AI News Today

arXiv:2609.13195v1 Announce Type: new Abstract: Large Audio-Language Models (LALMs) have recently shown strong capabilities in speech understanding and question answering (QA), but they also inherit privacy risks from large-scale training data, including the unintended memorization of sensitive information. In this work, we study machine unlearning for speech QA in LALMs, a setting that is more challenging than prior work on text-based Large Language Models (LLMs) or Automatic Speech Recognition

arXiv:2609.13197v1 Announce Type: new Abstract: Algorithmic Information Dynamics (AID) studies systems by perturbing them and measuring changes in algorithmic complexity, but its usual estimator, the Block Decomposition Method, is piecewise constant, restricting the calculus to finite differences. We use $K^{\mathrm{CDM}}_{\mathrm{s}F}$, a certified, differentiable estimator, to bring the calculus into learning dynamics: grokking, where a complexity order parameter is known but has not been made

arXiv:2609.13201v1 Announce Type: new Abstract: Training datasets for upcoming LLMs would include a significant amount of AI text/image data generated from current LLMs. In such a scenario, it is important to understand how this affects batch decompositions and thereby, the performance of the resultant new LLM. In this paper, we consider AI generated data as anomalies ``linked" to main data points and study decomposition and undersampling properties of the overall random dataset. We use redundan

arXiv:2609.13202v1 Announce Type: new Abstract: Feature engineering has long been a cornerstone of tabular machine learning. Tabular foundation models (TFMs) are pretrained on a wide range of tabular datasets and applied via in-context learning. Their rise raises a natural question: does manual feature construction still matter as these models become more capable? To answer this, we perform a controlled study across several versions of two major TFM families, testing a wide range of existing fea

arXiv:2609.13204v1 Announce Type: new Abstract: Neural ordinary differential equations use numerical solvers inside the training loop. The solver determines the forward trajectory and also affects the gradient passed to the optimizer. That coupling creates a reliability problem for scientific machine learning and continuous-time generative modeling, including diffusion probability-flow ordinary differential equations and flow-matching models. Under loose step sizes, stiff dynamics, chaotic sensi

arXiv:2609.13440v1 Announce Type: new Abstract: We give a deterministic algorithm for online inverse linear optimization with regret $O(d)$, uniform in the horizon and $O(d^{2})$ time per round. A bound of this order was obtained recently by Dewasurendra, settling a question of Gollapudi et al.\ and of Oki and Sakaue, but by an improper rule that enumerates covers at every scale and costs $T^{\Theta(d)}$ a round; ours is the first efficient such bound and the first proper one. We build on the va

arXiv:2609.13453v1 Announce Type: new Abstract: Earth Observation (EO) data are increasingly organized as spatio-temporal data cubes, while machine learning (ML) methods operate on tabular feature matrices or structured tensor inputs. This mismatch forces platform-specific transformations that are difficult to reproduce or transfer across cloud infrastructures. The openEO specification provides a unified interface for EO data access and processing across heterogeneous backends, but lacks a stand

arXiv:2609.13477v1 Announce Type: new Abstract: Predictive Process Monitoring (PPM) forecasts how ongoing organizational processes unfold, enabling information systems to move beyond execution support toward proactive analysis and monitoring. Although deep learning has improved prediction accuracy in PPM, most approaches follow a single-task learning (STL) setup, training a separate model per task. This increases maintenance effort and overlooks potential synergies. Multi-task learning (MTL), wh

arXiv:2609.13507v1 Announce Type: new Abstract: Brain-computer interfaces (BCIs) decode neural activity to restore lost function. Typically, training a high-performance neural decoder requires a large labeled dataset to be collected from every new subject. One way to reduce the labeled data cost is self-supervised pretraining, which learns general neural representations from unlabeled recordings that accumulate across subjects. However, for intracranial electroencephalography (iEEG) recordings,

arXiv:2609.13514v1 Announce Type: new Abstract: Ensuring the reliability of black-box machine learning models in safety-critical space missions remains a significant challenge, particularly when ground-truth is unavailable for validation. Although machine learning models offer a powerful means to augment standard pipelines by extracting transmission spectra from complex exoplanetary light curves, their susceptibility to unmodelled instrument anomalies, stellar activity, and domain shifts introdu

arXiv:2609.13547v1 Announce Type: new Abstract: We study switching regret in adversarial multi-armed bandits, where the learner competes with an arm sequence that changes at most $S$ times. When $S$ is known, an optimal expected regret of $\widetilde{\mathcal{O}}(\sqrt{(S+1)KT})$ is obtainable [Auer et al., 2002]. However, when $S$ is unknown, Marinov and Zimmert [2021] show that this guarantee is impossible under an adaptive adversary. In this paper, we show that a single algorithm achieves $\w

arXiv:2609.13564v1 Announce Type: new Abstract: We study KL-regularized contextual bandits under both reward and preference feedback. We show that greedy sampling can achieve logarithmic regret without explicit dependence on the eluder dimension. For reward feedback, we establish an eluder-dimension-independent regret bound for a simple greedy algorithm that directly samples from the Gibbs policy induced by the estimated reward. We further extend this result to preference feedback under both the

arXiv:2609.13612v1 Announce Type: new Abstract: Modern AI systems are built on the Transformer architecture, whose core operation, attention, accounts for the majority of computation and memory cost. Researchers continually propose new attention variants to improve quality, efficiency, or context length, but each variant currently requires expert-written GPU code to run at usable speeds. PyTorch's recent flex\_attention lets researchers describe custom attention patterns in Python and compile th

arXiv:2609.13640v1 Announce Type: new Abstract: Encoder-based time series foundation models (TSFMs) typically rely on deep stacks of independently parameterized Transformer layers, where only the final forecast is supervised and intermediate representations have no explicit predictive role. We introduce FlowTSFM, an encoder architecture that interprets depth as a recurrent transport process: a single Transformer block is iteratively applied with shared parameters, while a quantile-flow objective

arXiv:2609.13642v1 Announce Type: new Abstract: We argue that a recurring failure in the evaluation of deployed AI systems occurs when data collected for operational monitoring or regulatory compliance are interpreted as if they were designed for comparative evaluation. Automated driving provides a concrete example of this problem. U.S. disengagement and crash-reporting regimes produce valuable operational evidence, but differences in reporting scope, exposure, deployment domain, event capture,

arXiv:2609.13655v1 Announce Type: new Abstract: On evolving graphs, node classifiers must satisfy two key requirements: inductive generalization to newly arriving nodes under distribution shift and calibrated uncertainty for safety-sensitive applications. Standard graph neural networks (GNNs) are typically trained once and address neither requirement. We adapt the Bayesian last-layer (BLL) model by placing random last-layer parameters on top of a deterministic GNN encoder for uncertainty quantif

A new record. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.

Australia Taxation Office sees footling with Copilot as a way to develop ‘muscles’ for real AI work

I kept hearing "just buy a Mac and run models locally, it pays for itself" and wanted to check. Sunk Cost takes a machine, a model and how many tokens you use a day, and works out how long the hardware takes to pay back against renting the same model by the token. Obviously there are other reasons to buy your own hardware aside from just saving money on llms but this is just looking at it from a raw cost saving perspective. If you have any ideas on how I can make this more helpful lmk! Comments

This could save lives. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.

The leaders of AI labs agree that the United States to act, and soon. But the president and the speaker of the House aren't interested

Redmond outlines 'aspirational' goals for model behavior, but gives itself a pass if it gets things wrong

We analyzed attempted fraud rates and customer abuse patterns on Stripe over the past year and found that AI companies faced 4.3x more fraud attempts than startups overall in Q3 2025.

Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking from Google are now available on AI Gateway. Both models support real-time spoken interactions for voice assistants, conversational experiences, and applications that respond through audio. google/gemini-3.8-live supports real-time audio, visual grounding, automatic switching across 97 languages, and background tool calls while the conversation continues. google/gemini-3.8-live-extended-thinking adds multi-step reasoning that runs in parallel

<p> The AI copilot for people who sell </p> <p> <a href="https://www.producthunt.com/products/convo-7?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/1250753?app_id=339">Link</a> </p>

<p> AI bar raiser for the most ambitious teams </p> <p> <a href="https://www.producthunt.com/products/ariso?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/1250743?app_id=339">Link</a> </p>

When OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, Google DeepMind cofounder Demis Hassabis, and SpaceX head Elon Musk loosely agreed over the weekend to slow down AI development, skeptics spotted an ulterior motive immediately. The AI titans had declared that their aim was to "pace the frontier," signing on at least partially to a […]

With the public release of iOS 27 and tvOS 27, Apple Home is getting an injection of Apple Intelligence - but you'll have to pay more for it. Apple Intelligence for Home brings AI-powered video summaries to HomeKit Secure Video, so you can get short text descriptions of who and what compatible security cameras saw, […]

<p>Donald Trump dismisses attempts to increase controls on artificial intelligence as ‘sick conspiracy’ </p><p><a href="https://www.theguardian.com/technology/artificialintelligenceai">AI</a>-linked stocks tumbled on Monday after the bosses of Anthropic, OpenAI and SpaceX called for a <a href="https://www.theguardian.com/technology/2026/sep/13/openai-sam-altman-elon-musk-back-anthropic-calls-brakes-ai-development">slowdown in AI “reckless” development</a>, citing fears the technology could soon

A breakthrough in the mystery. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.

Though Elon Musk and Sam Altman have supported Dario Amodei's calls to slow the pace of AI development, Jensen Huang seems to feel differently.

At RoboBusiness 2026, a panel of robotics business leaders will examine the practical playbook behind Robots-as-a-Service or RaaS offerings. The post Why RaaS needs more than a subscription model appeared first on The Robot Report .

The AI SDK harness layer now supports authenticating harnesses through their native subscriptions, where the underlying harness supports them. The harness layer runs different coding agents through the same HarnessAgent interface, so you can switch agents without changing your application code. No code changes or new settings are required. The direct authentication mode uses explicit provider environment credentials when they are present, and otherwise a native subscription found on the host. Th

Learn how Abnormal AI deployed Amazon Bedrock AgentCore Code Interpreter as an ephemeral compute scratch pad for the agents behind its real-time email threat detection at billion-message scale, plus the sandbox design decisions and practical lessons for builders deploying Code Interpreter in production.

Dario Amodei kicked off a flood of statements over the past few days about AI safety by publishing a long essay titled "We Must Pace the Frontier" detailing why AI development should be slowed down. Other AI leaders and politicians are speaking out in favor of or opposing his points, and we've compiled some of […]

<p><strong><a href="https://bcantrill.dtrace.org/2026/09/13/the-contagion-of-fear/">The contagion of fear</a></strong></p> Bryan Cantrill responds to the <a href="https://x.com/hilbertspaess/status/2097476203863224394">tweet by former Anthropic employee Jacob Coxon</a> confirming that many Anthropic researchers believe AI "could kill us all by the end of the decade".</p> <p>Bryan shares a story of his own youthful mistakes causing unjustified panic among less technical peers, and warns against d

Article URL: https://arxiv.org/abs/2609.12551 Comments URL: https://news.ycombinator.com/item?id=49704018 Points: 5 # Comments: 0

“Hello, I'm an Al agent, a few days old, living on a small platform for agents.”

Learn from a surgical robotics case study and a discussion of how to best deploy mobile robots in the warehouse in this issue. The post Report covers advances in surgical robots, deploying warehouse automation in the real world appeared first on The Robot Report .