Dean.Bot was the brainchild of Silicon Valley entrepreneurs Matt Krisiloff and Jed Somers, who had started a super PAC supporting Phillips (Minn.) The PAC had received 1 million from hedge fund manager Bill Ackman, the billionaire activist who led the charge to oust Harvard University president Claudine Gay.

Sadeghi and a colleague first noticed in April that there were a lot of posts on X that contained error messages they recognized from ChatGPT, suggesting accounts were using the chatbot to compose tweets automatically. They began searching for those phrases elsewhere online, including in Google search results, and found dozens of websites purporting to be news outlets that contained the telltale error messages.

OpenAI Chief Executive Officer Sam Altman, who has been working to raise billions of dollars from global investors for a chip venture, aims to use the funds to set up a network of factories to manufacture semiconductors, according to several people with knowledge of the plans. Altman has had conversations with several large potential investors in the hopes of raising the vast sums needed for chip fabrication plants, or fabs, as they're known colloquially, said the people, who requested anonymity because the conversations are private. Firms that have held discussions with Altman include Abu Dhabi-based G42, people told Bloomberg last month, and SoftBank Group, some of the people said. The project would involve working with top chip manufacturers, and the network of fabs would be global in scope, some of the people said.

Text-to-image diffusion models are a class of deep generative models that have demonstrated an impressive capacity for high-quality image generation. However, these models are susceptible to implicit biases that arise from web-scale text-image training pairs and may inaccurately model aspects of images we care about. This can result in suboptimal samples, model bias, and images that do not align with human ethics and preferences. In this paper, we present an effective scalable algorithm to improve diffusion models using Reinforcement Learning (RL) across a diverse set of reward functions, such as human preference, compositionality, and fairness over millions of images. We illustrate how our approach substantially outperforms existing methods for aligning diffusion models with human preferences. We further illustrate how this substantially improves pretrained Stable Diffusion (SD) models, generating samples that are preferred by humans 80.3% of the time over those from the base SD model while simultaneously improving both the composition and diversity of generated samples.

This paper studies long-term fair machine learning which aims to mitigate group disparity over the long term in sequential decision-making systems. To define long-term fairness, we leverage the temporal causal graph and use the 1-Wasserstein distance between the interventional distributions of different demographic groups at a sufficiently large time step as the quantitative metric. Then, we propose a three-phase learning framework where the decision model is trained on high-fidelity data generated by a deep generative model. We formulate the optimization problem as a performative risk minimization and adopt the repeated gradient descent algorithm for learning. The empirical evaluation shows the efficacy of the proposed method using both synthetic and semi-synthetic datasets.

Untrustworthy users can misuse image generators to synthesize high-quality deepfakes and engage in unethical activities. Watermarking deters misuse by marking generated content with a hidden message, enabling its detection using a secret watermarking key. A core security property of watermarking is robustness, which states that an attacker can only evade detection by substantially degrading image quality. Assessing robustness requires designing an adaptive attack for the specific watermarking algorithm. When evaluating watermarking algorithms and their (adaptive) attacks, it is challenging to determine whether an adaptive attack is optimal, i.e., the best possible attack. We solve this problem by defining an objective function and then approach adaptive attacks as an optimization problem. The core idea of our adaptive attacks is to replicate secret watermarking keys locally by creating surrogate keys that are differentiable and can be used to optimize the attack's parameters. We demonstrate for Stable Diffusion models that such an attacker can break all five surveyed watermarking methods at no visible degradation in image quality. Optimizing our attacks is efficient and requires less than 1 GPU hour to reduce the detection accuracy to 6.3% or less. Our findings emphasize the need for more rigorous robustness testing against adaptive, learnable attackers.

OpenAI, the parent company of the popular artificial intelligence chatbot platform ChatGPT, altered its usage policy to get rid of a prohibition on using their technology for "military and warfare." OpenAI's usage policy specifically banned the use of its technology for "weapons development, military and warfare" before January 10 of this year, but that policy has since been updated to only disallow use that would "bring harm to others," according to a report from Computer World. "Our policy does not allow our tools to be used to harm people, develop weapons, for communications surveillance, or to injure others or destroy property," an OpenAI spokesperson told Fox News Digital. "There are, however, national security use cases that align with our mission. For example, we are already working with DARPA to spur the creation of new cybersecurity tools to secure open source software that critical infrastructure and industry depend on. It was not clear whether these beneficial use cases would have been allowed under'military' in our previous policies. So the goal with our policy update is to provide clarity and the ability to have these discussions."

In recent years, generative AI has undergone major advancements, demonstrating significant promise in augmenting human productivity. Notably, large language models (LLM), with ChatGPT-4 as an example, have drawn considerable attention. Numerous articles have examined the impact of LLM-based tools on human productivity in lab settings and designed tasks or in observational studies. Despite recent advances, field experiments applying LLM-based tools in realistic settings are limited. This paper presents the findings of a field randomized controlled trial assessing the effectiveness of LLM-based tools in providing unmonitored support services for information retrieval.

Machine learning (ML), the process of leveraging algorithms and optimization to infer strategies for solving learning tasks, has enabled some of the greatest developments in artificial intelligence (AI) in the last decade, enabling the automated segmentation or class identification of images, the ability to answer nearly any text-based question, and the ability to generate images never seen before. In biomedical research, many of these ML models are quickly being applied to medical images and decision support systems in conjunction with a significant shift from traditional and statistical methods to increasing application of deep learning models. At the same time, the importance of both plentiful and well-curated data has become better understood, coinciding as of the time of writing this article with the incredible premise of OpenAI's ChatGPT and GPT-4 engines as well as other generative AI models which are trained on a vast, well-curated, and diverse array of content from across the internet [1]. As more data has become available, a wider selection of datasets containing more than one modality has also enabled growth in the multimodal research sphere. Multimodal data is intrinsic to biomedical research and clinical care.

Meta is reorganizing its AI teams as it joins the growing ranks of companies trying to create artificial general intelligence, or AGI. Mark Zuckerberg, who has been increasingly focused on the company's AI research, said the change would help the company "accelerate" its research and, eventually, improve the metaverse. Meta currently has two teams pursuing AI research: the Fundamental AI Research (FAIR) team, started in 2013, and a team solely focused on generative AI experiences for users of its apps. With the change, Zuckerberg said, the company would bring the two "closer together" as it looks to expand both groups. Meta's CEO didn't say how many workers it might add to its AI efforts, but the expansion is notable considering the company has shed more than 20,000 jobs since 2022.