It's a seductive distraction from the advances in AI that are most likely to improve or even save your life On April 28, 2022, at a highly anticipated concert in Spokane, Washington, the musician Paul McCartney astonished his audience with a groundbreaking application of AI: He began to perform with a lifelike depiction of his long-deceased musical partner, John Lennon. Using recent advances in audio and video processing, engineers had taken the pair's final performance (London, 1969), separated Lennon's voice and image from the original mix and restored them with lifelike clarity. For years, researchers like me had taught machines to "see" and "hear" in order to make such a moment possible. As McCartney and Lennon appeared to reunite across time and space, the arena fell silent; many in the crowd began to cry. As an AI scientist and lifelong Beatles fan, I felt profound gratitude that we could experience this truly life-changing moment. Later that year, the world was captivated by another major breakthrough: AI conversation.

Research into whether drugs like ayahuasca can mitigate the effects of traumatic brain injury is in its infancy. Pro athletes like the Buffalo Bills' Jordan Poyer are forging ahead anyway. Roam the wide-open halls and cavernous showrooms of the Colorado Convention Center during Psychedelic Science, the world's largest psychedelics conference, and you'll see exhibitors hawking everything from mushroom jewelry, to chewable gummies containing extracts of the psychoactive succulent plant kanna, to broad flat-brim baseball caps emblazoned with "MDMA" and "IBOGA." Booths publicize organizations such as the Ketamine Taskforce and the Psychedelic Parenthood Community, and even, a live-action feature film looking to attract investors. It's a motley, multifarious symposium where indigenous-plant-medicine healers mingle with lanyard-clad pharma-bros, legendary underground LSD chemists, and workaday stoners tottering around in massive red and white toadstool hats that make them look like that cute little mushroom guy from . And yet, oddest among such oddities may be the sight of enormously burly NFL tough guys talking candidly about their feelings.

Amazon has deployed over 750,000 robots to its fulfillment centers over the last decade or so, but now there's a new, shall we say, more sensitive addition. The company has announced Vulcan, its first robot with a sense of touch. It's one in a series of new robots introduced today at Amazon's Delivering the Future event in Germany. Vulcan uses force feedback sensors to monitor how much it's pushing or holding on to an object and, ideally, not damage it. "In the past, when industrial robots have unexpected contact, they either emergency stop or smash through that contact. They often don't even know they have hit something because they cannot sense it."

Amazon has developed a new warehouse robot that uses touch to rummage around shelves to find the right product to ship to customers. The robot, called Vulcan, is a meaningful step towards making robots less sausage-fingered compared to human beings. Honing robots' tactile abilities further may allow them to take on more fulfillment and manufacturing work in the years ahead. Aaron Parness, Amazon's director of robotics AI who led the development of Vulcan, explains that touch sensing helps the robot push items around on a shelf and identify what it's after. "When you're trying to stow [or pick] items in one of these pods, you can't really do that task without making contact with the other items," he says.

High-level synthesis (HLS) has significantly advanced the automation of digital circuits design, yet the need for expertise and time in pragma tuning remains challenging. Existing solutions for the design space exploration (DSE) adopt either heuristic methods, lacking essential information for further optimization potential, or predictive models, missing sufficient generalization due to the time-consuming nature of HLS and the exponential growth of the design space. To address these challenges, we propose Deep Inverse Design for HLS (DID4HLS), a novel approach that integrates graph neural networks and generative models. DID4HLS iteratively optimizes hardware designs aimed at compute-intensive algorithms by learning conditional distributions of design features from post-HLS data. Compared to four state-of-the-art DSE baselines, our method achieved an average improvement of 42.5% on average distance to reference set (ADRS) compared to the best-performing baselines across six benchmarks, while demonstrating high robustness and efficiency.

Today, the detection of AI-generated content is receiving more and more attention. Our idea is to go beyond detection and try to recover the prompt used to generate a text. This paper, to the best of our knowledge, introduces the first investigation in this particular domain without a closed set of tasks. Our goal is to study if this approach is promising. We experiment with zero-shot and few-shot in-context learning but also with LoRA fine-tuning. After that, we evaluate the benefits of using a semi-synthetic dataset. For this first study, we limit ourselves to text generated by a single model. The results show that it is possible to recover the original prompt with a reasonable degree of accuracy.

The robustness of recent Large Language Models (LLMs) has become increasingly crucial as their applicability expands across various domains and real-world applications. Retrieval-Augmented Generation (RAG) is a promising solution for addressing the limitations of LLMs, yet existing studies on the robustness of RAG often overlook the interconnected relationships between RAG components or the potential threats prevalent in real-world databases, such as minor textual errors. In this work, we investigate two underexplored aspects when assessing the robustness of RAG: 1) vulnerability to noisy documents through low-level perturbations and 2) a holistic evaluation of RAG robustness. Furthermore, we introduce a novel attack method, the Genetic Attack on RAG (\textit{GARAG}), which targets these aspects. Specifically, GARAG is designed to reveal vulnerabilities within each component and test the overall system functionality against noisy documents. We validate RAG robustness by applying our \textit{GARAG} to standard QA datasets, incorporating diverse retrievers and LLMs. The experimental results show that GARAG consistently achieves high attack success rates. Also, it significantly devastates the performance of each component and their synergy, highlighting the substantial risk that minor textual inaccuracies pose in disrupting RAG systems in the real world.

Instruction Tuning (IT), the process of training large language models (LLMs) using instruction-response pairs, has emerged as the predominant method for transforming base pre-trained LLMs into open-domain conversational agents. While IT has achieved notable success and widespread adoption, its limitations and shortcomings remain underexplored. In this paper, through rigorous experiments and an in-depth analysis of the changes LLMs undergo through IT, we reveal various limitations of IT. In particular, we show that (1) IT fails to enhance knowledge or skills in LLMs. LoRA fine-tuning is limited to learning response initiation and style tokens, and full-parameter fine-tuning leads to knowledge degradation. (2) Copying response patterns from IT datasets derived from knowledgeable sources leads to a decline in response quality. (3) Full-parameter fine-tuning increases hallucination by inaccurately borrowing tokens from conceptually similar instances in the IT dataset for generating responses. (4) Popular methods to improve IT do not lead to performance improvements over a simple LoRA fine-tuned model. Our findings reveal that responses generated solely from pre-trained knowledge consistently outperform responses by models that learn any form of new knowledge from IT on open-source datasets. We hope the insights and challenges revealed inspire future work.

Abstract-- We present a contact-based phenotyping robot platform that can autonomously insert nitrate sensors into cornstalks to proactively monitor macronutrient levels in crops. This task is challenging because inserting such sensors requires sub-centimeter precision in an environment which contains high levels of clutter, lighting variation, and occlusion. To address these challenges, we develop a robust perceptionaction pipeline to detect and grasp stalks, and create a custom robot gripper which mechanically aligns the sensor before inserting it into the stalk. Through experimental validation on 48 unique stalks in a cornfield in Iowa, we demonstrate our platform's capability of detecting a stalk with 94% success, grasping a stalk with 90% success, and inserting a sensor with 60% success. In addition to developing an autonomous phenotyping research platform, we share key challenges and insights obtained from deployment in the field. With the development of artificial intelligence in computer vision and robotics, the agricultural sector is poised to implement precision agriculture methods to enhance crop production efficiency and minimize environmental footprint [1]. Figure 1: Robot inserting sensors into cornstalks to monitor plant nitrate concentration in Curtiss Farm, Iowa.

Our test sets were too small and too haphazard to support statistically valid conclusions, but they were suggestive of a number of conclusions. We summarize these here, and discuss them at greater length in section 7. Over the kinds of problems tested, GPT-4 with either plug-in is significantly stronger than GPT-4 by itself, or, almost certainly, than any AI that existed a year ago. However it is still far from reliable; it often outputs a wrong answer or fails to output any answer. In terms of overall score, we would judge that these systems performs on the level of a middling undergraduate student. However, their capacities and weaknesses do not align with a human student; the systems solve some problems that even capable students would find challenging, whereas they fail on some problems that even middling high school students would find easy.