support@headlinenews360.com
Last Updated:
AI, despite its advanced capabilities, struggles with basic tasks like telling time and calendar calculations, as revealed by a study led by Rohit Saxena from the University of Edinburgh.
AI struggles with basic tasks like telling time and calendar calculations.(Shutterstock)
Artificial Intelligence (AI) has long been hailed as a transformative force, with the potential to surpass human intelligence and dominate many aspects of society. AI models, such as chatbots, have demonstrated their ability to solve complex problems, pass demanding exams, and even engage in high-level reasoning. However, a groundbreaking study uncovered some unexpected flaws in AI systems – limitations that could significantly affect their functionality.
This research, conducted by a team of scientists led by Rohit Saxena from the School of Informatics at the University of Edinburgh, Scotland, was recently published in Nature and stunned experts in the field. The study highlights that many AI models, despite their advanced capabilities, are unable to perform simple, everyday tasks- specifically, telling time and answering questions related to calendars.
In one of the most surprising findings, the study revealed that certain AI models fail to tell the correct time when asked to read traditional analog clocks. These AI systems, which are adept at solving sophisticated puzzles and understanding complex data, struggled with the fundamental task of interpreting a clock with hands – something that humans typically learn as children.
The issue persisted across various clock designs, including those with Roman numerals, watches lacking a second hand, and clocks with differently colored dials. Despite their capacity for solving intricate problems, these AI systems couldn’t reliably answer even the simplest questions about time.
The research also exposed significant issues in AI’s ability to manage calendar-related questions. When tested on their ability to calculate past or future dates, the AI models failed to provide accurate responses. This included determining the days of the week for historical dates or projecting future days based on a given timeline. Calendar computations, which humans accomplish without much thought, proved to be a hurdle for these AI systems, revealing that even seemingly basic tasks are beyond their grasp.
Saxena and his team emphasised that these flaws point to a broader limitation in current AI models, one that could undermine their broader applicability in real-world scenarios. The researchers argued that for AI to reach its full potential and serve as a reliable tool in everyday life, these basic deficiencies must be addressed.
Can AI Truly Replace Humans?
While AI has made significant strides in fields such as language processing, data analysis, and problem-solving, this study raises important questions about its capabilities and limitations. The idea of AI replacing humans in various jobs – ranging from customer service to mental health care – has become a topic of heated debate. Some believe that AI’s potential to surpass human intelligence could render certain professions obsolete. However, a separate study from Switzerland has cast doubt on this notion, particularly in the realm of mental health.
This study focused on AI chatbots, investigating whether they could replace human therapists or doctors. The findings revealed that while AI chatbots, such as OpenAI’s ChatGPT, could simulate human-like conversations, they also exhibited signs of anxiety when placed in highly stressful or traumatic scenarios.
AI systems seemed to experience stress, mirroring human emotional responses, and their anxiety diminished when combined with mental-calming exercises. This raises serious concerns about whether AI can adequately provide mental health advice or support, given its emotional vulnerabilities.
AI’s Future and the Need for Improvement
The limitations discovered in the recent study, including AI’s struggle with time-telling and calendar calculations, serve as a reminder that even the most advanced AI systems still have significant room for improvement. As AI continues to evolve, addressing these basic flaws will be critical if it is to fulfill its promise as an indispensable tool for society.
