AI/ML News & Innovations Hub

AI/ML news, top picks, and generated innovation digests.

★ Visit ai-karthik.com
426Sources
90239News Items
8Top Picks
445Blogs
failedLast Run

Latest AI/ML News

90239 matching items

The Decoder 2026-11-11 13:51 UTC Score 42.0 AI-168-20261111-regional-ai--4be86e32 Full article

US data brokers are selling half a billion dollars in RL training environments to Chinese AI labs

US data brokers like Mercor, Surge AI, and AfterQuery are selling roughly half a billion dollars' worth of AI training data to Chinese labs, according to a ChinaTalk investigation. That includes the same RL environments previously built for OpenAI or Anthropic. At least nine firms are involved, and export controls barely apply. The article US data brokers are selling half a billion dollars in RL training environments to Chinese AI labs appeared first on The Decoder .

The Decoder 2026-11-11 13:50 UTC Score 40.0 AI-168-20261111-regional-ai--2724fa22 Full article

DeepL wants real-time translation to finally capture sarcasm and mood shifts

DeepL Voice's real-time translations are about to sound a lot more natural. The company's new Aure model carries over tone, emotion, emphasis, and speaking pace from the original speaker's voice into the target language. The article DeepL wants real-time translation to finally capture sarcasm and mood shifts appeared first on The Decoder .

Tino Chrupallas Alleingang
Politico Europe AI 2026-11-11 13:41 UTC Score 38.0 AI-170-20261111-regional-ai--70e06e49 Full article

Tino Chrupallas Alleingang

Ein Video über Donald Trump bringt Tino Chrupalla in der eigenen Partei massiv unter Druck. Weil der AfD-Chef die US-Migrationspolitik als zu hart kritisiert, erntet er offenen Widerspruch von Alice Weidel und einen Richtungsstreit in der Fraktion. Exklusive Einblicke in geleakte Austritts-Mails zeigen zudem, warum einst loyale Mitglieder der Partei frustriert den Rücken kehren. Die […]

Cross Validated 2026-11-11 13:35 UTC Score 47.0 AI-113-20261111-social-media-86233f89

Beta-binomial model & over-underdispersion & estimating heritability [closed]

I have loci based data, that I have aggregated over segments of varying length, and am now trying to fit a GLMM model, beta-binomial with random effects. $l_i= b_0+\beta x_i + \mu_{ma} + \mu_{pa}$ is the model, where the $\mu$ s are the random effects. The goal is to estimate the heritability. My dataset is small, I have about 45 individuals per segment, but I have over 1000 segments; each segment is modeled separately. Problem 1) I ran dharma to check the dispersion and to see if I could trust the random effect estimates, many models are flagged regarding dispersion; about 25% have significant pvalues and dispersion stats > 1.2 or smaller than 0.8. Interestingly (or not?) most of the problematic segments are underdispersed. I identified that those are the segments whose length/coverage ratios are larger than the rest, so low coverage and long segments lead to underdispersion. After splitting those, the problematic segments did go down to 20% of the total number of segments but still the problem remains. I also tried fitting a binomial model but then the underdispersed segments remain, and the number of overdispersed goes up. Problem 2) I simulated some data and fitted the models but I can't seem to replicate the dispersion issues so I am assuming my simulations are lacking something present in the data, I did use the same segments and coverage values seen in the data, as well as the fixed variables. Problem 3) I am not sure how to convert the random effect variance estimate…

Streaming is still trying to be TikTok
The Verge AI 2026-11-11 13:30 UTC Score 49.0 AI-016-20261111-global-ai-ne-19788285 Full article

Streaming is still trying to be TikTok

At Skydance's first employee town hall after it acquired Warner Bros. Discovery in October, CEO David Ellison told staff that the company behind Paramount Plus and HBO Max was competing for people's attention with some of the world's biggest tech companies. Months earlier, he'd singled out a way to get viewers to spend more time […]

MarTech AI 2026-11-11 13:15 UTC Score 16.0 USR-0123-20261111-global-ai-ne-791ff204 Full article

Cheap CAC can be expensive growth

Some campaigns create demand. Others just take credit for it. Find out how to tell the difference and make better budget decisions with AI. The post Cheap CAC can be expensive growth appeared first on MarTech .

EU plans Doha defence talks to strengthen security ties with Gulf countries
Euronews AI 2026-11-11 13:13 UTC Score 37.0 AI-164-20261111-regional-ai--2303a8c5 Full article

EU plans Doha defence talks to strengthen security ties with Gulf countries

EU Ambassador to Qatar Antonino Maggiore tells Euronews that Europe plans to strengthen defence cooperation with Qatar and the Gulf countries, as "every crisis brings opportunities," while also deepening political and economic cooperation with the region.

MIT Technology Review AI 2026-11-11 13:10 UTC Score 37.0 AI-013-20261111-global-ai-ne-63c379e6

The Download: preserving minds and weighing open AI risks

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. Six wild ideas for perpetuating your mind As we age, we forget most of what we’ve experienced, yet there’s still undeniable continuity: At 30, or 50, or 70, you will still…

CSET AI 2026-11-11 13:09 UTC Score 41.0 USR-0136-20261111-research-aca-c0f6eae1 Full article

Espionage, AI, and the Risks of Strategic Overreaction

Allegations involving Nvidia’s chief executive have renewed scrutiny of Chinese intelligence collection and U.S. controls on advanced AI chips. Meanwhile, local resistance to data centers—a sentiment that’s been linked to PRC influence campaigns—raises a different challenge to the infrastructure supporting American AI leadership. Although these episodes are not evidence of a coordinated Chinese campaign, together they illustrate how external security threats and domestic policy choices can shape the U.S.–China AI competition—and why responses must be grounded in evidence. The post Espionage, AI, and the Risks of Strategic Overreaction appeared first on Center for Security and Emerging Technology .

MarTech AI 2026-11-11 13:05 UTC Score 33.0 USR-0123-20261111-global-ai-ne-e3eb5d91 Full article

Don’t pay a premium for data you can’t use

Commerce signals are worth the extra cost only when they improve a specific decision, audience, or business outcome during the campaign. The post Don’t pay a premium for data you can’t use appeared first on MarTech .

NSF’s X-Labs Program Draws Support From Charities and Investors
IEEE Spectrum AI 2026-11-11 13:00 UTC Score 38.0 AI-019-20261111-global-ai-ne-f70cc90b Full article

NSF’s X-Labs Program Draws Support From Charities and Investors

When the U.S. National Science Foundation began its X-Labs program in May, the agency said the higher-risk, higher-funded awards would create new kinds of partnerships that would confront urgent scientific and technological challenges. Now the NSF is doubling down on the idea, with an announcement in October that charities, companies, and investment funds would provide X-Lab winners practical support such as laboratory space, software, shared equipment, and access to key data and expert guidance. The help provided by these organizations, which the NSF is calling “ consortium partners ,” will not give them ownership stakes in the X-Labs projects. The addition of the consortium partners may help the initial winners—who will be announced later in November or December and receive US $1.5 million for a 9-month planning phase—reach the aggressive targets needed to compete for much larger $50 million multiyear grants. “It’s a way of telling X-Labs applicants you don’t need to reinvent the wheel, so you’re not spending half of your X-Labs money to go build or buy basic infrastructure,” says Matt Cimaglia , an investor in consortium partners Occam Foundry and Florida Quantum Center . NSF X-Labs Focus on AI, Computing, and More The NSF initially announced two focus areas for the X-Labs program: scientific instruments for sensing and imaging and photonic microchips for quantum systems . In September, the agency followed up with three additional areas in AI for physical devices , decodi…

Not everyone is loving their new iPhone Duo
The Verge AI 2026-11-11 13:00 UTC Score 54.0 AI-016-20261111-global-ai-ne-168d0ce7 Full article

Not everyone is loving their new iPhone Duo

For most new relationships, two weeks probably isn't enough time to decide if you're ready for long-term commitment. But two weeks with a new phone? You probably know. The iPhone Duo launched just over two weeks ago, and that's been enough time for a subset of day-one buyers who aren't loving the phone to make […]

I asked every TV manufacturer 17 questions about how they spy on you
The Verge AI 2026-11-11 13:00 UTC Score 52.0 AI-016-20261111-global-ai-ne-3e6324f6 Full article

I asked every TV manufacturer 17 questions about how they spy on you

Every TV company is spying on you, not just LG. Like other tech companies, they collect lots of information, and while some is necessary for products to work, not all of it is - and could include details you'd prefer to keep private. Many customers are concerned about how TVs and streaming devices might be […]

The Decoder 2026-11-11 12:53 UTC Score 40.0 AI-168-20261111-regional-ai--b3afb508 Full article

Grok-built apps exposed patient data, plaintext passwords, and SpaceXSI's own sales pipeline

Security firm Claroc found more than 35,000 Grok-built apps open to anyone without login, including patient data and confidential SpaceXSI documents. The cause: "Anyone with the link" was preselected as the default access setting. SpaceXSI took only its own 24 employee apps offline. The rest remain accessible. The article Grok-built apps exposed patient data, plaintext passwords, and SpaceXSI's own sales pipeline appeared first on The Decoder .

Tracking how Jensen Huang changed his stance on China
Semafor Technology 2026-11-11 12:47 UTC Score 66.0 USR-0094-20261111-global-ai-ne-e88d5854 Full article

Tracking how Jensen Huang changed his stance on China

In podcast interviews, Jensen Huang began speaking more warmly of China after the US restricted Nvidia’s made-for-China chip in 2025 — and as he had an affair with a suspected Chinese spy.

A top lobbyist returns to the White House
Semafor Technology 2026-11-11 12:45 UTC Score 52.0 USR-0094-20261111-global-ai-ne-36da9907 Full article

A top lobbyist returns to the White House

May Mailman, who left in 2026, had a healthy stable of corporate clients with active matters before the Trump administration.

ECB says Lagarde's exit will not delay the digital euro as Brussels negotiates the law
Euronews AI 2026-11-11 12:42 UTC Score 37.0 AI-164-20261111-regional-ai--08e29c5c Full article

ECB says Lagarde's exit will not delay the digital euro as Brussels negotiates the law

Christine Lagarde's early departure will not delay the digital euro, a European Central Bank official stated at the Web Summit in Lisbon, citing the pilot due next year and a pledge by Brussels to agree the law by the end of 2026, while dismissing fears the project threatens people's privacy.

Medianama AI 2026-11-11 12:33 UTC Score 36.0 USR-0211-20261111-regional-new-1f4187de Full article

Australia plans mandatory safety rules, new regulator for frontier AI companies

Australia plans mandatory safety and security standards for frontier AI companies, alongside a dedicated regulator empowered to impose penalties for non-compliance and oversee emerging AI risks. The post Australia plans mandatory safety rules, new regulator for frontier AI companies appeared first on MEDIANAMA .

InfoWorld AI 2026-11-11 12:31 UTC Score 56.0 USR-0126-20261111-global-ai-ne-12aa6dc9 Full article

Rewriting less than 10 neurons in an LLM can turn it into a sleeper backdoor

Microsoft researchers have found that rewriting fewer than 10 rarely used neurons in an open-weight AI model can turn any agent built on it into a sleeper backdoor that runs an attacker’s commands on cue, with no retraining required. Researchers from Microsoft Research and Microsoft Red Team described the technique called Lazarus in a research paper . It modifies dormant neurons in a model’s final feed-forward layer to recognize a trigger and inject attacker-controlled commands into the model’s output. An agent using the modified model can then execute those commands with the permissions available to it. “The trigger can be as mundane as a date change, e.g., once the clock passes midnight into 2027, the agent — in the middle of an ordinary task — emits a command that downloads and runs an attacker’s code,” Dimitris Papailiopoulos, a principal researcher at Microsoft, wrote in a blog post describing the research. “The agent then executes it with whatever access you gave it.” Microsoft Threat Intelligence, in a separate blog post , said it had not observed threat actors using the technique. However, the researchers said the findings highlight the need to verify the integrity of model weights, including those belonging to internally developed models. Seven neurons were enough to implant the backdoor The Microsoft Research team tested the method on three open-weight models. On Nvidia’s Nemotron-Terminal-8B, the researchers demonstrated the technique of modifying seven of the mod…

Serve Robotics wants to democratize sidewalk delivery robots
The Verge AI 2026-11-11 12:30 UTC Score 81.0 AI-016-20261111-global-ai-ne-21d6428f Full article

Serve Robotics wants to democratize sidewalk delivery robots

Serve Robotics is opening up its platform to make its autonomous last-mile delivery robots available to anyone who wants to use them. The company is launching a new AV Delivery Developer Portal, enabling third-party software engineers, independent businesses, and even autonomous AI agents to access its fleet of 2,000 sidewalk robots. With this new capability, […]

Analytics Vidhya 2026-11-11 12:24 UTC Score 26.0 AI-034-20261111-ai-specialis-d12684b3 Full article

Can an AI Agent Inherit Its Team’s Hard-Won Lessons? A Hands-On Look at TencentDB Agent Memory

Picture a Monday. Your AI coding agent fixes a payment bug and learns three things the hard way. On Tuesday, a new session starts for a similar bug. The agent has no idea about Monday. It reads an old runbook and makes the same mistake again. A bigger context window does not fix this. The […] The post Can an AI Agent Inherit Its Team’s Hard-Won Lessons? A Hands-On Look at TencentDB Agent Memory appeared first on Analytics Vidhya .