Revolutionary AI Speeds Up Ad Launches with Minimal Data
In brief
- A new AI system called LLM-HYPER has been deployed to help online advertising platforms launch new promotional ads faster.
- The system uses large language models to predict how likely users are to click on an ad, even when there is little data available about the ad or its audience.
- LLM-HYPER works by analyzing text and images from ads and using a technique called few-shot Chain-of-Thought prompting.
- This allows the system to generate predictions without needing extensive training.
- In tests, the system improved a key performance metric called NDCG@10 by more than 55%.
- It also reduced the time it takes for new ads to become effective, which is especially important for e-commerce platforms.
- Researchers say this could change how companies launch new ads, making the process faster and more efficient.
- Watch for how this technology is used in other areas of advertising and marketing.
Terms in this brief
- Chain-of-Thought prompting
- A technique where an AI generates predictions by simulating a series of reasoning steps, similar to how a human would think through a problem. This allows the system to make decisions with limited data by emulating a chain of logical deductions.
- NDCG@10
- Normalized Discounted Cumulative Gain at position 10 — a metric used to evaluate the effectiveness of search or recommendation algorithms. It measures how well the top 10 results match what a user is looking for, with higher scores indicating better performance.
Read full story at arXiv CS.AI →
More briefs
AI Tech Lets Brands Disappear into the Scene
A new AI tool developed by a team including Netflix employees can insert products into TV shows more naturally than ever before. This technology uses advanced algorithms to make virtual product placements look realistic and seamless within scenes. This development could change how brands advertise on television. Instead of interrupting viewers with traditional commercials, products can appear naturally in the background or in the hands of characters. Early tests show the tool can place items with high accuracy, making it easier for content creators to integrate ads without disrupting the story. Watch for how this technology might influence future TV productions and whether it will become a common practice in the industry.
AI-Powered Tips Keep Fans Hooked on Sports Sites
A sports publisher is using artificial intelligence to keep readers interested on its website. The company, which covers high school and college sports, is testing AI tools that suggest related articles to readers. This helps keep people on the site longer and increases the chances they will read more content. This approach is becoming more common as publishers face declining traffic from search engines and social media. By focusing on keeping visitors engaged once they arrive, the publisher hopes to increase revenue. The company operates a network of more than 70 fan sites, and improving engagement could help them maintain their audience despite the challenges. Watch for more publishers to adopt similar AI-driven strategies in the coming months.
Advertisers test how chatbots shape shopping experiences
Major online retailers are using chatbots to guide customers through product choices and purchases. They want to know if these automated assistants can make shopping faster and more convenient. Some companies are running experiments to see how well chatbots understand customer needs and preferences. This matters because it could change how ads are shown and how people buy things online. Early tests suggest chatbots can help customers find products quickly. But they also raise questions about privacy and whether companies can truly understand what shoppers want. The results could influence how advertising and e-commerce develop in the coming years. Watch for new tools that help chatbots better understand customer behavior and improve the shopping experience.
OpenAI Quietly Unveils Ad Manager in ChatGPT, Lowering Barriers for Marketers
OpenAI has quietly introduced an ad manager to its ChatGPT platform, offering a new tool for businesses to experiment with AI-driven marketing. The move comes as the company continues to expand access to its beta testing program, making it easier for more advertisers to join. This integration allows marketers to craft tailored messages using ChatGPT's conversational capabilities, potentially revolutionizing how brands engage with consumers. The ad manager is part of OpenAI's ongoing efforts to explore commercial applications of its AI technology beyond chatbots. By lowering the entry barriers for its pilot program, the company aims to gather feedback from a broader range of users, which could shape future features and pricing models. Currently, costs are reportedly low—around $20 per message—but exact details remain under wraps. For developers and researchers, this development signals OpenAI's pivot toward practical, monetizable tools. It also underscores the growing competition in AI-powered advertising, with other platforms like Google and Meta racing to integrate similar features. As OpenAI continues to refine its ad manager, expect more businesses to experiment with AI-driven marketing strategies, blurring the lines between human creativity and machine efficiency.
AI Breakthroughs Turbocharge Genetic Research and Fraud Detection
In a major leap for science and technology, researchers have unveiled two groundbreaking advancements in artificial intelligence. One innovation, ERAS, has revolutionized genetic research by swiftly analyzing 1 billion biological sequences-unlocking new possibilities for understanding genetic relationships and speeding up critical discoveries. Meanwhile, SMT-AD, another AI breakthrough, has demonstrated exceptional skill in detecting anomalies, such as fraudulent transactions, with minimal computational overhead. These tools not only enhance efficiency but also pave the way for smarter, faster solutions across industries. With ERAS handling biological data at unprecedented scales and SMT-AD pinpointing anomalies with remarkable accuracy, these advancements promise to accelerate progress in fields from medicine to finance. As these technologies evolve, expect even more powerful applications that push the boundaries of what AI can achieve.