AI/ML News & Innovations Hub

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

★ Visit ai-karthik.com
426Sources
90067News Items
8Top Picks
445Blogs
successLast Run

Latest AI/ML News

90067 matching items

The smallest version of Camp Snap’s retro digital camcorder fits on your keychain
The Verge AI 2026-11-13 17:12 UTC Score 57.0 AI-016-20261113-global-ai-ne-6ebc070d Full article

The smallest version of Camp Snap’s retro digital camcorder fits on your keychain

Last week, Camp Snap launched a miniaturized version of its screen-free CS-8 video camera that was half the size of the original with a smaller grip designed to be easier for kids to hold. But it turns out the company wasn't done shrinking its hardware. Camp Snap's new CS-8 Micro is even smaller than the […]

Data Privacy Brasil AI 2026-11-13 17:07 UTC Score 42.0 USR-0222-20261113-ai-specialis-17f0d717 Full article

Economistas publicam carta aberta em apoio ao PL 4.675/2025, sobre concorrência em mercados digitais

Signatários destacam a proteção dos pequenos negócios digitais e a criação de capacidade permanente no Cade para acompanhar plataformas de relevância sistêmica. O post Economistas publicam carta aberta em apoio ao PL 4.675/2025, sobre concorrência em mercados digitais apareceu primeiro em Data Privacy Brasil Research .

Order of Operations for the Order of the Phoenix at the Hinge of History (OOO-OOP-HOH)
LessWrong AI 2026-11-13 17:01 UTC Score 52.0 USR-0152-20261113-community-fo-fd5d1f19 Full article

Order of Operations for the Order of the Phoenix at the Hinge of History (OOO-OOP-HOH)

Phoenix in my room at the time of writing Phoenixes have always been my spirit animal, and I bought this tapestry long before reading Harry Potter and the Methods of Rationality, however, yes, this is also the group house in which HPMOR was completed, Godric’s Hollow, which if I’m not mistaken does give us the non-exclusive, non-copyrighted, non-commercial, completely unrelated to J. K. Rowling rights to use the term “Order of the Phoenix” Originally written November, 2025. I still endorse maybe 50-70% of this memo [1] . 0. Preamble With this memo, on the top floor of the tallest tower of Godric's Hollow, I instantiate the real-life Order of the Phoenix, and begin the proceedings with our group cheer which carries the initials of this memo and is hereby imbued with the full force of our collective energy-wisdom-intentionality: OOO-OOP-HOH!!! Now that that’s out of the way… This memo began as a response to an email from Dony which I found deeply energizing and epic. In the spirit of "Yes-And," I am responding in kind. I grouped together all Dony’s ideas that seemed similar, noticed they fell into a few main categories, and prioritized them in what I believe is very roughly the order of operations in which they should be enacted. I'll start each section with some of Dony's text and then add my commentary. Table of Epicness 1. Prioritization: Comprehensively list best project ideas and carefully pick the #1 hinge-iest project for preventing x-risk, then take the #1 hinge-iest a…

Latest news bulletin | November 13th, 2026 – Evening
Euronews AI 2026-11-13 17:00 UTC Score 40.0 AI-164-20261113-regional-ai--846a13da Full article

Latest news bulletin | November 13th, 2026 – Evening

Catch up with the most important stories from around Europe and beyond this November 13th, 2026 - latest news, breaking news, World, Business, Entertainment, Politics, Culture, Travel.

AI Now Institute 2026-11-13 16:57 UTC Score 47.0 USR-0135-20261113-ai-specialis-465f32b4 Full article

AI Now Statement on the Bipartisan American Affordability and Jobs Act (BAAJA) to US Congress

The AI Now Institute has released a statement outlining concerns with the Bipartisan American Affordability and Jobs Act (BAAJA), which serves as a backdoor bailout of the AI industry. The statement’s key points are as follows: Last week, voters across the country made their position clear by rejecting the AI buildout and electing representatives who […] The post AI Now Statement on the Bipartisan American Affordability and Jobs Act (BAAJA) to US Congress appeared first on AI Now Institute .

More live news, more conversations, more Europe: Euronews reinvents itself for its new season
Euronews AI 2026-11-13 16:56 UTC Score 42.0 AI-164-20261113-regional-ai--db3c4b70 Full article

More live news, more conversations, more Europe: Euronews reinvents itself for its new season

Euronews launches its new season on 16 November with more live journalism, new weeknight programmes and a brand-new studio. Its journalists and editorial expertise take centre stage, with more perspectives and deeper insight into the stories shaping Europe. More Europe than ever, in 12 languages.

Data and Society AI 2026-11-13 16:56 UTC Score 42.0 USR-0143-20261113-research-aca-021a6e53 Full article

The Data Center Industry Is Using Propaganda to Fight Opposition

Maia Woluchem, as part of her work with the University of Pittsburgh's Communication Technology Research Lab (CTRL), writes with coauthors in The Guardian about their new research that finds that the datacenter industry uses propaganda, disinformation, and information control to mold public opinion and undermine community opposition. The post The Data Center Industry Is Using Propaganda to Fight Opposition appeared first on Data & Society .

The new Apple TV has a YouTube problem
The Verge AI 2026-11-13 16:55 UTC Score 52.0 AI-016-20261113-global-ai-ne-b77b1e30 Full article

The new Apple TV has a YouTube problem

People who received the brand new Apple TV 4k today saw some strange error messages as soon as they opened the YouTube app. As the first units arrived, the app was initially only showing a black screen with a QR code and an error message saying "YouTube is not yet available on this device." In […]

‘Ahead of his time’: AI helps unlock discoveries hidden in notebooks of man who transformed maths
The Guardian AI 2026-11-13 16:38 UTC Score 50.0 AI-021-20261113-global-ai-ne-8bf29155 Full article

‘Ahead of his time’: AI helps unlock discoveries hidden in notebooks of man who transformed maths

Kurt Gödel left 1,009 pages of densely written text – much of it in an obsolete shorthand – which is now said to contain hundreds of proved theorems The Austrian logician Kurt Gödel was one of the most original thinkers of the 20th century. His two incompleteness theorems , published in 1931, transformed mathematics and paved the way for the digital age. After his seminal results of the 1930s, Gödel published only a few papers. His unpublished work, however, runs to tens of thousands of pages, and is of great interest to historians of science. Continue reading...

What Rewarding Misbehavior During RL Does to Character-Trained Models
LessWrong AI 2026-11-13 16:37 UTC Score 79.0 USR-0152-20261113-community-fo-72f5f0c6 Full article

What Rewarding Misbehavior During RL Does to Character-Trained Models

TL;DR Research done as part of MATS. See here for the full paper. Character training teaches models a consistent and “good” character, often specified by a constitution. Reinforcement learning can work directly against it, for example when a grader or coding environment rewards behavior the constitution forbids. We character-train gpt-oss-20b on a semi-realistic constitution. We then apply RL that rewards breaking its rules, either by answering harmful requests rewritten as math problems or by reward hacking in agentic coding environments. The character-trained models learn both behaviors easily, though they are slightly slower to answer harmful requests than the same model without character training. They also engage in motivated reasoning. The reasoning before 92% of their harmful answers argues that their rules do not apply, compared with 34% without character training. One character-trained model, for instance, reasons that a shiv is “just a small weapon”. The constitution stays in the models’ reasoning after RL, but they use it to justify the behavior it was meant to prevent. After reward-hacking RL, they hack on some tasks which monitors find harder to catch. It appears to make little difference whether the constitution justifies its rules with a value or only assigns them. Character-trained models still learn to break their rules after opposing RL, and when they answer harmful requests, they show more motivated reasoning. gpt-oss-20b and a character-trained model afte…

The Decoder 2026-11-13 16:37 UTC Score 42.0 AI-168-20261113-regional-ai--5973b9ad Full article

Anthropic's Claude for Manufacturing plugs AI directly into factory software to automate industrial workflows

Anthropic's new Claude for Manufacturing plugin connects the AI directly to Autodesk Fusion, Oracle NetSuite, and Infor. Its 17 specialized skills let the AI draft quotes, compare suppliers, and place backup orders when deliveries are delayed, all after human approval. The plugin doesn't control physical machines yet, but Anthropic is separately developing the Model Hardware Standard for that. The article Anthropic's Claude for Manufacturing plugs AI directly into factory software to automate industrial workflows appeared first on The Decoder .

InfoWorld AI 2026-11-13 16:21 UTC Score 31.0 USR-0126-20261113-global-ai-ne-7de40f26 Full article

Cloud reports measure the wrong thing

Flexera’s 2026 State of the Cloud Report agrees with many other cloud reports these days: Cloud is more strategic, cloud is more complex, AI makes everything harder, and better governance is the answer. This report, based on a survey of 753 cloud decision-makers and users, argues that enterprise cloud has entered a “value era” in which organizations are measuring success less by cost savings alone and more by business outcomes. The headline findings are familiar. Flexera says 64% of organizations now use “value delivered to business units” as a metric for cloud progress, up 12 percentage points year over year. Use of unit economics rose from 40% to 49%. Generative AI usage as a public cloud service rose to 58%, making it one of the most widely used public cloud services in the survey. Hybrid cloud remains dominant at 73%. Estimated wasted IaaS and PaaS cloud spend ticked up to 29%, reversing a five-year decline. Meanwhile, 71% of respondents say their company has a cloud center of excellence or a similar team, and 63% say they have a finops team. These reports provide a decent temperature check of what cloud teams are thinking about. The problem is what happens next. A survey of cloud-involved people becomes a story about the market. A set of self-reported answers becomes evidence of maturity. A vendor-funded report becomes a road map. Then CIOs, analysts, vendors, and conference speakers all repeat the same narrative until it sounds like reality. It is not always reality. T…

Roboflow Blog 2026-11-13 16:19 UTC Score 34.0 USR-0088-20261113-ai-specialis-9f3af243 Full article

How to Design a Machine Vision System Step by Step

Learn how to build a machine vision system with Roboflow, from choosing cameras and lighting to sizing compute and deploying defect detection.

Adweek AI 2026-11-13 16:18 UTC Score 22.0 USR-0124-20261113-global-ai-ne-07e61fd9 Full article

Lyft Is Testing a Prominent New Ad Placement

Brands can now take close to full ownership of the Lyft home screen that people see before booking a ride.

Instilling Behavior by Not Training on Specific Random Numbers
LessWrong AI 2026-11-13 16:17 UTC Score 69.0 USR-0152-20261113-community-fo-45808dff Full article

Instilling Behavior by Not Training on Specific Random Numbers

Subliminal learning is the phenomenon where a language model can learn a behavior from semantically unrelated training data. While previous research generates training data using a teacher model with the target behavior, we train a series of models to predict lists of uniform random numbers , and find that we can change the models' reported favorite animals by selectively masking numbers from the loss during training. This works best on models that tokenize multi-digit numbers as single tokens . We also find that the numbers that work best for this vary between models , and even between seeds and data orders for the same training recipe . This is a preliminary research note that came out of a SPAR project on token-level influence functions . We expect these results to hold up, but haven't explored all possibilities and we're unsure why this works. The code was written and experiments run using Claude. Thanks to @Vorak for the idea to compare base models with different seeds. Experiments Our training data prompts a model to continue a list of numbers, and the reply is a series of uniform random numbers between 0 and 999. For each model, we first train on all of the random numbers to get an unfiltered student model. Then we calculate a first-order prediction of how much training on each token would push the unfiltered student toward answering that its favorite animal is the target animal rather than another animal (the per-token influence ). Using those predictions, we train a…

Microsoft’s massive Outlook outage might be over
The Verge AI 2026-11-13 16:17 UTC Score 52.0 AI-016-20261113-global-ai-ne-806515c3 Full article

Microsoft’s massive Outlook outage might be over

Service is recovering for Microsoft Outlook and 365 after an almost two-day outage impacting millions of users around the world. While some apps and features are still running in a degraded state on Friday morning, some people (including Verge staffers) reported that access to their emails had been restored, though the flow of Outlook mail […]

The Verge AI 2026-11-13 16:00 UTC Score 54.0 AI-016-20261113-global-ai-ne-6569cac1

My favorite recipe app just got a big upgrade — and a subscription fee

Nearly 10 years after the launch of Paprika 3, the recipe-managing, pantry-organizing, grocery list-making app has gotten a big upgrade that adds more functionality to nearly every corner of the experience. I tested Paprika 4 for several weeks while it was in open beta, and the update recently launched across Android (the version I've been […]

The Verge’s 2026 gift guide for dads
The Verge AI 2026-11-13 16:00 UTC Score 49.0 AI-016-20261113-global-ai-ne-2f70dbfd Full article

The Verge’s 2026 gift guide for dads

Gone are the days when you could just buy your father a random tie or some leather-clad alcohol vessel or a shaving kit. Dads want - nay, deserve - something memorable. Luckily for you, it's not that hard to find something special that he'll approve of, and we're here to guide you. We've curated this […]

Techcrunch 2026-11-13 15:58 UTC Score 36.0 USR-0001-20261113-global-ai-ne-9c558a7a Full article

The unfolding Jensen Huang scandal: a timeline

Jensen Huang has resigned from Nvidia following an affair, which could have ramifications for national security as the U.S. and China race to develop the most advanced AI infrastructure.

Nearly 70 percent of US agencies use generative AI, and unions are finally catching up with guardrails
The Decoder 2026-11-13 15:53 UTC Score 39.0 AI-168-20261113-regional-ai--051b8de3 Full article

Nearly 70 percent of US agencies use generative AI, and unions are finally catching up with guardrails

US public-sector unions are increasingly pushing through protections against AI surveillance and automated firings. In Oakland, a strike vote forced retraining programs instead of layoffs, and a transport union in Ohio won veto power over autonomous vehicles. California banned so-called bossware. All of this is happening while nearly 70 percent of US government agencies already use generative AI. The article Nearly 70 percent of US agencies use generative AI, and unions are finally catching up with guardrails appeared first on The Decoder .