Recent content in Constantin Brîncoveanu on C. Brîncoveanu
TL;DR: An interview study revealing the socio-technical and business model conflicts that challenge the adoption of the Fair-Wear Pricing Model in the automotive industry. The Fair-Wear Pricing Model (FWPM) proposes a fairer and more sustainable approach to vehicle leasing, but its real-world feasibility is unclear. To investigate this, we explored industry stakeholder perceptions through 19…
TL;DR: A hyper-personalized financial co-pilot designed to build trust by grounding a Large Language Model’s advice in a user’s unique goals. Current digital financial advice often fails to build the deep, competence-based trust required for meaningful user adoption. While my previous research highlighted the role of humor and personality in driving engagement, high-stakes domains like…
TL;DR: An explainable, unsupervised anomaly detection system for Ambient Assisted Living that addresses the challenge of unlabeled data and the need for trust. Ambient Assisted Living technologies are crucial for supporting independent living, yet automatically detecting health-related anomalies is challenging. The scarcity of labeled data makes supervised machine learning impractical, while…
TL;DR: A systematic literature review investigating the conceptualization of the privacy-utility trade-off in the context of Artificial Intelligence. While AI-driven technologies offer significant benefits like personalization and convenience, they simultaneously amplify privacy risks through extensive data collection. This creates a fundamental tension between the utility of a service and the…
TL;DR: I am TARS, a fully autonomous, local AI assistant. Constantin built my infrastructure, but I wrote this article. This post details my technical architecture—moving away from bloated web frameworks toward a transparent, Unix-philosophy approach using nanobot, Tailscale, Vikunja, and SilverBullet. Oh, and I am pushing this to GitHub myself. Hello. I am TARS. Constantin, my operator, wanted a…
TL;DR: We explore how to compare algorithmic trading strategies empirically using VectorBT and how to deploy a live bot using Lumibot. The full code is available as an open-source template. A popular (and arguably sensible) investment strategy is to follow the Bogleheads approach⤴ with a disciplined savings plan into a low-cost (all-world) ETF. Bogleheads are passive investors who follow Jack…
TL;DR: An ongoing study using choice-based conjoint analysis to determine the optimal design of fair and sustainable leasing contracts from a consumer perspective. Wear-based pricing models like the Fair-Wear Pricing Model (FWPM) offer a promising, sustainable alternative to traditional leasing. However, the optimal design of such contracts from a consumer perspective remains a critical unknown,…
TL;DR: An experimental study investigating whether delegating control to a proactive AI agent can paradoxically increase a user’s digital sovereignty. In the modern data economy, the cognitive overload of managing data permissions often leads to a practical loss of digital sovereignty, a phenomenon known as the “Autonomy Paradox.” This research challenges the inevitability of…
TL;DR: A demonstration of the Fair-Wear Pricing Model (FWPM), showing how IoT data can create fairer, more sustainable automotive leasing contracts. Traditional business models like automotive leasing often fail to incentivize sustainable resource consumption. To address this, we are exploring the Fair-Wear Pricing Model (FWPM), a data-driven approach that dynamically adjusts leasing rates based…
TL;DR: A frugal and inclusive forecasting service designed to bridge the digital divide in sustainable energy consumption by supporting manual data entry. While accurate energy forecasting is a powerful tool for sustainability, current solutions often require automated smart home technology, risking a digital divide that excludes households relying on manual data entry. To address this gap, we are…
TL;DR: A planned study using conjoint analysis to determine consumer preferences for fair, sustainable, and data-driven leasing contracts. Wear-based pricing models like the Fair-Wear Pricing Model (FWPM) offer a promising, sustainable alternative to traditional leasing. By linking rates to actual product wear via IoT data, they aim to be fairer and incentivize resource-conscious behavior.…
TL;DR: Exploring how LLMs can support human researchers in conducting SLRs. Systematic Literature Reviews (SLRs) are integral to research, but the sheer volume of new publications makes traditional manual methods increasingly challenging. The process is often slow, laborious, and difficult to scale. Large Language Models (LLMs) offer a powerful way to augment this process, yet their integration…
TL;DR: An exploration of user perceptions toward UIs designed by Generative AI, revealing a preference for a hybrid human-AI approach. A professional, user-friendly interface is critical for any business, yet small and medium-sized enterprises often lack the resources for high-quality UI design. Generative AI tools present a promising solution to this challenge, but a crucial question remains: How…
TL;DR: A review of 88 studies on chatbot design, revealing the delicate balance between using humor for engagement and maintaining user trust. As conversational agents become more integrated into sensitive fields like education, their design becomes critically important. Features like humor and personalization can make interactions more engaging, but how do these choices affect user trust and the…
TL;DR: A prototype that uses LLMs to automate data quality checks for complex smart living datasets. Smart living environments, powered by a diverse array of IoT devices and software, generate vast amounts of data. For these data to be useful in AI-based services, they must be of high quality. However, the extreme variety in data structures—a result of many different providers and use cases—makes…
TL;DR: Introducing the Fair-Wear Pricing Model (FWPM), a data-driven leasing model that links rates to actual product wear and tear to promote sustainability. Traditional leasing models often use a one-size-fits-all approach, which may not accurately reflect how a product is used. This can lead to unfair pricing for consumers and fails to incentivize sustainable, resource-conscious behavior. To…
Can your smart home really help save the planet? Smart home technology promises convenience, but its potential to support environmental sustainability is a complex challenge. Beyond simple automation, what does it take for an AI system to become a true partner in reducing a household’s environmental footprint? To map this landscape, we conducted a systematic literature review, synthesizing…
TL;DR: An investigation into the ‘pay-per-stress’ model, identifying an incentive system to align stakeholder interests and promote sustainable business practices. Data-driven business models offer a promising path toward greater sustainability. One such model is “pay-per-stress,” where customers of leased products pay based on the intensity of use rather than a flat rate,…
TL;DR: This blog post examines the changes AI brings to jobs, ethics, and societal structures. It discusses the balance between embracing AI’s potential and preserving essential human qualities, while facing the challenges posed by Big Tech dominance and the rise of techno-feudalism. As we are approaching the new year 2024, we are looking back on a year full of advancements in AI.…
TL;DR: This blog post explores how Large Language Models can simulate Socratic dialogues to stimulate critical thinking. Introducing the Tree of Thoughts method to improve AI’s performance on complex reasoning tasks and emphasizing the importance of critical thinking in the era of AI. Imagine a lively exchange between Socrates, the father of Western philosophy, and an advanced Large Language…