Clean Data, Sustainable AI Take Centre Stage at QUEST-AI Conference
Experts at QUEST-AI conference highlight need for clean data, caution on synthetic data, and discuss quantum-AI integration.
The quality of data, rather than its sheer volume, will determine the reliability of future artificial intelligence systems, according to experts speaking at the international conference on Quantum Enhanced Sustainable Technologies for AI Systems (QUEST-AI), organised by Andhra University Engineering College.
Nilanjan Dey, Professor at Techno International New Town, told the gathering that machine-learning algorithms are fundamentally dependent on the data they are trained on. He argued that poor-quality data can lead to flawed outcomes, making it essential to prioritise clean and diverse datasets over the current obsession with 'big data'. He also drew attention to the importance of 'small data' scenarios, where information is scarce and requires careful handling.
Addressing the growing use of synthetic data, Prof. Dey acknowledged its utility when real-world information is limited. However, he cautioned that excessive reliance on it could increase errors in large language models (LLMs), and stressed that its application must be governed by clearly defined limits.
On the conference's concluding day, S. Praveen Kumar, a senior scientist at BARC, offered a glimpse into the future of computing. He predicted that quantum technology and AI would increasingly work in tandem, potentially giving rise to 'quantum-enhanced AI data centres'. He also flagged the substantial electricity demands of such facilities, suggesting nuclear power as a viable option to meet these future energy needs.
Other speakers included Greg Skulmoski of Bond University, Australia, who emphasised the importance of translating university research into practical, real-world applications with tangible business value. NVSN Sharma of NIT-Warangal also presented on Terahertz technology, detailing its potential applications in defence and healthcare sectors.