Goto

Collaborating Authors

 query


Ads in Chatbots Have Arrived. Here's What's Working--and What's Not

WIRED

Ads in Chatbots Have Arrived. As Anthropic maintains its vow to keep ads out of Claude, the rest of the AI chatbot industry is racing to win over advertisers without antagonizing users. Anthropic is holding firm on the pledge made in its Super Bowl commercial this year to keep Claude ad-free. The company's recent IPO filing didn't even mention the possibility of selling ads to make money from the AI chatbot. It's easy to maintain the high ground with about $112 billion in investor cash.


When You Should--and Shouldn't--Use AI to Search the Web

WIRED

Chatbots are often quite good at retrieving the information you want from the web. But sometimes you'll get better results by typing your query into a good old search engine. If you've interacted with the likes of Claude, ChatGPT, and Gemini, you've noticed that in addition to everything else they do, these AI chatbots can search the web for you. Depending on what you're looking for, it can be easier to find more relevant results through a chatbot, which may explain why Google is so keen to push AI Mode for its own search engine. When asked to search the web, the latest versions of these bots are combining elements of traditional web search (they're essentially just googling on your behalf) with knowledge they are able to pull out of their training data.


How to Use AI With Your Privacy Intact

WIRED

Your conversations with AI chatbots are both highly personal and deeply vulnerable to surveillance. Here's how you can protect yourself. If the tech industry sought to create a method of seducing users into sending their deepest, most sensitive secrets to a server in a faraway data center, it would be hard-pressed to create a better honeypot than an AI chatbot . OpenAI's ChatGPT, Anthropic's Claude, Google's Gemini, and their countless smaller competitors have become therapists, sounding boards, and virtual confession booths for millions of people around the world. Almost all of those AI models are set by default to collect and store that highly private data with, in many cases, no restrictions on how it's shared or sold, used to further train the tools, or handed over to any lawsuit plaintiff or law enforcement agency that demands it through a legal process.


AI Agents Are Thirsty for Power

WIRED

Silicon Valley is shifting away from chatbot queries toward a future filled with resource-intensive agentic AI--and it's driving the data center buildout. Welcome back to Power Play! Each week, senior writer Molly Taft tackles a topic around this midterm season's biggest issue: data centers. If you've got a question or thought for the column, feel free to shoot Molly an email at [email protected] or reach them securely on Signal at mollytaft.76. "What on earth are they building all of these data centers for?" an exasperated friend asked me recently.


Why you shouldn't trust everything you find in an AI web search

Popular Science

Technology AI Why you shouldn't trust everything you find in an AI web search Your new web search option is far from infallible. Here's how to weight the answers it gives you. More information Adding us as a Preferred Source in Google by using this link indicates that you would like to see more of our content in Google News results. All the popular AI chatbots, including ChatGPT, can search the web. Breakthroughs, discoveries, and DIY tips sent six days a week.


How an Atlanta Suburb Ended Up Sharing Flock Data With More Than 2,000 Organizations

WIRED

Alpharetta, Georgia, cops share data with thousands of Flock users, ranging from federal agencies to a fish and wildlife commission. The reasons why show how vast--and invasive--the network has become. Alpharetta, a prosperous suburb of Atlanta, Georgia, is home to about 67,000 people, served by about 120 local police officers. It also hosts several dozen cameras sold by Flock Safety, the increasingly controversial surveillance company that collects license plate data and other information and makes it searchable by police. WIRED has made this article free for all to read because it is primarily based on reporting from public records requests. Please consider subscribing to support our journalism.


CCL: Causal-aware In-context Learning for Out-of-Distribution Generalization

Neural Information Processing Systems

In-context learning (ICL), a nonparametric learning method based on the knowledge of demonstration sets, has become a de facto standard for large language models (LLMs). The primary goal of ICL is to select valuable demonstration sets to enhance the performance of LLMs. Traditional ICL methods choose demonstration sets that share similar features with a given query. However, we have found that the performance of these traditional ICL approaches is limited on out-of-distribution (OOD) datasets, where the demonstration set and the query originate from different distributions. To ensure robust performance in OOD datasets, it is essential to learn causal representations that remain invariant between the source and target datasets. Inspired by causal representation learning, we propose causal-aware in-context learning (CCL). CCL captures the causal representations of a given dataset and selects demonstration sets that share similar causal features with the query. To achieve this, CCL employs a novel VAE-based causal representation learning technique. We demonstrate that CCL improves the OOD generalization performance of LLMs both theoretically and empirically.


Graph Few-Shot Learning via Adaptive Spectrum Experts and Cross-Set Distribution Calibration

Neural Information Processing Systems

Graph few-shot learning has attracted increasing attention due to its ability to rapidly adapt models to new tasks with only limited labeled nodes. Despite the remarkable progress made by existing graph few-shot learning methods, several key limitations remain. First, most current approaches rely on predefined and unified graph filters (e.g., low-pass or high-pass filters) to globally enhance or suppress node frequency signals. Such fixed spectral operations fail to account for the heterogeneity of local topological structures inherent in real-world graphs. Moreover, these methods often assume that the support and query sets are drawn from the same distribution. However, under few-shot conditions, the limited labeled data in the support set may not sufficiently capture the complex distribution of the query set, leading to suboptimal generalization.


Lookahead Routing for Large Language Models

Neural Information Processing Systems

Large language model (LLM) routers improve the efficiency of multi-model systems by directing each query to the most appropriate model while leveraging the diverse strengths of heterogeneous LLMs. Most existing approaches frame routing as a classification problem based solely on the input query. While this reduces overhead by avoiding inference across all models, it overlooks valuable information that could be gleaned from potential outputs and fails to capture implicit intent or contextual nuances that often emerge only during response generation. These limitations can result in suboptimal routing decisions, particularly for complex or ambiguous queries that require deeper semantic understanding. To address this challenge, we propose Lookahead, a routing framework that "foresees" potential model outputs by predicting their latent representations and uses these predictions to guide model selection, thus enabling more informed routing without full inference. Within this framework, we implement two approaches based on causal and masked language models. Empirical evaluations across seven public benchmarks--spanning instruction following, mathematical reasoning, and code generation--show that Lookahead consistently outperforms existing routing baselines, achieving an average performance gain of 7.7% over the state-of-the-art.


Counterfactual Image Editing with Disentangled Causal Latent Space

Neural Information Processing Systems

The process of editing an image can be naturally modeled as evaluating a counterfactual query: "What would an image look like if a particular feature had changed?" While recent advances in text-guided image editing leverage powerful pre-trained models to produce visually appealing images, they often lack counterfactual consistency - ignoring how features are causally related and how changing one may affect others. In contrast, existing causal-based editing approaches offer solid theoretical foundations and perform well in specific settings, but remain limited in scalability and often rely on labeled data. In this work, we aim to bridge the gap between causal editing and large-scale text-to-image generation through two main contributions. First, we introduce Backdoor Disentangled Causal Latent Space (BD-CLS), a new class of latent spaces that allows for the encoding of causal inductive biases. One desirable property of this latent space is that, even under weak supervision, it can be shown to exhibit counterfactual consistency. Second, and building on this result, we develop BD-CLS-Edit, an algorithm capable of learning a BD-CLS from a (non-causal) pre-trained Stable Diffusion model. This enables counterfactual image editing without retraining. Our method ensures that edits respect the causal relationships among features, even when some features are unlabeled or unprompted and the original latent space is oblivious to the environment's underlying cause-and-effect relationships.