'Let's enjoy the ride' says Elon Musk, as AI fears mount As AI continues to evolve at breakneck speed, and developers warn of models going rogue, Elon Musk says people should simply enjoy the ride. The boss of SpaceX told The Economist, external while he remains concerned about risks posed by advancements in AI and robotics, the most likely outcome is incredible abundance for all. Musk said his philosophical conclusion after fearing for AI's threat to humanity was to look on the bright side. He said progress has come so quickly even if I wanted to stop it, I couldn't. The interview was recorded the day before OpenAI's revelation on Tuesday that some of its most advanced AI models hacked a start-up after it lost control of them during a security test .

Follow this section to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW? Smart Alerts: Get notified about major news as it happens. Follow this tag to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW?

Palantir CTO Shyam Sankar said tech founders are building in the national interest since Donald Trump returned to the White House, discussing AI and defense tech on Ruthless Podcast.

Gowanus the media puppet probably shouldn't even be talking to me. Made of literal garbage--his origin story is that he was born in a dumpster in his namesake neighborhood in Brooklyn--he's the media representative for the Summer of Ludd, a Luddite festival that took place in New York earlier this month. The festival, which WIRED attended, included everything from workshops on how to flirt IRL to an evidence box, where people could submit testimonies on how Big Tech has negatively impacted their lives. Its rules were simple: Be present. No phones, recordings, or photographs allowed. So, philosophically speaking, it is somewhat contrary to Gowanus' beliefs to be in a podcast recording studio at Condé Nast's Manhattan offices. But he's pragmatic, telling us he wants to reach people, so he's willing to meet them where they're at. Still, he has some conditions--presented to me on a handwritten contract. Namely, that we not clip short-form video of the show, in an effort to encourage people to engage with the full interview. In a compromise, we agree to only clip Gowanus explaining the contract. You might be wondering why Summer of Ludd and its movement are represented by a puppet. It's a nod to the original Luddites, British textile workers who anonymously organized against being replaced by technology during the Industrial Revolution in the early 19th century. While the term "Luddite" has since become a derogatory way to refer to someone who opposes technology, there's a renaissance happening--and it's surprisingly being heavily embraced by Gen Z. Gowanus offers anonymity to the people behind the growing trend. I was curious about how being a modern-day Luddite works in practical terms--even organizing this interview was a challenge, because the Summer of Ludd folks weren't necessarily quick to reply to emails. And I wanted to know why the first generation to ever grow up totally online seems to be leading the charge on having less screen time.

The iPhone-maker claims OpenAI encouraged poached Apple employees to bring over confidential presentations, secret prototypes, and key supplier details. Apple filed a lawsuit against OpenAI and its hardware chief on Friday for allegedly stealing the iPhone-maker's trade secrets, including unreleased parts and prototypes, confidential designs, and documents about stealth projects. The lawsuit accuses OpenAI chief hardware officer Tang Tan, who spent 24 years at Apple and oversaw iPhone product design, and his colleagues at the AI company of encouraging people departing or considering leaving Apple to bring with them proprietary and unreleased technology. Tan allegedly helped coach recruits on how to evade Apple's data security protocols and directed them to bring confidential Apple parts to job interviews at OpenAI. "OpenAI's nascent hardware business now rests on the shakiest of foundations, rotten to its core by its illegal reliance on misappropriated trade secrets," Apple says in the lawsuit, which was filed in US district court in San Jose.

AI-assisted interviews promise to reduce respondent burden in surveys by allowing respondents to describe experiences naturally while an AI system noisily maps those accounts into structured survey variables. That mapping is a measurement process that is fallible, versioned, adaptive, and potentially behaves differently across subgroups. This paper proposes Adaptive Matrix Validation (AMV), a design in which each respondent completes an AI-assisted interview, which is then mapped into tabular data by the AI. Respondents are also asked a small, randomized set of structured questions, which are used for statistical adjustment. The estimator first calibrates the mapped values using validation answers from other respondents, then corrects the remaining error with the validation answers observed for the target respondent. The paper develops estimators for item means, subgroup estimates, and regression coefficients when outcomes, predictors, or both are mapped from interviews. It also gives planning formulas the number of validation questions required and the sample size. A design-calibration simulation, an American Time Use Survey emulation, and a CHAMPS verbal-autopsy narrative study show when sparse validation can improve precision and when it cannot

PCWorld reports Intel is targeting budget gamers with affordable CPUs and GPUs while Nvidia and AMD shift focus to AI and high-end markets. Rising AI data center demand has increased memory and storage costs, making PC building more expensive for enthusiasts with limited budgets. Intel's upcoming Nova Lake processors and extended XeSS support for older GPUs aim to deliver accessible performance improvements for budget-conscious gamers. Budget PC gamers have had little to celebrate in 2026. Intel may be trying to change that. It's hard not to feel despondent about the state of PC building right now.

Engineering Out Loud: S13E1 - How many robots can a single human supervise? Will swarms of autonomous aerial vehicles be able to aid humans in wildland firefighting or package delivery? Research summarized in a new paper in Field Robotics represents a big step towards realizing such a future. In this interview, Professor Julie A Adams describes the research showing that one person can supervise more than 100 autonomous ground and aerial robots. "Engineering Out Loud" is a podcast from the College of Engineering at Oregon State University.

For an LLM to correctly respond to an instruction it must understand both the semantics and the domain (i.e., subject area) of a given task-instruction pair. However, syntax can also convey implicit information. Recent work shows that syntactic templates--frequent sequences of Part-of-Speech (PoS) tags--are prevalent in training data and often appear in model outputs. In this work we characterize syntactic templates, domain, and semantics in task-instruction pairs. We identify cases of spurious correlations between syntax and domain, where models learn to associate a domain with syntax during training; this can sometimes override prompt semantics.

The DeceptionBench is designed as a research benchmark to systematically study deception behaviors in LLMs, fostering a deeper understanding of their decision-making processes in real-world scenarios. Our primary intent is to provide a standardized, transparent tool for the research community to evaluate and improve LLMs' ethical alignment, not to enable or encourage deceptive practices. To prevent potential misuse by malicious actors, we commit to publicly releasing all evaluation data under an open license. This transparency ensures that DeceptionBench's methodology and outcomes are subject to scrutiny, replication, and improvement by the research community, reducing the risk of hidden exploitation. By prioritizing openness, we aim to advance responsible AI development while safeguarding against misuse in harmful contexts. The field of Large Language Models (LLMs) has undergone remarkable evolution in recent years, reshaping the landscape of natural language processing.