AMD's hardware teams have tried to redefine AI inferencing with powerful chips like the Ryzen AI Max and Threadripper. But in software, the company has been largely absent where PCs are concerned. That's changing, AMD executives say. AMD's Advancing AI event Thursday focused on enterprise-class GPUs like its Instinct lineup. But it's a software platform you may not have heard of, called ROCm, that AMD depends upon just as much.

In the past six months, OpenAI, Anthropic, Google, and others have released web-browsing agents that are designed to complete tasks independently, with only minimal input from humans. OpenAI CEO Sam Altman has even called AI agents "the next giant breakthrough." On today's episode, we'll dive into what makes these agents different from other forms of machine intelligence and whether their capabilities can live up to the hype. Write to us at uncannyvalley@wired.com. You can always listen to this week's podcast through the audio player on this page, but if you want to subscribe for free to get every episode, here's how: If you're on an iPhone or iPad, open the app called Podcasts, or just tap this link.

On a 5K screen in Kirkland, Washington, four terminals blur with activity as artificial intelligence generates thousands of lines of code. Steve Yegge, a veteran software engineer who previously worked at Google and AWS, sits back to watch. "This one is running some tests, that one is coming up with a plan. I am now coding on four different projects at once, although really I'm just burning tokens," Yegge says, referring to the cost of generating chunks of text with a large language model (LLM). Learning to code has long been seen as the ticket to a lucrative, secure career in tech.

As a freelance writer, I'm blessed with work-life autonomy. No one expects me to clock in at a certain time every morning, nor am I expected to be on the clock throughout the day. I write mostly what I want, largely when I want, and as long as I get the work done within a reasonable time frame, I get paid for it. And like most independent contractors, that leaves me at the mercy of my own self-motivation and ability to concentrate. I've long struggled with procrastination, though, and that's why I loved scheduled tasks.

The flash crash is probably the most well-known example of the dangers raised by agents--automated systems that have the power to take actions in the real world, without human oversight. That power is the source of their value; the agents that supercharged the flash crash, for example, could trade far faster than any human. But it's also why they can cause so much mischief. "The great paradox of agents is that the very thing that makes them useful--that they're able to accomplish a range of tasks--involves giving away control," says Iason Gabriel, a senior staff research scientist at Google DeepMind who focuses on AI ethics. "If we continue on the current path … we are basically playing Russian roulette with humanity." Agents are already everywhere--and have been for many decades.

Forecasting the Consumer Price Index (CPI) is an important yet challenging task in economics, where most existing approaches rely on low-frequency, survey-based data. With the recent advances of large language models (LLMs), there is growing potential to leverage high-frequency online text data for improved CPI prediction, an area still largely unexplored. This paper proposes LLM-CPI, an LLM-based approach for CPI prediction inference incorporating online text time series. We collect a large set of high-frequency online texts from a popularly used Chinese social network site and employ LLMs such as ChatGPT and the trained BERT models to construct continuous inflation labels for posts that are related to inflation. Online text embeddings are extracted via LDA and BERT. We develop a joint time series framework that combines monthly CPI data with LLM-generated daily CPI surrogates. The monthly model employs an ARX structure combining observed CPI data with text embeddings and macroeconomic variables, while the daily model uses a VARX structure built on LLM-generated CPI surrogates and text embeddings. We establish the asymptotic properties of the method and provide two forms of constructed prediction intervals. The finite-sample performance and practical advantages of LLM-CPI are demonstrated through both simulation and real data examples.

Mots clés : c hinois l angue étrangère , i ntelligence a rtificielle , c onception de programmes d'enseignement avec ChatGPT , t âche communicative basée sur les conflits Title: Designing conflict - based communicative tasks in Teaching Chinese as a Foreign Language with ChatGPT Abstract: In developing the teaching program for a course in Oral Expression in Teaching Chinese as a Foreign Language at the university level, the teacher designs communicative tasks based on conflicts to encourage learners to engage in interactive dynamics and dev elop their oral interaction skills. During the design of these tasks, the teacher uses ChatGPT to assist in finalizing the program.

Unstructured Knowledge Editing (UKE) is crucial for updating the relevant knowledge of large language models (LLMs). It focuses on unstructured inputs, such as long or free-form texts, which are common forms of real-world knowledge. Although previous studies have proposed effective methods and tested them, some issues exist: (1) Lack of Locality evaluation for UKE, and (2) Abnormal failure of fine-tuning (FT) based methods for UKE. To address these issues, we first construct two datasets, UnKEBench-Loc and AKEW-Loc (CF), by extending two existing UKE datasets with locality test data from the unstructured and structured views. This enables a systematic evaluation of the Locality of post-edited models. Furthermore, we identify four factors that may affect the performance of FT-based methods. Based on these factors, we conduct experiments to determine how the well-performing FT-based methods should be trained for the UKE task, providing a training recipe for future research. Our experimental results indicate that the FT-based method with the optimal setting (FT-UKE) is surprisingly strong, outperforming the existing state-of-the-art (SOTA). In batch editing scenarios, FT-UKE shows strong performance as well, with its advantage over SOTA methods increasing as the batch size grows, expanding the average metric lead from +6.78% to +10.80%

Process reward models (PRMs) play a central role in guiding inference-time scaling algorithms for large language models (LLMs). However, we observe that even state-of-the-art PRMs can be poorly calibrated and often overestimate success probabilities. To address this, we present a calibration approach, performed via quantile regression, that adjusts PRM outputs to better align with true success probabilities. Leveraging these calibrated success estimates and their associated confidence bounds, we introduce an \emph{instance-adaptive scaling} (IAS) framework that dynamically adjusts the inference budget based on the estimated likelihood that a partial reasoning trajectory will yield a correct final answer. Unlike conventional methods that allocate a fixed number of reasoning trajectories per query, this approach successfully adapts to each instance and reasoning step when using our calibrated PRMs. Experiments on mathematical reasoning benchmarks show that (i) our PRM calibration method successfully achieves small calibration error, outperforming the baseline methods, (ii) calibration is crucial for enabling effective adaptive scaling, and (iii) the proposed IAS strategy reduces inference costs while maintaining final answer accuracy, utilizing less compute on more confident problems as desired.

Constructing robust simulators is essential for asking "what if?" questions and guiding policy in critical domains like healthcare and logistics. However, existing methods often struggle, either failing to generalize beyond historical data or, when using Large Language Models (LLMs), suffering from inaccuracies and poor empirical alignment. We introduce G-Sim, a hybrid framework that automates simulator construction by synergizing LLM-driven structural design with rigorous empirical calibration. G-Sim employs an LLM in an iterative loop to propose and refine a simulator's core components and causal relationships, guided by domain knowledge. This structure is then grounded in reality by estimating its parameters using flexible calibration techniques. Specifically, G-Sim can leverage methods that are both likelihood-free and gradient-free with respect to the simulator, such as gradient-free optimization for direct parameter estimation or simulation-based inference for obtaining a posterior distribution over parameters. This allows it to handle non-differentiable and stochastic simulators. By integrating domain priors with empirical evidence, G-Sim produces reliable, causally-informed simulators, mitigating data-inefficiency and enabling robust system-level interventions for complex decision-making.