Models, products and AI startups, near-duplicates collapsed to the most credible source
SK Hynix Pulls Solidigm's $10 Billion IPO on Regulatory PushbackSK Hynix scrapped a planned US listing of its Solidigm unit valued at about $10 billion after strong opposition from regulators and shareholders. It becomes the latest company to pull an IPO as markets swing on doubts about the AI boom.
ThinkingBox uses database checks as RL rewardsThe team reused ThinkingBox's executable checks of final database state as training rewards, building an offline task pool. With 50 GRPO steps on Qwen3.8-27B, average success rose from 51.70% to 60.78%, and tasks solved correctly in all 20 attempts went from 38 to 107.
US Treasury proposes curbs on China tech investmentThe US Treasury proposed a rule expanding outbound-investment restrictions on AI, supercomputing, semiconductors, quantum and biotech in China and other countries of concern, implementing the 2025 Comprehensive Outbound Investment National Security Act. It applies only to investments above $50,000 and excludes many financial transactions.
Pivotal Advisors CEO: AI Trade Needs Revenue ProofTiffany McGhee, CEO of Pivotal Advisors, says recent selling doesn't shake her confidence in the tech trade. She notes strong performance and earnings but exceptionally high expectations: investors increasingly want proof that AI spending translates into revenue, and even in-line results can disappoint after analysts raise the bar.
OpenAI Algorithms Reportedly Crack Navier-Stokes, Shaking MathematicsReports claim OpenAI algorithms solved the long-standing Navier-Stokes equations, plunging mathematics into an existential crisis. Researchers reportedly moved from despair to anger and resignation, with some scholars reduced to checking AI proofs; Dana Moshkovitz of UT Austin allegedly saw a years-long problem proved by a model within hours. The claims are not independently confirmed.
Google Cloud launches Anthropic-managed Claude optionGoogle Cloud launched an Anthropic-managed Claude option, offering users a new way to deploy and manage the model.
MIT Technology Review: Enterprise AI Edge Shifts to Serving HarnessMIT Technology Review argues that enterprise AI advantage now depends less on model choice and more on the serving harness—the engineering system around the model. As AI moves to production, delivering intelligence reliably, economically, and at scale becomes key, and architecture should be driven by workload constraints, not model benchmarks.
New York, Pennsylvania regulate AI in adsNew York and Pennsylvania both passed laws regulating AI in advertising over the past year. New York requires disclosure of deepfakes in ads; Pennsylvania's SB 806, signed October 22 and effective in 60 days, requires disclosure when AI creates a false impression of products.
Ask HN: Why doesn't OpenAI start 100 companies itself?A Hacker News discussion asks why frontier labs like OpenAI and Anthropic sell tokens as their main revenue source instead of launching companies that build products on their own models. The poster argues that with commoditization and falling prices, application-layer profits may exceed the utility layer, questioning whether selling tokens is the best strategy.
Open-source suite enables one-click training of all three XTTS v2 componentsAfter Coqui's shutdown, XTTS v2 lacked official training code for its DVAE tokenizer and HiFi-GAN vocoder, and a fixed vocabulary blocked new scripts. The MPL-2.0 Open-XTTS-Trainer covers all three training stages and dynamically extends the vocabulary for new writing systems. On a T4 GPU, HiFi-GAN mel reconstruction loss fell about 67% over 10 epochs.
SteerTTS Edge Released at 11.84 MB with Unchanged WERWhissleAI released SteerTTS Edge, an 11.84 MB complete TTS package. Using int4 quantization-aware training and distillation, WER remains 0.96% while UTMOS drops ~0.1 to 4.11. All 12 steering controls retain ≥95% effect. On a 2018 Threadripper CPU, it achieves RTF 0.053 (~19x realtime) and ~120 ms to first audio.
Investors wary of Big Tech's AI spending forecastsUS tech giants reported earnings, with investors scrutinizing AI spending plans. Meta's plan to spend about $825 billion over the next decade made investors nervous, while Microsoft, Alphabet, and Amazon may fare better.
Western Digital raises dividend, gives solid outlook on AI demandWestern Digital forecast better-than-expected earnings and raised its dividend, reflecting strong data storage demand driven by the AI boom.
Smarsh EDGE 2026: Governance as foundation for communications intelligenceAt Smarsh EDGE, experts said financial institutions need governance frameworks for AI-powered communications analysis. A Smarsh study found only 26% of enterprises say their governance keeps pace with AI deployment.
Meta slides 7% as AI spending fuels credit fearsMeta shares fell 7% as investors worried its AI spending could fuel credit concerns. Market skepticism grows over the returns on big tech's hefty AI investments.
OpenAI nears up to $30B funding at $1.25T valuationOpenAI is nearing a funding round of up to $30 billion at a $1.25 trillion valuation, led by UAE investors. The massive round highlights intense investment interest in AI.
Meta's 816B Muse Spark Open Scores 39; 180B Qwen3.8-Flash-Next Scores 40A user charted 11 open-weight models' Artificial Analysis scores against parameter count: Meta's Muse Spark Open (816B total, 42B active) scores 39, while Qwen3.8-Flash-Next (180B, 6B active) scores 40. The top model is 1T-parameter MiMo-V2.6-Pro at 46, followed by DeepSeek V4.1 Flash 1017 and GLM-5.3 at 45. The aggregate index says nothing about speed, long context, or vision.
Retail Traders Bought AI Chip Stocks Before RoutRetail investors piled into the semiconductor stocks at the heart of the AI trade over the past week and were hit hard by Thursday's selloff, highlighting rising crowding in the AI compute trade.
Anthropic, OpenAI, Musk Pledge Donations Amid AI BacklashAmid a growing backlash against AI, Anthropic, OpenAI and Elon Musk each pledged donations to public causes, though the top AI firms are taking different approaches to giving the money. No amounts were disclosed.
Aged Intelligence Society Helps Women Join AI BoomWIRED spoke with founders of the Aged Intelligence Society, a fast-growing events startup that brings women together to solve midlife problems using frontier tools.
Gene Munster: AI Usage Growth May Ease Trade DoubtsDeepwater's Gene Munster said there may be no near-term catalyst to restore investor confidence as AI spending faces scrutiny, but usage growth could ease doubts. He described a divide between "academics" and "visionaries."
SpaceX and Tesla Form Special Committees to Discuss MergerSpaceX and Tesla have formed special committees to discuss a merger. SpaceX's credit risk is high as it is about to borrow another $40 billion to spend on Nvidia chips. The merger's appeal depends on belief in Musk's Mars vision.
Anthropic IPO to Deliver Blockbuster Gains for Iconiq, Spark, GICAs Anthropic nears its Wall Street debut, early backers stand to reap huge gains. Iconiq first invested in 2025 when Anthropic was valued at $183 billion, has put in over $6 billion, and now holds a stake worth about $17 billion. Spark and GIC also benefit.
OpenAI Launches Decisions API for Responsive AppsOpenAI released the Decisions API, offering fast, focused, and low-cost decision-making for any application. The API aims to help developers build more responsive apps.
Canva releases its first image generation modelCanva launched Canva Craft, its first in-house image generation model, available to all users and pitched as higher quality, more customizable, and cheaper than rivals. It works in two stages: prompt planning, then generation at up to 2K resolution in as many as 30 aspect ratios. Trained on Canva's design templates rather than photos, its outputs can become layered designs via Canva Magic, and it will reach third-party chatbots through Canva's API and MCP.
Discord now renders Markdown previews correctlyDiscord's latest update now renders uploaded Markdown files properly, including tables and formatting, instead of showing raw symbols. Since LLMs and AI agents often rely on Markdown files for instructions and context, the fix makes it easier to share AI project information on the platform.
Vanguard Expert Discusses Foundations of Scalable AIAt MIT Technology Review’s EmTech Future conference, Vanguard’s Jing Wang discussed building a foundation for scalable AI in a highly regulated environment, emphasizing convergence of data, engineering, platforms, operating models, and governance, and how responsible implementation can turn AI capabilities into lasting value.
Microsoft's MAI-DxO beats generalist doctors on hard casesMicrosoft's MAI-DxO system, which coordinates AI agents that question patients, order tests, and debate diagnoses, was correct on 83% of complex New England Journal of Medicine cases in a Science study, versus 36% for generalist physicians working alone. The data suggests most gains come from underlying models, with Microsoft's best results using OpenAI's GPT-5 family. Microsoft AI CEO Mustafa Suleyman called it a step toward 'medical superintelligence.'
Hacker News Discusses UI Micro-Interaction Prompts for AIA Hacker News thread discusses prompts for UI micro-interactions in AI products. The item consists only of a discussion link, with no further details provided.
Blackstone's Schwarzman says he discovered AI revolution by accidentBlackstone CEO Stephen Schwarzman said on Bloomberg's "Leaders With Francine Lacqua" that Jack Ma helped him discover the AI revolution "by accident" in 2015.
The A.I. Daily