How might messages about large language models (LLMs) found in public discourse influence the way people think about and interact with these models? To explore this question, we randomly assigned participants (N = 470) to watch short informational videos presenting LLMs as either machines, tools, or companions -- or to watch no video. We then assessed how strongly they believed LLMs to possess various mental capacities, such as the ability to have intentions or remember things. We found that participants who watched video messages presenting LLMs as companions reported believing that LLMs more fully possessed these capacities than did participants in other groups. In a follow-up study (N = 604), we replicated these findings and found nuanced effects on how these videos also impact people's reliance on LLM-generated responses when seeking out factual information. Together, these studies suggest that messages about LLMs -- beyond technical advances -- may shape what people believe about these systems and how they rely on LLM-generated responses.

A quintessential feature of human intelligence is the ability to create ad hoc conventions over time to achieve shared goals efficiently. We investigate how communication strategies evolve through repeated collaboration as people coordinate on shared procedural abstractions. To this end, we conducted an online unimodal study (n = 98) using natural language to probe abstraction hierarchies. In a follow-up lab study (n = 40), we examined how multimodal communication (speech and gestures) changed during physical collaboration. Pairs used augmented reality to isolate their partner’s hand and voice; one participant viewed a 3D virtual tower and sent instructions to the other, who built the physical tower. Participants became faster and more accurate by establishing linguistic and gestural abstractions and using cross-modal redundancy to emphasize key changes from previous interactions. Based on these findings, we extend probabilistic models of convention formation to multimodal settings, capturing shifts in modality preferences. Our findings and model provide building blocks for designing convention-aware intelligent agents situated in the physical world.

Sketching serves as a versatile tool for externalizing ideas, enabling rapid exploration and visual communication that spans various disciplines. While artificial systems have driven substantial advances in content creation and human-computer interaction, capturing the dynamic and abstract nature of human sketching remains challenging. In this work, we introduce SketchAgent, a language-driven, sequential sketch generation method that enables users to create, modify, and refine sketches through dynamic, conversational interactions. Our approach requires no training or fine-tuning. Instead, we leverage the sequential nature and rich prior knowledge of off-the-shelf multimodal large language models (LLMs). We present an intuitive sketching language, introduced to the model through in-context examples, enabling it to "draw" using string-based actions. These are processed into vector graphics and then rendered to create a sketch on a pixel canvas, which can be accessed again for further tasks. By drawing stroke by stroke, our agent captures the evolving, dynamic qualities intrinsic to sketching. We demonstrate that SketchAgent can generate sketches from diverse prompts, engage in dialogue-driven drawing, and collaborate meaningfully with human users.

B-roll is vital when producing high-quality videos, but finding the right images can be difficult and time-consuming. Moreover, what B-roll is most effective can depend on a video content creator’s intent—is the goal to entertain, to inform, or something else? While new text-to-image generation models provide promising avenues for streamlining B-roll production, it remains unclear how these tools can provide support for content creators with different goals. To close this gap, we aimed to understand how video content creator’s goals guide which visual concepts they prioritize for B-roll generation. Here we introduce a benchmark containing judgments from > 800 people as to which terms in 12 video transcripts should be assigned highest priority for B-roll imagery accompaniment. We verified that participants reliably prioritized different visual concepts depending on whether their goal was help produce informative or entertaining videos. We next explored how well several algorithms, including heuristic approaches and large language models (LLMs), could predict systematic patterns in human judgments. We found that none of these methods fully captured human judgments in either goal condition, with state-of-the-art LLMs (i.e., GPT-4) even underperforming a baseline that sampled only nouns or nouns and adjectives. Overall, our work identifies opportunities to develop improved algorithms to support video production workflows.

Realizing a designer’s intent in software currently requires tedious manipulation of geometric primitives, such as points and curves. By contrast, designers routinely communicate more abstract design goals to one another using an efficient combination of natural language and drawings. What would it take to develop artificial systems that understand how humans naturally convey design intent, and thereby enable more seamless interactions between humans and machines throughout the design process? First, it is vital to establish benchmarks that showcase the full range of strategies that humans use to successfully communicate about design intent. Here we take initial steps towards that goal by conducting an online study in which pairs of human participants – a “Designer” and “Maker” – collaborated over multiple turns to recreate target designs. In each turn, Designers sent messages containing language, drawings, or both to the Maker, describing how to modify an existing design toward the target. We found a preference for communicating using drawings in early turns and observed several multimodal strategies for conveying design intent. By comparing how human Makers and GPT-4V carried out instructions, we identify a gap in human and machine understanding of multimodal instructions and suggest a path for bridging this gap.

Humans effortlessly grasp the connection between sketches and real-world objects, even when these sketches are far from realistic. Moreover, human sketch understanding goes beyond categorization -- critically, it also entails understanding how individual elements within a sketch correspond to parts of the physical world it represents. What are the computational ingredients needed to support this ability? Towards answering this question, we make two contributions: first, we introduce a new sketch-photo correspondence benchmark, PSC6k, containing 150K annotations of 6250 sketch-photo pairs across 125 object categories, augmenting the existing Sketchy dataset with fine-grained correspondence metadata. Second, we propose a self-supervised method for learning dense correspondences between sketch-photo pairs, building upon recent advances in correspondence learning for pairs of photos. Our model uses a spatial transformer network to estimate the warp flow between latent representations of a sketch and photo extracted by a contrastive learning-based ConvNet backbone. We found that this approach outperformed several strong baselines and produced predictions that were quantitatively consistent with other warp-based methods. However, our benchmark also revealed systematic differences between predictions of the suite of models we tested and those of humans. Taken together, our work suggests a promising path towards developing artificial systems that achieve more human-like understanding of visual images at different levels of abstraction. Project page: https://photo-sketch-correspondence.github.io

Sketching is a powerful tool for creating abstract images that are sparse but meaningful. Sketch understanding poses fundamental challenges for general-purpose vision algorithms because it requires robustness to the sparsity of sketches relative to natural visual inputs and because it demands tolerance for semantic ambiguity, as sketches can reliably evoke multiple meanings. While current vision algorithms have achieved high performance on a variety of visual tasks, it remains unclear to what extent they understand sketches in a human-like way. Here we introduce $\texttt{SEVA}$, a new benchmark dataset containing approximately 90K human-generated sketches of 128 object concepts produced under different time constraints, and thus systematically varying in sparsity. We evaluated a suite of state-of-the-art vision algorithms on their ability to correctly identify the target concept depicted in these sketches and to generate responses that are strongly aligned with human response patterns on the same sketch recognition task. We found that vision algorithms that better predicted human sketch recognition performance also better approximated human uncertainty about sketch meaning, but there remains a sizable gap between model and human response patterns. To explore the potential of models that emulate human visual abstraction in generative tasks, we conducted further evaluations of a recently developed sketch generation algorithm (Vinker et al., 2022) capable of generating sketches that vary in sparsity. We hope that public release of this dataset and evaluation protocol will catalyze progress towards algorithms with enhanced capacities for human-like visual abstraction.

General physical scene understanding requires more than simply localizing and recognizing objects -- it requires knowledge that objects can have different latent properties (e.g., mass or elasticity), and that those properties affect the outcome of physical events. While there has been great progress in physical and video prediction models in recent years, benchmarks to test their performance typically do not require an understanding that objects have individual physical properties, or at best test only those properties that are directly observable (e.g., size or color). This work proposes a novel dataset and benchmark, termed Physion++, that rigorously evaluates visual physical prediction in artificial systems under circumstances where those predictions rely on accurate estimates of the latent physical properties of objects in the scene. Specifically, we test scenarios where accurate prediction relies on estimates of properties such as mass, friction, elasticity, and deformability, and where the values of those properties can only be inferred by observing how objects move and interact with other objects or fluids. We evaluate the performance of a number of state-of-the-art prediction models that span a variety of levels of learning vs. built-in knowledge, and compare that performance to a set of human predictions. We find that models that have been trained using standard regimes and datasets do not spontaneously learn to make inferences about latent properties, but also that models that encode objectness and physical states tend to make better predictions. However, there is still a huge gap between all models and human performance, and all models' predictions correlate poorly with those made by humans, suggesting that no state-of-the-art model is learning to make physical predictions in a human-like way. These results show that current deep learning models that succeed in some settings nevertheless fail to achieve human-level physical prediction in other cases, especially those where latent property inference is required. Project page: https://dingmyu.github.io/physion_v2/

While most crowd work seeks consistent answers, creative domains often seek more diverse input. The typical crowd mechanisms for controlling quality may stifle creativity, yet removing them altogether could just produce noise. Schemas and metadata provide two mechanisms for embedding existing knowledge into task environments. Schemas are expert-derived patterns designed to structure how people think through a problem. Metadata, on the other hand, illustrate a range of creative input that fits within the structure of a schema. To understand the relative effects of schemas and metadata, we conducted a study where crowd workers are asked to generate creative interpretations for a set of placemaking examples. Crowd workers were guided either by schema plus metadata, schema alone, or neither. We found that showing schema along with crowd-produced metadata helped workers contribute interpretations that are both more on-topic and diverse, compared to using the schema alone or no schema. We discuss the implications on how crowds can creatively build on insights shared by others.

Sketching is one of the most accessible techniques for communicating our ideas quickly and for collaborating in real time. Here we present a web-based environment for collaborative sketching of everyday visual concepts. We explore the integration of an artificial agent, instantiated as a recurrent neural network, who is both cooperative and responsive to actions performed by its human collaborator. To evaluate the quality of the sketches produced in this environment, we conducted an experimental user study and found that sketches produced collaboratively carried as much semantically relevant information as those produced by humans on their own. Further control analyses suggest that the semantic information in these sketches were indeed the product of collaboration, rather than attributable to the contributions of the human or the artificial agent alone. Taken together, our findings attest to the potential of systems enabling real-time collaboration between humans and machines to create novel and meaningful content.