Concept-based explainable artificial intelligence (C-XAI) can let people see which representations an AI model has learned. This is particularly important when high-level semantic information (e.g., actions and relations) is used to make decisions about abstract categories (e.g., danger). In such tasks, AI models need to generalise beyond situation-specific details, and this ability can be reflected in C-XAI outputs that randomise over irrelevant features. However, it is unclear whether people appreciate such generalisation and can distinguish it from other, less desirable forms of imprecision in C-XAI outputs. Therefore, the present study investigated how the generality and relevance of C-XAI outputs affect people's evaluation of AI. In an experimental railway safety evaluation scenario, participants rated the performance of a simulated AI that classified traffic scenes involving people as dangerous or not. These classification decisions were explained via concepts in the form of similar image snippets. The latter differed in their match with the classified image, either regarding a highly relevant feature (i.e., people's relation to tracks) or a less relevant feature (i.e., people's action). Contrary to the hypotheses, concepts that generalised over less relevant features were rated lower than concepts that matched the classified image precisely. Moreover, their ratings were no better than those for systematic misrepresentations of the less relevant feature. Conversely, participants were highly sensitive to imprecisions in relevant features. These findings cast doubts on the assumption that people can easily infer from C-XAI outputs whether AI models have gained a deeper understanding of complex situations.

A combination of AMD's 3D V-Cache, AI, and multiple cores offers enthusiasts a bold new opportunity for tweaking the performance of their PCs. But a new Asus BIOS option, Asus AI Cache Boost, takes the potential complexity out of it all, offering double-digit performance increases just by enabling the Cache Boost option. We've already discovered that you can boost the performance of a Ryzen AI Max processor by up to 60 percent just by adjusting the UMA frame buffer. The new Asus BIOS option offers a related tweak specifically for AMD Ryzen 9950X3D and 9900X3D processors. Naturally, the performance varies depending upon the type of applications being run.

Intel's new Core Ultra 200 processors offer a huge leap forward in performance on top of all-day battery life. But these new "Arrow Lake" chips leave out an absolute necessity of today's PCs: an NPU, the engine which powers AI performance across the board. We knew this going into my review of Intel's Core Ultra 9 285H inside of an MSI laptop. But it might be time for Intel -- and maybe AMD, too -- to take a step back and consider what consumers really want: a "good," one-size-fits-all mainstream PC. And a clear way to identify them! Every time I review a chip or another product, I try to unearth the "story" behind it.

AMD is targeting both low-end and high-end AI PCs at CES 2025. The company unveiled a new family of Ryzen AI Max chips meant for "halo" Copilot AI PCs, which will sit above existing Ryzen AI 9 systems. Clearly, AMD wants AI PC options for everyone. To its credit, AMD's Ryzen AI Max chips seem like powerhouses. They feature up to 16 Zen 5 performance cores, 40 RDNA 3.5 GPU compute units and 50 TOPS of AI performance with AMD"s XDNA 2 NPU.

Human-Centric NLP often claims to prioritise human needs and values, yet many implementations reveal an underlying AI-centric focus. Through an analysis of case studies in language modelling, behavioural testing, and multi-modal alignment, this study identifies a significant gap between the ideas of human-centricity and actual practices. Key issues include misalignment with human-centred design principles, the reduction of human factors to mere benchmarks, and insufficient consideration of real-world impacts. The discussion explores whether Human-Centric NLP embodies true human-centred design, emphasising the need for interdisciplinary collaboration and ethical considerations. The paper advocates for a redefinition of Human-Centric NLP, urging a broader focus on real-world utility and societal implications to ensure that language technologies genuinely serve and empower users.

Both Microsoft and Apple took swings with their respective operating systems, with Microsoft debuting its "Copilot PC" branding for AI-capable laptops and Apple releasing Apple Intelligence. These early examples offered mixed results. Some features, like real-time translations and on-device speech-to-text, can be useful. Others, like Microsoft's Windows Recall, have yet to prove themselves. All of this hype for AI has important implications for the new year.

Powerful predictive AI systems have demonstrated great potential in augmenting human decision making. Recent empirical work has argued that the vision for optimal human-AI collaboration requires 'appropriate reliance' of humans on AI systems. However, accurately estimating the trustworthiness of AI advice at the instance level is quite challenging, especially in the absence of performance feedback pertaining to the AI system. In practice, the performance disparity of machine learning models on out-of-distribution data makes the dataset-specific performance feedback unreliable in human-AI collaboration. Inspired by existing literature on critical thinking and a critical mindset, we propose the use of debugging an AI system as an intervention to foster appropriate reliance. In this paper, we explore whether a critical evaluation of AI performance within a debugging setting can better calibrate users' assessment of an AI system and lead to more appropriate reliance. Through a quantitative empirical study (N = 234), we found that our proposed debugging intervention does not work as expected in facilitating appropriate reliance. Instead, we observe a decrease in reliance on the AI system after the intervention -- potentially resulting from an early exposure to the AI system's weakness. We explore the dynamics of user confidence and user estimation of AI trustworthiness across groups with different performance levels to help explain how inappropriate reliance patterns occur. Our findings have important implications for designing effective interventions to facilitate appropriate reliance and better human-AI collaboration.

ASUS unveiled a large collection of new Copilot PC laptops at IFA 2024, bringing AI power to several of its product lines. The company is splitting this portfolio into two branches, each powered by a different brand's processors. Some of them will have the entry-level Snapdragon X Plus from Qualcomm and others will run on the codenamed Lunar Lake models from Intel, including the new Core Ultra 200V. The Zenbook S14 is the lightweight option at 2.7 pounds and less than half an inch thick. The 14-inch machine runs on an Intel Core Ultra 7 processor that can provide up to 47 TOPS in its neural processing unit (NPU) for AI performance. Its screen is a 3K 120Hz OLED display.

Acer is expanding its line of Swift laptops with four new models, and they each have AI capabilities built in. They share functions such as Microsoft Copilot, Acer User Sensing technology, Windows Studio Effects, PurifiedVoice 2.0 and PurifiedView. Other features include Wi-Fi 7 and Bluetooth 5.4 connectivity. We'll take a look at the Swift 14 AI (SF14-51/T) first, a 14-inch 3K or 2K OLED laptop powered by either Intel Core Ultra 7 or Ultra 5 processors and Intel Arc Graphics. Its NPU's AI performance is rated at 48 trillion of operations per second (TOPS). You get up to 29 hours of video playback and 23 hours of web browsing thanks to the 65Wh battery, perfect for those working on the go.

Windows laptops are in a bit of transition thanks to the recent introduction of Microsoft's Copilot PCs. However, that designation currently only applies to systems featuring Qualcomm's Snapdragon X Elite and X Plus chips. But now, with some help from AMD, HP's OmniBook Ultra 14 is packing even better AI performance in a thin and light chassis. Powered by AMD's Ryzen AI 300 series chips, the OmniBook UItra 14 is said to deliver up to 55 TOPS of AI performance, which is more than the 45 TOPS you get from the Snapdragon X Elite and X Plus' Hexagon NPU. HP claims this will support a range of new features including faster AI image generation, improved camera effects in video calls and more.