A specialized glass layer could make tomorrow's computers faster and more energy efficient. An early version of the glass substrate developed by Absolics. Human-made glass is thousands of years old. But it's now poised to find its way into the AI chips used in the world's newest and largest data centers. This year, a South Korean company called Absolics is planning to start commercial production of special glass panels designed to make next-generation computing hardware more powerful and energy efficient. Other companies, including Intel, are also pushing forward in this area.

Hosted by Eleanor Drage and Kerry McInerney, The Good Robot is a podcast which explores the many complex intersections between gender, feminism and technology. In this episode, we talk to Tomasz Hollanek, researcher at the Leverhulme Centre for the Future of Intelligence at the University of Cambridge. Tomasz argues that design is central to AI ethics and explores the role designers should play in shaping ethical AI systems. The conversation examines the importance of AI literacy, the responsibilities of journalists in reporting on AI technologies, and how design choices embed social and political values into AI. Together, we reflect on how critical design can challenge existing power dynamics and open up more just and inclusive approaches to human-AI interaction.

Imagine observing someone scratching their arm; to understand why, additional context would be necessary. However, spotting a mosquito nearby would immediately offer a likely explanation for the person's discomfort, thereby alleviating

Designing reward functions for efficiently guiding reinforcement learning (RL) agents toward specific behaviors is a complex task.

Itisoften the responsibility of the agent designer toconstruct thistargetwhich,inrichandcomplexenvironments,constitutesaonerousburden; without full knowledge of the environment itself, a designer may forge a suboptimal learning target that poorly balances the amount ofinformation an agent must acquire to identify the target against the target's associated performance shortfall.

Volvo's Electric EX60 SUV Has a 400-Mile Range--and Rethinks the Humble Seat Belt The Swedish brand's latest computer-packed EV hopes to take on and beat the BMW iX3. Alongside the chosen few in WIRED's breakdown of the most anticipated EVs coming this year, the arrival of the Volvo EX60 has also been eagerly awaited. This is mainly because of the impressive stats surrounding the car; the headline claim is a range of more than 400 miles. Sitting between the EX40 and EX90, the new EV looks more like a sibling of the entry-level EX30, which is a good car but too fast for its own good . Plus, the reveal images here from Volvo initially seem to show that the design team has figured out a way to remove the unsightly lidar roofline bulges that in some eyes ruined the finished aesthetic of the EX90.

Designing reward functions for efficiently guiding reinforcement learning (RL) agents toward specific behaviors is a complex task.This is challenging since it requires the identification of reward structures that are not sparse and that avoid inadvertently inducing undesirable behaviors. Naively modifying the reward structure to offer denser and more frequent feedback can lead to unintended outcomes and promote behaviors that are not aligned with the designer's intended goal. Although potential-based reward shaping is often suggested as a remedy, we systematically investigate settings where deploying it often significantly impairs performance. To address these issues, we introduce a new framework that uses a bi-level objective to learn \emph{behavior alignment reward functions}. These functions integrate auxiliary rewards reflecting a designer's heuristics and domain knowledge with the environment's primary rewards.

To regulate a social system comprised of self-interested agents, economic incentives are often required to induce a desirable outcome. This incentive design problem naturally possesses a bilevel structure, in which a designer modifies the payoffs of the agents with incentives while anticipating the response of the agents, who play a non-cooperative game that converges to an equilibrium. The existing bilevel optimization algorithms raise a dilemma when applied to this problem: anticipating how incentives affect the agents at equilibrium requires solving the equilibrium problem repeatedly, which is computationally inefficient; bypassing the time-consuming step of equilibrium-finding can reduce the computational cost, but may lead the designer to a sub-optimal solution. To address such a dilemma, we propose a method that tackles the designer's and agents' problems simultaneously in a single loop. Specifically, at each iteration, both the designer and the agents only move one step. Nevertheless, we allow the designer to gradually learn the overall influence of the incentives on the agents, which guarantees optimality after convergence. The convergence rate of the proposed scheme is also established for a broad class of games.