We consider a decision aggregation problem with two experts who each make a binary recommendation after observing a private signal about an unknown binary world state. An agent, who does not know the joint information structure between signals and states, sees the experts' recommendations and aims to match the action with the true state. Under the scenario, we study whether supplemented additionally with second-order information (each expert's forecast on the other's recommendation) could enable a better aggregation. We adopt a minimax regret framework to evaluate the aggregator's performance, by comparing it to an omniscient benchmark that knows the joint information structure. With general information structures, we show that second-order information provides no benefit -- no aggregator can improve over a trivial aggregator, which always follows the first expert's recommendation. However, positive results emerge when we assume experts' signals are conditionally independent given the world state. When the aggregator is deterministic, we present a robust aggregator that leverages second-order information, which can significantly outperform counterparts without it. Second, when two experts are homogeneous, by adding a non-degenerate assumption on the signals, we demonstrate that random aggregators using second-order information can surpass optimal ones without it. In the remaining settings, the second-order information is not beneficial. We also extend the above results to the setting when the aggregator's utility function is more general.
论文检索
输入标题、作者或关键词，从 432 篇学术成果中精准定位
Recent advances in Machine Learning (ML) and Artificial Intelligence (AI) follow a familiar structure: A firm releases a large, pretrained model. It is designed to be adapted and tweaked by other entities to perform particular, domain-specific functions. The model is heralded as 'general-purpose,' meaning it can be transferred to a wide range of downstream tasks, in a process known as adaptation or fine-tuning. Understanding this process - the strategies, incentives, and interactions involved in the development of AI tools - is crucial for making conclusions about societal implications and regulatory responses, and may provide insights beyond AI about general-purpose technologies. We propose a model of this adaptation process. A Generalist brings the technology to a certain level of performance, and one or more Domain specialist(s) adapt it for use in particular domain(s). Players incur costs when they invest in the technology, so they need to reach a bargaining agreement on how to share the resulting revenue before making their investment decisions. We find that for a broad class of cost and revenue functions, there exists a set of Pareto-optimal profit-sharing arrangements where the players jointly contribute to the technology. Our analysis, which utilizes methods based on bargaining solutions and sub-game perfect equilibria, provides insights into the strategic behaviors of firms in these types of interactions. For example, profit-sharing can arise even when one firm faces significantly higher costs than another. We show that any potential domain specialization will either contribute, free-ride, or abstain in their uptake of the technology, and provide conditions yielding these different responses.
This paper explores the design of a balanced data-sharing marketplace for entities with heterogeneous datasets and machine learning models that they seek to refine using data from other agents. The goal of the marketplace is to encourage participation for data sharing in the presence of such heterogeneity. Our market design approach for data sharing focuses on interim utility balance, where participants contribute and receive equitable utility from refinement of their models. We present such a market model for which we study computational complexity, solution existence, and approximation algorithms for welfare maximization and core stability. We finally support our theoretical insights with simulations on a mean estimation task inspired by road traffic delay estimation.
We study the price of anarchy of the generalized second-price auction where bidders are value maximizers (i.e., autobidders). We show that in general the price of anarchy can be as bad as 0. For comparison, the price of anarchy of running VCG is 1/2 in the autobidding world. We further show a fined-grained price of anarchy with respect to the discount factors (i.e., the ratios of click probabilities between lower slots and the highest slot in each auction) in the generalized second-price auction, which highlights the qualitative relation between the smoothness of the discount factors and the efficiency of the generalized second-price auction.
Clickbait vs. Quality: How Engagement-Based Optimization Shapes the Content Landscape in Online Platforms
PDF ↗Online content platforms commonly use engagement-based optimization when making recommendations. This encourages content creators to invest in quality, but also rewards gaming tricks such as clickbait. To understand the total impact on the content landscape, we study a game between content creators competing on the basis of engagement metrics and analyze the equilibrium decisions about investment in quality and gaming. First, we show the content created at equilibrium exhibits a positive correlation between quality and gaming, and we empirically validate this finding on a Twitter dataset. Using the equilibrium structure of the content landscape, we then examine the downstream performance of engagement-based optimization along two axes. Perhaps counterintuitively, the average quality of content consumed by users can decrease at equilibrium as gaming tricks become more costly for content creators to employ. Moreover, engagement-based optimization can perform worse in terms of user utility than a baseline with random recommendations. Altogether, our results highlight the need to consider content creator incentives when evaluating a platform's choice of optimization metric.
In this research, we study the problem that a collector acquires items from the owner based on the item qualities the owner declares and an independent appraiser's assessments. The owner is interested in maximizing the probability that the collector acquires the items and is the only one who knows the items' factual quality. The appraiser performs her duties with impartiality, but her assessment may be subject to random noises, so it may not accurately reflect the factual quality of the items. The main challenge lies in devising mechanisms that prompt the owner to reveal accurate information, thereby optimizing the collector's expected reward. We consider the menu size of mechanisms as a measure of their practicability and study its impact on the attainable expected reward. For the single-item setting, we design optimal mechanisms with a monotone increasing menu size. Although the reward gap between the simplest and optimal mechanisms is bounded, we show that simple mechanisms with a small menu size cannot ensure any positive fraction of the optimal reward of mechanisms with a larger menu size. For the multi-item setting, we show that an ordinal mechanism that only takes the owner's ordering of the items as input is not incentive-compatible. We then propose a set of Union mechanisms that combine single-item mechanisms. Moreover, we run experiments to examine these mechanisms' robustness against the independent appraiser's assessment accuracy and the items' acquiring rate.
In today's online advertising markets, it is common for advertisers to set long-term budgets. Correspondingly, advertising platforms adopt budget control methods to ensure that advertisers' payments lie within their budgets. Most budget control methods rely on the value distributions of advertisers. However, due to the complex advertising landscape and potential privacy concerns, the platform hardly learns advertisers' true priors. Thus, it is crucial to understand how budget control auction mechanisms perform under unassured priors. This work answers this problem from multiple aspects. Specifically, we examine five budget-constrained parameterized mechanisms: bid-discount/pacing first-price/second-price auctions and the Bayesian revenue-optimal auction. We consider the unassured prior game among the seller and all buyers induced by these five mechanisms in the stochastic model. We restrict the parameterized mechanisms to satisfy the budget-extracting condition, which maximizes the seller's revenue by extracting buyers' budgets as effectively as possible. Our main result shows that the Bayesian revenue-optimal mechanism and the budget-extracting bid-discount first-price mechanism yield the same set of Nash equilibrium outcomes in the unassured prior game. This implies that simple mechanisms can be as robust as the optimal mechanism under unassured priors in the budget-constrained setting. In the symmetric case, we further show that all these five (budget-extracting) mechanisms share the same set of possible outcomes. We further dig into the structural properties of these mechanisms. We characterize sufficient and necessary conditions on the budget-extracting parameter tuple for bid-discount/pacing first-price auctions. Meanwhile, when buyers do not take strategic behaviors, we exploit the dominance relationships of these mechanisms by revealing their intrinsic structures. In summary, our results establish vast connections among budget-constrained auctions with unassured priors and explore their structural properties, particularly highlighting the advantages of first-price mechanisms.
The gig economy features dynamic arriving agents and on-demand services provided. In this context, instant and irrevocable matching decisions are highly desirable due to the low patience of arriving requests. In this paper, we propose an online-matching-based model to tackle the two fundamental issues, matching and pricing, existing in a wide range of real-world gig platforms, including ride-hailing (matching riders and drivers), crowdsourcing markets (pairing workers and tasks), and online recommendations (offering items to customers). Our model assumes the arriving distributions of dynamic agents (e.g., riders, workers, and buyers) are accessible in advance, and they can change over time, which is referred to as Known Heterogeneous Distributions (KHD). In this paper, we initiate variance analysis for online matching algorithms under KHD. Unlike the popular competitive-ratio (CR) metric, the variance of online algorithms' performance is rarely studied due to inherent technical challenges, though it is well linked to robustness. We focus on two natural parameterized sampling policies, denoted by ATT(γ) and SAMP(γ), which appear as foundational bedrock in online algorithm design. We offer rigorous competitive ratio (CR) and variance analyses for both policies. Specifically, we show that ATT(γ) with γ ∈ [0,1/2] achieves a CR of γ and a variance of γ ∙ (1-γ) ∙ B on the total number of matches with B being the total matching capacity. In contrast, SAMP(γ) with γ ∈ [0,1] accomplishes a CR of γ (1-γ) and a variance of γ (1-γ)∙ B with γ = min(γ,1/2). All CR and variance analyses are tight and unconditional of any benchmark. As a byproduct, we prove that ATT(γ=1/2) achieves an optimal CR of 1/2.
Large language models have substantially advanced the state of the art in various AI tasks, such as natural language understanding and text generation, and image processing, and multimodal modeling. In this talk, we will first introduce the development of AI in the past decades, in particular from the angle of China. We will also talk about the opportunities, challenges, and risks of AGI in the future, and its impact on the Web. In the second part of the talk, we will use ChatGLM, an alternative but open sourced model to ChatGPT, as an example to explain our understandings and insights derived during the implementation of the model.
Advances in generative AI and the increasingly easy availability of tools for creating text, code, audio, and images have impacted almost all industry sectors, promising new efficiencies and changing work patterns. The darker side of this same technology is the problematic case of deepfakes created by AI and spread online to humiliate, manipulate, trick, or defraud ordinary individuals and public figures. Transparency, fairness, and beneficence are vital values of responsible and ethical AI. All of these values would preclude harmful uses of AI deep fakes. However, harmful deepfakes are usually the work of fraudsters with little regard for ethics and beyond the reach of the law. So, who should be responsible? Arguably, principles of responsible AI require tech companies and digital platforms to take responsibility for reducing harmful uses of deepfakes. These entities are gatekeepers to the creation and distribution of deepfakes. Therefore, they are ethically obligated to respond to the foreseeable consequential harms arising from generative AI. Increasingly, this is the response of lawmakers. Gatekeeper responsibility envisages that tech producers and platforms will proactively invest in technical solutions to harmful deepfakes, such as watermarking, finetuning, red teaming or automated content moderation, and proactive take-down responses. This response is compelling and might seem straightforward. As always, the details are more complex. The efficiency of the proposed technical responses is still emerging. They raise as yet unaddressed implications for smaller providers and the relations between tech companies and digital platforms. Moreover, even beginning to respond to online deepfakes requires social policy decisions that assess and weigh incommensurable considerations, including retaining trust on the Web, keeping vulnerable groups safe, preserving free speech and creativity, and not stifling the development of potentially beneficial technology. This presentation addresses these problematic choices in responding to the 'wicked' challenge of AI deepfakes on the Web. It proposes a networked response to the problem, embracing multiple relevant actors and influences.
Artificial intelligence (AI) based self-learning or self-improving material discovery systems will enable next-generation material discovery. Herein, we demonstrate how to combine accurate prediction of material performance via first-principles calculation and Bayesian optimization-based active learning to realize a self-improving discovery system for high-performance photosensitizers (PSs). Through self-improving cycles, such a system can improve the model prediction accuracy (best mean absolute error of 0.090 eV for singlet--triplet spitting) and high-performance PS search ability, realizing efficient discovery of PSs. From a molecular space with more than 7 million molecules, 5357 potential high-performance PSs were discovered. Four PSs were further synthesized to show performance comparable with or superior to commercial ones. This work highlights the potential of active learning in first principle-based materials design, and the discovered structures could boost the development of photosensitization-related applications, which is one of the typical examples of how AI can be used to accelerate materials innovation and facilitate science development in general.
One of the fundamental problems that platform algorithms face is the process of inferring user preferences from observed behavior; the vast amounts of data a platform collects become much less useful if they cannot effectively inform this type of inference. Traditional approaches to this problem rely on an often unstated revealed-preference assumption: that choice reveals preference. Yet a long line of work in psychology and behavioral economics reveals the gaps that can open up between choice and preference, and experience with platform dynamics makes clear how it can arise in some of the most basic online settings; for example, we might choose content to consume in the present and then later regret the time we spent on it. More generally, behavioral biases and inconsistent preferences make it highly challenging to appropriately interpret the user data that we observe. We discuss a set of models and algorithms that address this challenge through a process of "inversion", in which an algorithm must try inferring mental states that are not directly measured in the data. The talk is based on joint work with Jens Ludwig, Sendhil Mullainathan, and Manish Raghavan.