Machine learning models are often personalized with categorical attributes that are protected, sensitive, self-reported, or costly to acquire. In this work, we show models that are personalized with group attributes can reduce performance at a group level. We propose formal conditions to ensure the "fair use" of group attributes in prediction tasks by training one additional model -- i.e., collective preference guarantees to ensure that each group who provides personal data will receive a tailored gain in performance in return. We present sufficient conditions to ensure fair use in empirical risk minimization and characterize failure modes that lead to fair use violations due to standard practices in model development and deployment. We present a comprehensive empirical study of fair use in clinical prediction tasks. Our results demonstrate the prevalence of fair use violations in practice and illustrate simple interventions to mitigate their harm.

We tested Intel's new tool, "FakeCatcher", on videos of Donald Trump and Joe Biden - with mixed results.

Meta's global policy head, Sir Nick Clegg, has backed calls for an international agency to guide the regulation of artificial intelligence if it becomes autonomous, saying governments globally should avoid "fragmented" laws around the technology. But Clegg downplayed suggestions of payment for content creators like artists or news outlets whose work is scraped to teach chatbots and generative AI, suggesting such information would be available under fair use arrangements. "Creators who lean in to using this technology, rather than trying to block it or slow it down or prevent it from drawing on their own creative output, will in the long run be better placed than those who set their face against this technology," Clegg told Guardian Australia. "We believe we're using [data] entirely in line with existing law. A lot of this data is being transformed in the way it's being deployed by these generative AI models. In the long run, I can't see how you put the genie back in the bottle, given that these models do use publicly available information across the internet, and not unreasonably so."

Americans reveled if they would let the AI dating website Keeper set them up on blind date. The website asks users a list of questions before matching them. NEW YORK CITY – New Yorkers revealed whether they would let an artificial intelligence-based program set them up on a date. "Would I ever do it? If got desperate, but I hope it would never get to that point," Nick, from New York, told Fox News.

The governor of Russia's Belgorod region has said that Ukraine fired cluster munitions at a village near the Ukrainian border on Friday, but that there were no casualties or damage. The governor made the statement on Saturday during a daily briefing on his Telegram channel, without providing visual evidence. There was no immediate comment from Ukrainian authorities. "In Belgorod district, 21 artillery shells and three cluster munitions from a multiple-launch rocket system were fired at the village of Zhuravlevka," Governor Vyacheslav Gladkov said. Ukraine received cluster bombs from the United States this month, but it has pledged to use them only to dislodge concentrations of enemy soldiers. They contain dozens of small bomblets that rain shrapnel over a wide area, but are banned in many countries due to the potential danger they pose to civilians.

Seekr Technologies CEO Pat Condo spoke with Fox News Digital about a partnership with Bear Grylls to encourage digital media literacy among young people. An aspiring British politician crowdsourced his platform and used artificial intelligence (AI) to build his manifesto, a "brave" measure despite its seeming failure, according to one expert. "Andrew Gray had a brave idea, but having finished 11th out of 13 candidates and with just 99 votes, I wouldn't expect mainstream politicians to rush to copy his tactics just yet," Alan Mendoza, co-founder and executive director of the Henry Jackson Society, told Fox News Digital. "That said, it's clear that AI is going to have an impact on how political parties in the U.K. source and target data going forwards, as well as focus their campaigns," he argued. "We may not have to wait that long for the first AI-inspired victorious candidate, but they will undoubtedly emerge from one of the major parties, with all the electoral advantages they already possess."

Washington – U.S. President Joe Biden evoked AI's "enormous" risk and promise Friday at a White House meeting with tech leaders who committed to guarding against everything from cyberattacks to fraud as the sector revolutionizes society. "It is astounding," Biden said, highlighting AI's "enormous, enormous promise of both risk to our society and our economy and our national security, but also incredible opportunities." Standing alongside top representatives from Amazon, Anthropic, Google, Inflection, Meta, Microsoft and OpenAI, Biden said the cutting-edge companies had made commitments to "guide responsible innovation" as AI rips ever deeper into personal and business life. This could be due to a conflict with your ad-blocking or security software. Please add japantimes.co.jp and piano.io to your list of allowed sites.

President Biden held a conference about artificial intelligence, but left as he was asked about a technological attack from China-based hackers. President Biden shuffled away from a question about Chinese hackers after remarks on AI at the White House, sparking laughter across Twitter. After quipping, "I'm the AI," and warning, "If any of you think I'm Abe Lincoln, blame it on the AI," Biden spoke about the future of artificial intelligence, noting that it "promises an enormous, enormous promise of both risk to our society and our economy and our national security, but also incredible opportunities." Among his calls for action, he demanded companies working with AI ensure they are "rooting out bias and discrimination." President, can you tell us about the hacking of cabinet officials by China and the threshold of concern you have about that, sir?" Biden ignored the question and asked his staff, "Ready?

Robust reinforcement learning (RL) seeks to train policies that can perform well under environment perturbations or adversarial attacks. Existing approaches typically assume that the space of possible perturbations remains the same across timesteps. However, in many settings, the space of possible perturbations at a given timestep depends on past perturbations. We formally introduce temporally-coupled perturbations, presenting a novel challenge for existing robust RL methods. To tackle this challenge, we propose GRAD, a novel game-theoretic approach that treats the temporally-coupled robust RL problem as a partially-observable two-player zero-sum game. By finding an approximate equilibrium in this game, GRAD ensures the agent's robustness against temporally-coupled perturbations. Empirical experiments on a variety of continuous control tasks demonstrate that our proposed approach exhibits significant robustness advantages compared to baselines against both standard and temporally-coupled attacks, in both state and action spaces.

Advancements in generative artificial intelligence (AI) and large language models (LLMs) have fueled the development of many educational technology innovations that aim to automate the often time-consuming and laborious tasks of generating and analysing textual content (e.g., generating open-ended questions and analysing student feedback survey) (Kasneci et al., 2023; Wollny et al., 2021; Leiker et al., 2023). LLMs are generative artificial intelligence models that have been trained on an extensive amount of text data, capable of generating human-like text content based on natural language inputs. Specifically, these LLMs, such as Bidirectional Encoder Representations from Transformers (BERT) (Devlin et al., 2018) and Generative Pre-trained Transformer (GPT) (Brown et al., 2020), utilise deep learning and self-attention mechanisms (Vaswani et al., 2017) to selectively attend to the different parts of input texts, depending on the focus of the current tasks, allowing the model to learn complex patterns and relationships among textual contents, such as their semantic, contextual, and syntactic relationships (Min et al., 2021; Liu et al., 2023). As several LLMs (e.g., GPT-3 and Codex) have been pre-trained on massive amounts of data across multiple disciplines, they are capable of completing natural language processing tasks with little (few-shot learning) or no additional training (zero-shot learning) (Brown et al., 2020; Wu et al., 2023). This could lower the technological barriers to LLMs-based innovations as researchers and practitioners can develop new educational technologies by fine-tuning LLMs on specific educational tasks without starting from scratch (Caines et al., 2023; Sridhar et al., 2023). The recent release of ChatGPT, an LLMs-based generative AI chatbot that requires only natural language prompts without additional model training or fine-tuning (OpenAI, 2023), has further lowered the barrier for individuals without technological background to leverage the generative powers of LLMs. Although educational research that leverages LLMs to develop technological innovations for automating educational tasks is yet to achieve its full potential (i.e., most works have focused on improving model performances (Kurdi et al., 2020; Ramesh and Sanampudi, 2022)), a growing body of literature hints at how different stakeholders could potentially benefit from such innovations.