When solving two-player zero-sum games, multi-agent reinforcement learning (MARL) algorithms often create populations of agents where, at each iteration, a new agent is discovered as the best response to a mixture over the opponent population. Within such a process, the update rules of who to compete with (i.e., the opponent mixture) and how to beat them (i.e., finding best responses) are underpinned by manually developed game theoretical principles such as fictitious play and Double Oracle. In this paper, we introduce a novel framework--Neural Auto-Curricula (NAC)--that leverages meta-gradient descent to automate the discovery of the learning update rule without explicit human design. Specifically, we parameterise the opponent selection module by neural networks and the best-response module by optimisation subroutines, and update their parameters solely via interaction with the game engine, where both players aim to minimise their exploitability. Surprisingly, even without human design, the discovered MARL algorithms achieve competitive or even better performance with the state-of-the-art population-based game solvers (e.g., PSRO) on Games of Skill, differentiable Lotto, non-transitive Mixture Games, Iterated Matching Pennies, and Kuhn Poker. Additionally, we show that NAC is able to generalise from small games to large games, for example training on Kuhn Poker and outperforming PSRO on Leduc Poker. Our work inspires a promising future direction to discover general MARL algorithms solely from data.

On this episode of, we discuss OpenAI's new teen safety features, the right's retaliation against critics of the late Charlie Kirk, and more of the week's biggest stories. Charlie Kirk (R) shaking hands with US President Donald Trump as he speaks on stage at America Fest 2024 in Phoenix, Arizona. All products featured on WIRED are independently selected by our editors. However, we may receive compensation from retailers and/or from purchases of products through these links. In today's episode, our host Zöe Schiffer is joined by WIRED's senior culture editor Manisha Krishnan to run through five of the best stories we published this week--from OpenAI implementing teen safety features to how human design is the new astrology. Zöe and Manisha also discuss the reverberating reactions to Charlie Kirk's death and why the work of many creators, from comic book artists to late night show hosts, is getting cancelled.

Human Design Is Blowing Up. The astrology-like system uses birthdates to break people into personality types and even find love and riches. From sleeping arrangements to diets, some are taking it very seriously. Travis Day regrets being so strict with his ex-partner about never sleeping in the same bed. According to human design, the New Age "synthesis" of astrology, the I Ching book of Chinese wisdom, Kabbalah, and the chakra system, "everyone should sleep in their own bed at night" to preserve their auras, says the blond 38-year-old Los Angeles county-based surfer, who's been following the practice rigidly for the last five years.

When solving two-player zero-sum games, multi-agent reinforcement learning (MARL) algorithms often create populations of agents where, at each iteration, a new agent is discovered as the best response to a mixture over the opponent population. Within such a process, the update rules of "who to compete with" (i.e., the opponent mixture) and "how to beat them" (i.e., finding best responses) are underpinned by manually developed game theoretical principles such as fictitious play and Double Oracle. In this paper, we introduce a novel framework--Neural Auto-Curricula (NAC)--that leverages meta-gradient descent to automate the discovery of the learning update rule without explicit human design. Specifically, we parameterise the opponent selection module by neural networks and the best-response module by optimisation subroutines, and update their parameters solely via interaction with the game engine, where both players aim to minimise their exploitability. Surprisingly, even without human design, the discovered MARL algorithms achieve competitive or even better performance with the state-of-the-art population-based game solvers (e.g., PSRO) on Games of Skill, differentiable Lotto, non-transitive Mixture Games, Iterated Matching Pennies, and Kuhn Poker.

Nvidia has been quick to hop on the artificial intelligence bus一with many of its consumer facing technologies, such as Deep Learning Super Sampling (DLSS) (opens in new tab) and AI-accelerated denoising exemplifying that. However, it has also found many uses for AI in its silicon development process and, as Nvidia's chief scientist Bill Dally (opens in new tab) said in a GTC conference, even designing new hardware. Dally outlines a few use cases for AI in its own development process of the latest and greatest graphic cards (opens in new tab) (among other things), as noted by HPC Wire (opens in new tab). "It's natural as an expert in AI that we would want to take that AI and use it to design better chips," Dally says. "We do this in a couple of different ways. The first and most obvious way is we can take existing computer-aided design tools that we have. For example, we have one that takes a map of where power is used in our GPUs, and predicts how far the voltage grid drops一what's called IR drop for current times resistance drop. Running this on a conventional CAD tool takes three hours."

AI is arguably the biggest tech topic of 2018. From Google Duplex's human imitations and Spotify's song recommendations to Uber's self-driving cars and the Pentagon's use of GoogleAI, the technology seems to offer everything to everyone. You could say AI has become synonymous with progress via computing. However, not all AI is created equal, and for AI to fulfill its many promises, it needs to be creative. Let's start by addressing what I mean by "creative."