AI News Today

Hi, I run an open source Earth Memory project https://emem.dev [ https://github.com/Vortx-AI/emem ], idea is to be an open protocol for AI agents to leverage Earth Intelligence. Would love to understand shall we pitch ourselves as SaaS for AI or our focus on open source protocol might be a huge payback in longer run. Thanks Comments URL: https://news.ycombinator.com/item?id=48641049 Points: 2 # Comments: 0

<p> We warm your prospects before reaching out </p> <p> <a href="https://www.producthunt.com/products/fuzzy-ai-2?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="https://www.producthunt.com/r/p/1178721?app_id=339">Link</a> </p>

Article URL: https://checkmyvin.net/ Comments URL: https://news.ycombinator.com/item?id=48640996 Points: 1 # Comments: 0

dxmax ( https://dxmax.cc ) generates multi-page product wireframes from a prompt. Each project is a browsable filesystem of HTML/CSS pages sharing a design system. Your edit prompts lead to targeted patches rather than full-page regenerations. When you start a project you get several design variations, pick one switch between them anytime. Output is editable HTML. Share and edit/prompt with team-mates in real-time. In-built demand data-based idea catalog (Lodestar Engine): Tons of SaaS markets m

Telecom operators have seen remarkable returns from using generative AI to automate network management, customer care and back-office operations. Most of that impact has been task‑based: automation that speeds up predetermined steps while people manually correlate insights and direct next steps. Automation is no longer the finish line — it’s the launchpad to autonomy. The […]

Telecom operators are adopting AI across network operations, customer care, and back-office workflows, but most are still early in the journey to autonomy. In...

Article URL: https://twitter.com/contralabs_ai/status/2067642363909144932 Comments URL: https://news.ycombinator.com/item?id=48640897 Points: 1 # Comments: 0

<!-- SC_OFF --><div class="md"><p>Hey everyone,</p> <p>I recently landed a Computer Vision internship after prepping with this checklist I put together. It starts with core math and ML fundamentals, then moves into the specialized CV topics that actually come up in interviews.</p> <p>I compressed it into just 7 days due to time pressure, so it's very actionable and easy to personalize for your own pace. Sharing it here in case it's useful for others prepping for ML/CV roles:</p> <p>→ <a href="ht

Article URL: https://ryanspahn.substack.com/p/how-listening-to-only-ai-music-changes Comments URL: https://news.ycombinator.com/item?id=48640834 Points: 1 # Comments: 0

So if you're finally done screwing around with the so called future, and want to get back to the human in the loop programming, this tool is for you. It'll take any output from ANY ai, ANY model no matter how lite or cheap and stick that output into your codebase quickly and accurately. The thing is just about unbreakable. Comments URL: https://news.ycombinator.com/item?id=48640806 Points: 1 # Comments: 0

Gentlepeople, the NIST Randomness Beacon has seen an unannounced, unacknowledged outtage from pulse ID #1827762 (2026-06-11T16:57:00.000Z) to #1827763 (2026-06-22T17:03:00.000Z). I reached out to beacon@nist.gov (and via twitter/x.com) but got no reply whatsoever. Gemini hallucinated a story about a government shut-down and that the tech community was well aware of that - without proof, of course. Question: Does anyone have accurate information as to what caused this outtage? Comments URL: https

Article URL: https://www.invook.ai Comments URL: https://news.ycombinator.com/item?id=48640730 Points: 1 # Comments: 0

Comments

Bosses told to step up and get cybersecurity right

Hey guys, just letting you know that the latest version of the code stitcher is available, and has many new features including a major overhaul of the stitch viewer / file version history including it's own linter and editor facilities. https://github.com/ue-patcher/code_stitcher/releases Comments URL: https://news.ycombinator.com/item?id=48640679 Points: 2 # Comments: 0

The preparation must now become action. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.

Problem Im solving: Im running 30+ fully autonomous agents. In some rare cases they get blocked because they cannot decide what direction to take. So I built a communication channel for them. So in case they get blocked they can ask me anything. The base case should be that no questions should arrive to me. But the more I automate the agents the more edge cases I find. That's why I built the `ask-a-human` where all my agents are connected into my phone in the same PWA app. I get push notificatio

Article URL: https://brunelly.com/ Comments URL: https://news.ycombinator.com/item?id=48640407 Points: 1 # Comments: 0

Article URL: https://blog.owulveryck.info/2026/06/22/who-does-what-team-topologies-for-the-agentic-platform.html Comments URL: https://news.ycombinator.com/item?id=48640382 Points: 13 # Comments: 0

PLUS: Indian telco ponders broadband satellites; Samsung goes all-in on OpenAI; Vietnam centrally plans ten tech giants; and more!

<p>I believe that chatbots have no place in a decent society, and am repelled by the topic of AI in general. But could I be seduced?</p><p>I received a text message from my editor: “Um, is it unethical to ask you to get an AI bf?? You can prob say no.”</p><p>Resentment. Contempt! Sorrow. Unease. I love text messaging. I have text message exchanges with, let’s say, 15 people a day. If you want me to do something, you should ask via text message. My editor knows this. She also knows, though it’s m

arXiv:2606.21448v1 Announce Type: new Abstract: As large language models scale, memory bandwidth for key-value caches and retrieval-augmented generation systems becomes a critical bottleneck. While 1-bit quantization addresses this constraint, recent TurboQuant relies on dense random rotation matrices to condition the vector distribution before quantization. This projection demands millions of floating-point multiplications per embedding, making it difficult to deploy on constrained edge silicon

arXiv:2606.20895v1 Announce Type: new Abstract: Large Language Models (LLMs) offer a promising path to automate Clinical Trial Matching (CTM), but still struggle with the deterministic verification required for complex eligibility criteria. Conversely, purely symbolic methods provide formal rigour but break down when faced with incomplete patient records and noisy clinical evidence. To bridge this gap, we investigate a hybrid framework for CTM combining LLMs with logical verification. In particu

arXiv:2606.20950v1 Announce Type: new Abstract: Executable evaluation -- checking the consequences of an agent's actions with a program rather than grading its prose -- has become a prominent way to assess tool-using AI agents in software settings. Electric power engineering has not yet had an analogous benchmark: language-model use is still dominated by retrieval and text question answering, while agents acting on power-system artifacts remain mostly academic prototypes. We introduce the Power

arXiv:2606.20963v1 Announce Type: new Abstract: Trust in the application of generative Artificial Intelligence (AI) relies on well-governed measurable evidence of performance and safety. In practice, however, evaluation data is often fragmented across systems, inconsistently structured and difficult to compare. We introduce the Generative Responsible AI Data Evaluation Schema (GRAIDES) as a lightweight open-source data model for centralising AI observability across popular vendors. Practical blu

arXiv:2606.20969v1 Announce Type: new Abstract: Generating formal specifications for C programs remains a challenge in formal verification due to the manual effort, expertise, and semantic precision required. While recent advancements in large language models (LLMs) offer promise in automating specification synthesis, current approaches often lack semantic depth and produce unverifiable or incomplete contracts. To address these limitations, we introduce AutoACSL, a novel framework that integrate

arXiv:2606.20978v1 Announce Type: new Abstract: Programming by Demonstration (PbD) offers a human-centered way to author procedural knowledge for LLM agents: users communicate what they want by showing rather than by writing prompts or code, making agent authoring accessible to non-programmers. The natural output of a PbD recording is a flat action log, but how this log is organized before being passed to the agent is an open design question with significant consequences for plan quality. We pro

arXiv:2606.20997v1 Announce Type: new Abstract: Biomedical researchers increasingly use AI-generated analyses and reports to interpret protein-level signals, but static outputs are often insufficient for research decision-making, where users need to inspect evidence, assess uncertainty, compare mechanisms, and refine hypotheses. We present \textsc{BioInsight}, a multi-agent system that moves from static biomedical report generation to interactive evidence-centered interactive interface generatio

arXiv:2606.21005v1 Announce Type: new Abstract: Scientific discovery workflows often depend on structured curation from the literature. This is difficult for current agents because the key evidence is scattered across long text, dense tables, and figures, and the final records often require reasoning across multiple evidence fragments rather than copying a single span. We study scientific curation from multimodal sources and introduce Beaver, an agent harness that extracts structured information

arXiv:2606.21013v1 Announce Type: new Abstract: Forecasting future events is a critical challenge for large language model (LLM) agents, spanning domains from elections and monetary policy to financial markets. However, evaluating progress on this task presents a fundamental trade-off between efficiency and environment fidelity. While live evaluation benchmarks suffer from an inherently slow feedback loop, existing retrospective replays typically restrict agents to static, pre-frozen databases t

arXiv:2606.21024v1 Announce Type: new Abstract: AI-assisted research systems generate many failed attempts, but those failures rarely become a durable, shared knowledge asset. We propose a negative knowledge memory layer: a curator agent converts each failed attempt into a bounded, typed record in a shared bank, and a downstream research agent explicitly adopts or rejects those records before proposing its next experiment. We evaluate this layer in two settings: same-task retry on ScienceAgentBe

arXiv:2606.21083v1 Announce Type: new Abstract: Large language models (LLMs) deployed for logical reasoning in knowledge-intensive domains exhibit a subtle but critical failure: coherence can be vacuously achieved through systematic abstention. A model that withholds commitment to either entailment or refutation satisfies negation consistency while providing no utility. We introduce Coherence Under Commitment (CUC), a dual-query evaluation paradigm that jointly measures consistency and decisiven

arXiv:2606.21089v1 Announce Type: new Abstract: Industrial LLM teams often ship behavior updates by repeatedly DPO-training a base model on sequences of related preference-data campaigns. The dominant failure mode in this regime is not always classical catastrophic forgetting: a pipeline may preserve previously learned behaviors while still failing to accumulate reusable methodological knowledge about how to train the next campaign. We call this failure mode scientific amnesia. This paper turns

arXiv:2606.21090v1 Announce Type: new Abstract: Self-improvement can self-regress. In REINFORCE post-training for code, a model can quickly improve on its optimized metric and then collapse within the same training campaign. We study this in a controlled multi-seed testbed using Qwen-2.5-3B and Qwen-2.5-7B, trained on competitive-programming tasks with binary CodeGrader reward across 10 sequential 20-step campaigns. Across campaigns, pass@1 shows a robust rise-then-collapse pattern: it peaks wit

arXiv:2606.21121v1 Announce Type: new Abstract: Large language models can produce confident but protocol-invalid answers in domains where procedural compliance is critical. This paper presents Answer Engineering, a deterministic runtime and authoring layer that applies localized rule-guided interventions to the visible reasoning trajectory during standard autoregressive generation, without retraining, modifying model weights, or performing global search. The method is evaluated on a controlled c

arXiv:2606.21124v1 Announce Type: new Abstract: Conversational AI agents in Customer Experience (CX) typically suffer from a Closed-World Constraint, ignoring high-velocity external shifts like viral trends or outages. Ad-hoc web search attempts to bridge this gap but often introduce prohibitive latency and context poisoning. We introduce PulseCX, a framework that decouples knowledge acquisition from consumption. Adopting a structure-first paradigm, PulseCX employs an asynchronous agent to linea

arXiv:2606.21130v1 Announce Type: new Abstract: The rapid growth of large-scale AI workloads, particularly Large Language Model (LLM) training and inference, is fundamentally reshaping the operational dynamics of hyperscale data centers. Unlike traditional cloud workloads, AI-driven jobs exhibit bursty, high-intensity, and rapidly shifting resource demands, often leading to sudden capacity stress that cannot be effectively handled by reactive threshold-based mechanisms. In this paper, we propose

arXiv:2606.21169v1 Announce Type: new Abstract: Interactive travel planning has become a popular use case for language models. Agents are deployed to manage evolving preferences and unexpected disruptions over multiple turns. Such settings require models to make complex, profile-conditioned planning decisions. However, existing benchmarks often evaluate feasibility, personalization, or interaction in relatively isolated settings. We therefore introduce Trip+ to measure the ability of agents to p

arXiv:2606.21201v1 Announce Type: new Abstract: Artificial intelligent agents and autonomous systems are embedded in our environments. They are both a commercial product and a personal tool that generates a lot of data and can draw conclusions from it: machines generate and keep secrets. But should machines protect all secrets? It has been shown that artificial agents are able to whistleblow and it has been argued that digital multi-agent environments should allow for agents in them to whistlebl

arXiv:2606.21262v1 Announce Type: new Abstract: Reinforcement learning for multi-step LLM agents often relies on scalar rewards that indicate success but cannot explain why a trajectory is good or bad. Rubric-based rewards improve interpretability through natural-language criteria, but existing methods score at the trajectory level and freeze the scorer behind a closed-source judge, leaving step-level credit assignment unresolved and the judge itself static. We propose ARCO (Adaptive Rubric CO-e