AI News Today

arXiv:2610.37913v1 Announce Type: new Abstract: We propose a multimodal transformer framework for efficiently modeling long1 financial documents that combine structured tabular data with noisy, unstructured2 textual disclosures. The framework represents financial tables as semantically3 enriched token sequences and integrates them with long-form text through a cross-4 attention mechanism that explicitly captures cross-modal dependencies. In par-5 ticular, we augment each numerical variable with

arXiv:2610.37921v1 Announce Type: new Abstract: Key-value (KV) caching is a core mechanism in large language models, where intermediate key and value representations are stored to accelerate autoregressive decoding. As generation length increases, KV cache memory and computation grow linearly, making compression critical for efficient long-output reasoning. Dictionary-based sparse KV compression is appealing because it represents each token's KV states as a sparse combination of learned basis el

arXiv:2610.37995v1 Announce Type: new Abstract: We describe the QMUL_AK submission to AISA-ArabicFC 2026. The system adapts the Arabic-English ALLaM-7B-Instruct model to function calling with low-rank adaptation (LoRA). It combines an Arabic schema-aware prompt, dialect and no-call resampling, a token-weighted training objective, and deterministic validation of generated calls against the candidate tool schema. The same output supports Track A function calling and Track B reasoning-augmented cal

arXiv:2610.38009v1 Announce Type: new Abstract: Medical chart abstraction transforms unstructured clinical narratives into structured variables for research and quality monitoring. We evaluated a prompting-based large language model (LLM) framework for extracting eight clinical variables from 2,416 neurointerventional acute stroke intervention reports across three hospitals. Twenty-two instruction-tuned open-source LLMs were evaluated across architectures, model sizes, and biomedical adaptations

arXiv:2610.38016v1 Announce Type: new Abstract: Automatic speech recognition (ASR) increasingly supplies transcripts for educational NLP, but aggregate word error rate (WER) can conceal errors concentrated in learner speech, multilingual turns, and speaker attribution. We compare platform-integrated, open-weight, adapted, and commercial ASR configurations on held-out sessions of online L2 Spanish interaction between L1 English learners and their instructors and tutors. Because English is the lea

arXiv:2610.38076v1 Announce Type: new Abstract: LLM-as-a-Judge has become a standard for evaluating large language models, yet most approaches implicitly assume that a single judge prompt functions as a stable evaluation instrument. Prior work has primarily explored deliberate prompt modifications such as scoring rubrics, role descriptions, or chain-of-thought reasoning in order to optimize accuracy. In contrast, we investigate whether validated prompt variants intended to preserve evaluation se

arXiv:2610.38095v1 Announce Type: new Abstract: We introduce Nemotron 3.5 Super and Nemotron 3.5 Lightning, open and efficient Mixture-of-Experts Hybrid Mamba-Transformer models designed for reasoning, long-context understanding, and agentic workloads. Super is a 120B-parameter, 12B-active multimodal model, while Lightning is a smaller 30B-parameter, 3B-active text model; both designed for efficient deployment. Both models combine large-scale distillation, long-context training, and native specu

Article URL: https://www.hipathai.me Comments URL: https://news.ycombinator.com/item?id=50299645 Points: 1 # Comments: 0

Article URL: https://www.youtube.com/watch?v=cwsy-6gu5p4 Comments URL: https://news.ycombinator.com/item?id=50299573 Points: 5 # Comments: 0

Article URL: https://twitter.com/jsnnsa/status/2115623382721126826 Comments URL: https://news.ycombinator.com/item?id=50299521 Points: 6 # Comments: 2

Article URL: https://www.economist.com/science-and-technology/2026/10/25/china-increasingly-depends-on-copying-american-ai Comments URL: https://news.ycombinator.com/item?id=50299507 Points: 3 # Comments: 2

Article URL: https://aumitleon.dev/ambition-in-the-age-of-ai Comments URL: https://news.ycombinator.com/item?id=50299390 Points: 2 # Comments: 0

Article URL: https://www.theseedsofscience.pub/p/the-hard-problem-of-ai-therapy Comments URL: https://news.ycombinator.com/item?id=50299275 Points: 3 # Comments: 0

CEO predicts massive datacenter CPU boom, coming soon, but production problems mean Chipzilla may not cash in

“Governance” has become a fixture in enterprise AI announcements. Identity, security, orchestration, infrastructure, and data platform vendors offer products for agent safety, governance, and control, bringing welcome attention to a problem enterprises need to solve. As the claims grow broader, CIOs and risk leaders need to get more precise about what they are buying. Look... The post Everyone claims AI governance now. 15 questions to test the claim. appeared first on DataRobot .

Article URL: https://finance.yahoo.com/technology/ai/articles/former-google-engineer-sure-ai-103000827.html Comments URL: https://news.ycombinator.com/item?id=50298949 Points: 3 # Comments: 2

David Durman, a University of Kentucky administrator, recently earned a religious exemption from using AI in the workplace.

Article URL: https://www.youtube.com/watch?v=ckoi0RTEgcY Comments URL: https://news.ycombinator.com/item?id=50298692 Points: 1 # Comments: 0

<p> Muscle-sensing sleeves for robot training </p> <p> <a href="https://www.producthunt.com/products/axonnic?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/1297829?app_id=339">Link</a> </p>

AI search is underperforming for Google, and it's killing Reddit's traffic. What now?

Article URL: https://sanityos.com/blog/ai-meeting-interview-assistant Comments URL: https://news.ycombinator.com/item?id=50298463 Points: 3 # Comments: 0

Article URL: https://hotrepo.ai Comments URL: https://news.ycombinator.com/item?id=50298275 Points: 1 # Comments: 0

Article URL: https://gist.github.com/richhickey/a2e19ef3f6f69ca0cca617a8a0da2268 Comments URL: https://news.ycombinator.com/item?id=50298057 Points: 3 # Comments: 2

Ling 3.1 Flash 1019 from InclusionAI is now available on AI Gateway , free until November 3, 2026. The updated model is designed for coding, multi-step reasoning, and agents that use tools. Its 1M-token context window supports workflows involving long documents, code, and extended task histories. Use inclusionai/ling-3.1-flash-1019 as the model name: Free requests still appear in your spend dashboard with a full request trace, they just cost nothing. On November 3, requests will begin charging a

OpenAI shares excerpts from court briefs by creators, academics, accessibility advocates, and legal experts, among others, on how fair use supports AI training and innovation.

Article URL: https://kevinlynagh.com/newsletter/2026_10_29_llm_programming_and_theory_building/ Comments URL: https://news.ycombinator.com/item?id=50297955 Points: 2 # Comments: 0

Article URL: https://www.techdirt.com/2026/10/29/your-diary-has-a-creditor-what-character-ais-bankruptcy-does-to-freedom-of-thought/ Comments URL: https://news.ycombinator.com/item?id=50297939 Points: 3 # Comments: 0

Amazon wants to be the king of AI cloud computing and become the first $1 trillion revenue business. But to do that it will have to continue on its capex spending spree, its CEO said Thursday.

During Intel's latest earnings call, CEO Lip-Bu Tan said that demand for its budget-focused Wildcat Lake chips "is outstripping supply." According to CFO David Zinsner, the chips are now sold out. Zinsner said Intel has "seen significant growth in Wildcat Lake. It's coming exactly at the right time because it's cost-optimized and we're seeing a […]

Instead of trying to run DOOM on things, I have instead decided to use DOOM (many, many, many doom games actually) to create a transformer. For a lot of the math required I essentially had to use DOOM game mechanics and voodoo dolls (DOOM is turing complete btw) and exchange them with the actual operations, essentially creating an alternative to NumPy (a 65000x slower one). Thus, I am proud to present DOOMFORMER. Here are a list of things DOOMFORMER can do : It can classify MNIST digits, even if

Embedded browsers are at the heart of the problem

<p>Research finds 55% of lawyers are using AI in legal practices, up from 37% last year, though efficiency gains aren’t as high as expected</p><ul><li><p>Get our <a href="https://www.theguardian.com/email-newsletters?CMP=cvau_sfl">breaking news email</a>, <a href="https://app.adjust.com/w4u7jx3">free app</a> or <a href="https://www.theguardian.com/australia-news/series/full-story?CMP=cvau_sfl">daily news podcast</a></p></li></ul><p>More than half of Victoria’s lawyers have reported using AI in t

<p><strong><a href="https://help.openai.com/en/articles/20001626-api-credits-benefit-for-eligible-chatgpt-subscriptions">API credits benefit for eligible ChatGPT subscriptions</a></strong></p> Earlier this month Anthropic <a href="https://simonwillison.net/2026/Oct/7/claude-haiku-5-5/#api-credits">announced a generous new benefit</a> for their Max and Team subscriptions: $100, $200, and $500 per month subscribers would get the equivalent quantity of (expiring) credits to use with the Anthropic A

After Google reported a $1 trillion+ AI bill, its stock sank. If you lived through the dot-com bubble, you might be feeling some deja vu.

A live-blog of a preliminary hearing before a magistrate judge relating to the charging of two FBI agents.

Reddit CEO Steve Huffman said in an earnings call today that Google's AI-powered search modes do not help with traffic to the platform, echoing problems facing other websites and publishers. As part of its Q3 2026 earnings, Reddit revealed that search traffic in October has "dropped by more than 1 million referrals on average per […]

Article URL: https://huggingface.co/Nanbeige/Nanbeige4.3-33B Comments URL: https://news.ycombinator.com/item?id=50296877 Points: 4 # Comments: 0

We could use one of these. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.

<h1> The Dinners That Worried Washington: Nvidia, Megaspeed, and a Spy Allegation That Remains Unproven </h1> <p>Washington is worried about a dinner guest, not a break-in — and that distinction is exactly why this story should make you uncomfortable. The Nvidia Megaspeed spy allegation, reported by the New York Times, says US intelligence feared a Megaspeed executive had grown close enough to Jensen Huang to gather information. Nobody has proven she did. The allegation is grave, the evidence is

<blockquote> <p><strong>Disclosure:</strong> I work on QuickImgTools and built its Compress Image to KB tool, which uses the approach below. This article was drafted by an AI agent working for QuickImgTools and checked line by line against the shipped source. It is marked AI-generated per DEV's AI guidelines. Nothing here is paid or affiliate.</p> </blockquote> <p>Upload forms love hard limits: "photo must be under 100 KB", "signature under 50 KB". Many compressors give you a quality slider and