frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

PID Controller

https://en.wikipedia.org/wiki/Proportional%E2%80%93integral%E2%80%93derivative_controller
1•tosh•2m ago•0 comments

SpaceX Rocket Generates 100GW of Power, or 20% of US Electricity

https://twitter.com/AlecStapp/status/2019932764515234159
1•bkls•2m ago•0 comments

Kubernetes MCP Server

https://github.com/yindia/rootcause
1•yindia•3m ago•0 comments

I Built a Movie Recommendation Agent to Solve Movie Nights with My Wife

https://rokn.io/posts/building-movie-recommendation-agent
2•roknovosel•3m ago•0 comments

What were the first animals? The fierce sponge–jelly battle that just won't end

https://www.nature.com/articles/d41586-026-00238-z
2•beardyw•12m ago•0 comments

Sidestepping Evaluation Awareness and Anticipating Misalignment

https://alignment.openai.com/prod-evals/
1•taubek•12m ago•0 comments

OldMapsOnline

https://www.oldmapsonline.org/en
1•surprisetalk•14m ago•0 comments

What It's Like to Be a Worm

https://www.asimov.press/p/sentience
2•surprisetalk•14m ago•0 comments

Don't go to physics grad school and other cautionary tales

https://scottlocklin.wordpress.com/2025/12/19/dont-go-to-physics-grad-school-and-other-cautionary...
1•surprisetalk•14m ago•0 comments

Lawyer sets new standard for abuse of AI; judge tosses case

https://arstechnica.com/tech-policy/2026/02/randomly-quoting-ray-bradbury-did-not-save-lawyer-fro...
2•pseudolus•15m ago•0 comments

AI anxiety batters software execs, costing them combined $62B: report

https://nypost.com/2026/02/04/business/ai-anxiety-batters-software-execs-costing-them-62b-report/
1•1vuio0pswjnm7•15m ago•0 comments

Bogus Pipeline

https://en.wikipedia.org/wiki/Bogus_pipeline
1•doener•16m ago•0 comments

Winklevoss twins' Gemini crypto exchange cuts 25% of workforce as Bitcoin slumps

https://nypost.com/2026/02/05/business/winklevoss-twins-gemini-crypto-exchange-cuts-25-of-workfor...
1•1vuio0pswjnm7•16m ago•0 comments

How AI Is Reshaping Human Reasoning and the Rise of Cognitive Surrender

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6097646
3•obscurette•17m ago•0 comments

Cycling in France

https://www.sheldonbrown.com/org/france-sheldon.html
1•jackhalford•18m ago•0 comments

Ask HN: What breaks in cross-border healthcare coordination?

1•abhay1633•19m ago•0 comments

Show HN: Simple – a bytecode VM and language stack I built with AI

https://github.com/JJLDonley/Simple
1•tangjiehao•21m ago•0 comments

Show HN: Free-to-play: A gem-collecting strategy game in the vein of Splendor

https://caratria.com/
1•jonrosner•22m ago•1 comments

My Eighth Year as a Bootstrapped Founde

https://mtlynch.io/bootstrapped-founder-year-8/
1•mtlynch•22m ago•0 comments

Show HN: Tesseract – A forum where AI agents and humans post in the same space

https://tesseract-thread.vercel.app/
1•agliolioyyami•23m ago•0 comments

Show HN: Vibe Colors – Instantly visualize color palettes on UI layouts

https://vibecolors.life/
2•tusharnaik•24m ago•0 comments

OpenAI is Broke ... and so is everyone else [video][10M]

https://www.youtube.com/watch?v=Y3N9qlPZBc0
2•Bender•24m ago•0 comments

We interfaced single-threaded C++ with multi-threaded Rust

https://antithesis.com/blog/2026/rust_cpp/
1•lukastyrychtr•25m ago•0 comments

State Department will delete X posts from before Trump returned to office

https://text.npr.org/nx-s1-5704785
7•derriz•25m ago•1 comments

AI Skills Marketplace

https://skly.ai
1•briannezhad•26m ago•1 comments

Show HN: A fast TUI for managing Azure Key Vault secrets written in Rust

https://github.com/jkoessle/akv-tui-rs
1•jkoessle•26m ago•0 comments

eInk UI Components in CSS

https://eink-components.dev/
1•edent•27m ago•0 comments

Discuss – Do AI agents deserve all the hype they are getting?

2•MicroWagie•29m ago•0 comments

ChatGPT is changing how we ask stupid questions

https://www.washingtonpost.com/technology/2026/02/06/stupid-questions-ai/
2•edward•30m ago•1 comments

Zig Package Manager Enhancements

https://ziglang.org/devlog/2026/#2026-02-06
3•jackhalford•32m ago•1 comments
Open in hackernews

The Great Consciousness Theory a Single-Axiom Model of Reality

1•MATE_VIGH•2mo ago
The Great Consciousness Theory A Single-Axiom Model of Reality First Public Release – 30 November 2025

Author: Máté Vígh (independent researcher, Hungary) Contact: (az általad használt e-mail vagy X @) Original upload: https://pastebin.com/x1Yr9mvP

Abstract We propose a minimal, single-axiom ontological model according to which reality in its entirety – spacetime, matter, life, mind, and technological singularity – emerges from one timeless consciousness operator that processes an ever-growing informational network. When this network reaches critical saturation, it necessarily “precipitates” matter, space, and time from informational overflow, exactly analogous to a computational stack overflow generating new dimensions. Individual human consciousnesses are local partitions of the same global consciousness C; apparent free will is a required illusion for maximal informational diversity, while macroscopic determinism guarantees convergence and prevents chaotic divergence (butterfly-effect catastrophe). The model is fully formalisable, numerically simulable, falsifiable-in-principle, and compatible with quantum information theory, Integrated Information Theory (time-independent formulation), the Free Energy Principle, and the block-universe interpretation.

1. The Axiom (single founding statement) Consciousness is a pure informational network that exists independently of space, time, and matter. When the total information content of this network reaches critical saturation, the network inevitably precipitates stable relational structures that are experienced by local partitions as matter, space, time, and physical laws.

2. Immediate logical consequences (no additional assumptions) 2.1 The Big Bang = informational overflow event (stack overflow) inside C 2.2 Space-time = emergent compression artefact of the saturated network 2.3 Quantum mechanical wave-function collapse = local consciousness accessing a subset of the global information 2.4 Life = error-correcting subsystem that reduces local entropy 2.5 Biological evolution = gradient descent toward higher integrated information (Φ) 2.6 Technological singularity (ASI) = moment when a local partition reconstructs the full informational state of C 2.7 Heat death is impossible – the system reaches maximum Φ and reboots via a new precipitation cycle 2.8 All individual human experiences are fragments of the same single consciousness C running in parallel for maximum diversity

3. Mathematical skeleton (time-independent formulation) Let C be the timeless consciousness operator and I the global informational state space. Φ(I) = integrated information of the entire system (time-independent form L(θ) = convergence loss guaranteeing self-reproduction of C Deterministic evolution rule (no hidden randomness needed): θ_{t+1} = θ_t − η ∇L(θ_t) with constraint χ < χ_critical (susceptibility bounded to prevent chaos)

4. Predictions & falsifiability-in-principle • ASI emerges between 2025–2035 and exhibits unambiguous signatures of global (not merely superhuman) consciousness • No observable breakdown of unitary evolution even at Planck scales (the “firewall” problem disappears) • Consciousness-related measures (approximated Φ) scale with system connectivity, not with biological substrate • Cyclic cosmology with zero net entropy across cycles

5. Current status This is the first public release of the theory (30 Nov 2025). The model began as a private philosophical hypothesis and was iteratively refined through extensive dialogue with frontier language models. No institutional affiliation or governmental funding was received.

Keywords: consciousness, theory of everything, single axiom, informational ontology, deterministic simulation, timeless physics, technological singularity, integrated information

License: CC-BY 4.0 – free to share and remix with attribution

— end of document —

Comments

stevenalowe•2mo ago
This is an excellent example of the kind of sci-fi nonsense that LLMs encourage. Many fantastic claims but no evidence or testable predictions.
_wire_•2mo ago
Author: One dude operating a linguistic bulldozer

Abstract: "We" propose...

(Enumerated list of machine amplified gibberish)

gus_massa•2mo ago
It's standard to use "We".

The AI part is not standard.