TLDR: CareerPooler is an AI-powered system that uses a pool-table metaphor to simulate career exploration, offering a spatial and narrative interaction instead of traditional chat interfaces. It allows users to strike balls representing career events, with hints and collisions mirroring real-world decision-making under uncertainty. A study found CareerPooler significantly improved user engagement, satisfaction, and career clarity compared to a chatbot, fostering experience-based learning and resilience through realistic setbacks. The research highlights the potential of visually grounded metaphorical interactions for more engaging and satisfying AI-assisted career guidance.
Navigating a career path can often feel like trying to solve a complex puzzle with missing pieces. It’s a journey filled with uncertainty, requiring decisions made with limited information and unpredictable outcomes. While artificial intelligence (AI) has opened new doors for career guidance, many existing systems, often relying on simple chat interfaces, tend to offer suggestions that are too broad, overly optimistic, and fail to capture the real-world complexities and non-linear nature of career development.
A new research paper introduces an innovative solution called CareerPooler, an AI-powered system that reimagines career exploration through the engaging metaphor of a pool table. Instead of a linear chat, CareerPooler simulates career development as a dynamic, spatial, and narrative interaction. Users interact with ‘balls’ on a virtual pool table, each representing different career events such as milestones, skills, or unexpected random occurrences. The act of striking these balls, observing hints, and experiencing collisions and rebounds, mirrors the process of making decisions under uncertainty in a career.
The creators of CareerPooler, Ziyi Wang, Ziwen Zeng, Yuan Li, and Zijian Ding, conducted a study with 24 participants to compare their system against a traditional chatbot. The results were compelling: CareerPooler significantly boosted user engagement, satisfaction, and clarity about their career paths. Qualitative feedback from participants highlighted that this unique spatial-narrative interaction fostered experience-based learning, helped build resilience in the face of simulated setbacks, and reduced the psychological pressure often associated with career planning.
The system’s design is built on three core goals: to create an engaging and satisfying experience for early-career individuals, to simulate real-world uncertainty through interactive spatial elements and partial information, and to generate realistic and personalized experiences that improve career clarity. CareerPooler achieves this by allowing users to invest ‘time’ by aiming and striking event balls. The drag distance of the cue ball corresponds to days invested, and the unpredictable collisions and rebounds symbolize unexpected real-world outcomes. The interface features a dual-panel layout, with the pool table on one side for action and a timeline and event log on the other for reflection. Hovering over skill or random event balls provides vague hints, maintaining the element of uncertainty while still offering some foresight.
The event generation pipeline uses large language and vision models to create structured and coherent events, ensuring a balance of positive, negative, and neutral outcomes tailored to the user’s background and goals. This approach helps to counteract the common issue of AI models generating overly successful or idealized trajectories due to biases in their training data.
CareerPooler’s success points to several important design implications for future AI-powered career exploration tools. It suggests that incorporating experience-based learning alongside conversational interfaces can lead to greater engagement and information retention. Using spatial metaphors, like the pool table, can effectively ground abstract concepts, making complex decision-making processes more tangible. The system also demonstrates the value of progressive information disclosure, breaking down complex information into manageable, contextually relevant segments to prevent cognitive overload. However, the study also highlighted areas for improvement, such as ensuring temporal consistency in AI-generated narratives, addressing user privacy concerns, and exploring long-term engagement strategies.
Also Read:
- Navigating AI in Education: How Conversational Styles Impact Instructor Learning
- Guiding AI: A Human-Centered Approach to Web Browsing Agents
Ultimately, CareerPooler offers a fresh perspective on how AI can support personal development, transforming the often daunting task of career planning into an interactive, insightful, and even enjoyable experience. You can read the full research paper at arxiv.org/pdf/2509.11461.
