago•0 comments

Bsky Tracker: v2.5.0 Is Live

https://bsky.app/profile/bluesky-tracker.bsky.social
1•pavlostze•36m ago•1 comments

Malicious Ruby Gems Used in Targeted Credential Theft Campaign

https://socket.dev/blog/60-malicious-ruby-gems-used-in-targeted-credential-theft-campaign
1•amalinovic•49m ago•0 comments

HBO Max is going to get more annoying about password sharing

https://www.theverge.com/news/754357/hbo-max-password-sharing-annoying-earnings
2•tosh•49m ago•0 comments

Ask HN: GPT-5 still needs a second nudge for calculation?

1•chandlertsien•52m ago•0 comments

EU Artificial Intelligence Act

https://artificialintelligenceact.eu/
2•jonbaer•54m ago•0 comments

White Mountain Direttissima

https://whitemountainski.co/pages/white-mountain-direttissima
1•oftenwrong•54m ago•0 comments

Linux Desktop Share Tops 6% in 15M-System Analysis

https://www.zdnet.com/article/think-linux-desktop-market-share-isnt-over-6-this-15-million-system-scan-says-otherwise/
3•naves•59m ago•1 comments

OpenAI CEO Sam Altman says GPT-5 scares him – 'what have we done?'

https://www.tomsguide.com/ai/openais-ceo-sam-altman-says-gpt-5-is-so-fast-it-actually-scares-him-maybe-its-great-maybe-its-bad-but-what-have-we-done
3•pera•1h ago•3 comments

Tokenization in Large Language Models

https://seantrott.substack.com/p/tokenization-in-large-language-models
2•tokfan•1h ago•0 comments

Comuniq – A lightweight space for publishing and discussing specific topics

1•01-_-•1h ago•0 comments

Nature study on economic damages from climate change revised

https://www.pik-potsdam.de/en/news/latest-news/nature-study-on-economic-damages-from-climate-change-revised
1•01-_-•1h ago•0 comments

ChatGPT5 can't answer "How many states have R in it's name?"

https://bsky.app/profile/radamssmash.bsky.social/post/3lvtzdl343c2r
4•mattigames•1h ago•1 comments

How to Run Your Own OpenAI GPT OSS Server for Fun and Profit

https://northcodie.blogspot.com/2025/08/how-to-run-your-own-openai-gpt-oss.html
4•nickly•1h ago•2 comments

Understanding Late Binding in Python Closures

https://pythonkoans.substack.com/p/the-forgetful-calligrapher
2•meander_water•1h ago•0 comments

Loyalty programmes are keeping America's airlines aloft

https://www.economist.com/business/2025/08/06/how-loyalty-programmes-are-keeping-americas-airlines-aloft
2•jmsflknr•1h ago•0 comments

US Adds Surprise Gold Bar Tariff in Blow to Switzerland

https://www.bloomberg.com/news/articles/2025-08-08/us-hits-gold-bars-with-tariffs-in-blow-to-switzerland-ft-report
5•petethomas•1h ago•0 comments

Can't disable copilot code reviews

https://github.com/orgs/community/discussions/169148
2•TonyTrapp•1h ago•0 comments

Ask HN: What is "response-level error rate" and how is it measured?

2•myyke•1h ago•0 comments