GENERATIVE ARTIFICIAL INTELLIGENCE AS A COGNITIVE CO-PILOT: THE IMPACT OF HIERARCHICAL GOAL DECOMPOSITION IN THE BLOCK AI APPLICATION
TL;DR
Practical testing shows that dividing goals into bite sized tasks, paired with built in timers and haptic feedback, reduces procrastination and improves daily user engagement, which shows strong promise for next generation personal productivity systems.
Abstract
Managing personal productivity has become increasingly difficult under the pressure of daily information overload. This study presents the design and evaluation of Block AI, a mobile application that links hierarchical goal decomposition with generative artificial intelligence. Grounded in Locke and Latham’s (1990) Goal Setting Theory and Allen’s (2016) Getting Things Done (GTD) framework, the app uses a three tier model Objectives, Goals, and Tasks to turn abstract ambitions into daily routines. Its generative AI engine runs on large language models (LLMs) through two modes: Creator, which outlines structured action plans from raw text descriptions, and Coach, which processes real time performance metrics to deliver contextual productivity advice. Methodologically, the work is an applied technological development utilizing React Native, Supabase, and the Gemini model. Practical testing shows that dividing goals into bite sized tasks, paired with built in timers and haptic feedback, reduces procrastination and improves daily user engagement. Ultimately, merging traditional time management principles with generative models shows strong promise for next generation personal productivity systems.