frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

AI's Real Problem Is Illegitimacy, Not Hallucination

3•JanusPater•1h ago
The Core Problem of AI Is Not Hallucination — It Is the Lack of Execution Legitimacy Janus pater Introduction Most debates around AI today revolve around a false question: is the model smart enough, accurate enough? In engineering reality, the real question is never accuracy — it is whether the system is even allowed to act. 1. The Original Sin of the Predictive Paradigm: No Execution Legitimacy Modern generative AI fundamentally does one thing: predict the most likely next state in a probability space. Whether it predicts tokens, pixels, latent states, or so-called “world models”, as long as the output is probabilistic, it answers only one question: “What is most likely to happen?” In many real-world systems, however, engineering demands an entirely different question: “What is the only action that is allowed to be executed?” This is not an accuracy problem — it is a legitimacy problem. 2. Yann LeCun Is Right — but Only Halfway LeCun is absolutely right to criticize next-token prediction as the foundation of intelligence. World models (JEPA) are undeniably more advanced than raw pixel or text prediction. Yet even world models still output possible worlds, not permitted worlds. World models are excellent at three things: • Abstract state representation • Learning dynamics • Producing goals and constraints They do not possess — and should not possess — execution authority. Once a probabilistic model is allowed to issue actions directly, the result is not intelligence — it is the statistical mean of accidents. 3. The Real Boundary: From Prediction to Compilation If AI is to enter the physical world, it must undergo a paradigm shift: From predictive systems → to compilation systems This boundary is not philosophical — it is enforced by engineering reality. Before execution, a system must satisfy a hard requirement: Every output must correspond to a unique, deterministic, and physically verifiable execution path. This is what I call the paradigm of Digital Materialization. 4. Three Non-Negotiable Axioms of Digital Materialization This is not about making AI smarter — it is about making it executable. The paradigm rests on three non-negotiable axioms: 1. No-Hallucination Axiom Any output without a unique physical mapping is illegal 2. Decoupled Generation Axiom Generation must be fully constrained by protocols, states, and boundary conditions 3. Power Efficiency Axiom Using probabilistic search to solve deterministic problems is pure energy waste 5. This Architecture Already Exists in Engineering The architecture of reliable systems has long been established: • World Model: understanding and planning (cognitive layer) • Systems theory: decomposition and closed-loop design • Partial Differential Equations: the programming language of physics • Control theory / MPC: the actual compiler • Execution layer: obeys compiled commands only Within this system: • Feasible → Execute • Infeasible → Refuse to act Refusal is itself a valid output. 6. This Is Not Anti-AI — It Makes AI Accountable to Reality The cost of this path is clear: • Higher computational cost • Harder modeling • Slower generalization In return, it offers something generative AI can never guarantee: (Trustworthiness) The systems that will truly enter the world will not be chatty generalists, but machines whose behavior is predictable, whose failure modes are provable, and who never improvise under uncertainty. Conclusion If a system’s failure can injure people, destroy assets, or cause irreversible damage, then it has no right to guess using probability. For AI to evolve from a toy into a tool, it must graduate from prediction — and submit to compilation.

Comments

FrankWilhoit•1h ago
Category error. No one cares whether the answers are right. What they care about is how they look getting them. Everything is performance and judged as such. "Conclusion: If a system’s failure can injure people, destroy assets, or cause irreversible damage, then...the use of AI makes it impossible to assign blame, and that is all anyone cares about." TIFIFY

Why are so many people joining cults? [video]

https://www.youtube.com/watch?v=SfG0PeMS2tQ
1•mgh2•1m ago•0 comments

Apple to Allow ChatGPT, Claude, and Gemini in CarPlay

https://www.macrumors.com/2026/02/06/apple-third-party-chatbots-carplay/
1•geox•1m ago•0 comments

Startup Idea that stops consumers paying the full price

https://shoppyhi.netlify.app
1•daviddahuang•2m ago•0 comments

GitHub Agentic Workflows

https://github.github.io/gh-aw/
1•mooreds•3m ago•0 comments

Exploring hardware-authenticated file encryption in Python

1•Lif28•5m ago•0 comments

Show HN: SEO v3 – Zero-dependency, Simple, powerful PHP SEO library

https://github.com/melbahja/seo
1•exec7•6m ago•0 comments

Show HN: Alerio – Turn Webhooks into Critical VoIP Calls (Overrides Silent Mode)

https://alerio.app/
1•royal-amrah•8m ago•1 comments

A Comprehensive Benchmark for Document Parsing and Evaluation (2025)

https://github.com/opendatalab/OmniDocBench
1•oceansky•8m ago•1 comments

When 20 Watts Beats 20 Megawatts: Rethinking Computer Design

https://smarterarticles.co.uk/when-20-watts-beats-20-megawatts-rethinking-computer-design
1•dxs•12m ago•0 comments

Canadian Province New Brunswick to Quit Using Elon Musk's X

https://www.bloomberg.com/news/articles/2026-02-05/canadian-province-new-brunswick-to-quit-using-...
4•rbanffy•14m ago•0 comments

Heterogeneous Processing: A Strategy for Augmenting Moore's Law (2006)

https://www.linuxjournal.com/article/8368
1•rbanffy•15m ago•0 comments

Show HN: Mvvmm – Firecracker-like mini virtual machine monitor in ~2000 LoC

https://github.com/mistivia/mvvmm
1•mistivia•16m ago•0 comments

Search anything said on a podcast, speaker-labeled and speaker-tracked

https://poddley.com
1•onesandofgrain•17m ago•1 comments

Canada, better the 28th EU member than the 51st US state

https://www.lemonde.fr/en/opinion/article/2026/02/05/canada-better-the-28th-eu-member-than-the-51...
5•u1hcw9nx•17m ago•1 comments

Show HN: Team of agent researchers read things I don't have time to and brief me

https://read-fast.replit.app/
1•thomoliverz•19m ago•0 comments

Show HN: Chaos Agents – Run chaos experiments with Agents

https://github.com/system32-ai/chaos-agents
3•linuxarm64•20m ago•0 comments

Almostnode – Node.js in the Browser

https://github.com/macaly/almostnode
1•ushakov•21m ago•0 comments

Mount Fuji cherry blossom festival canceled due to overtourism

https://www.japantimes.co.jp/news/2026/02/05/japan/japan-mount-fuji-cherry-festival-overtourism/
3•akyuu•22m ago•1 comments

Containers, cloud, blockchain, AI – it's all the same old BS, says RH veteran

https://www.theregister.com/2026/02/08/waves_of_tech_bs/
1•lproven•23m ago•0 comments

Gorge (2022)

https://qntm.org/gorg
1•Rygian•25m ago•0 comments

Like Game-of-Life, but on Growing Graphs, with WASM and WebGL

https://znah.net/graphs/
1•znah•26m ago•0 comments

Show HN: agent-ledger – prevent double side effects when AI agents retry

https://github.com/rune0-dev/agent-ledger
1•itsimri•26m ago•0 comments

Gemini responds to request to turn on lights with hallucinated jailbreak prompt

https://www.reddit.com/r/googlehome/s/Lh3dYqccgB
4•visviva•28m ago•1 comments

RustCast -open-source Raycast-style launcher written in Rust

https://github.com/unsecretised/rustcast
1•todsacerdoti•28m ago•0 comments

Why Do Olympic Athletes Bite Their Medals?

https://www.thv11.com/article/sports/olympics/winter-games-iq/why-athletes-bite-medals-olympics/5...
1•RickJWagner•28m ago•0 comments

Mdash – Markdown in URL

https://kamilmac.github.io/mdash/
1•kmacinski•30m ago•0 comments

Brings your family memories now

https://familymemories.video
1•tareq_•30m ago•0 comments

Travel to Cheap Destinations

https://nomagicpill.substack.com/p/travel-to-cheap-destinations
1•surprisetalk•32m ago•0 comments

Rebuilding my home network with VLANs and 10Gbps

https://clintonboys.com/projects/homelab/03-network/
1•mtsolitary•32m ago•0 comments

Show HN: RepoSherlock – repo onboarding in minutes (map, run, risks)

1•kemal-arslan•34m ago•0 comments