Deep Learning Limitations and Flaws

Abstract

With today's growing interest toward Artificial Intelligence (AI) and its augmentation as part of integrated business from banking to eCommerce, medical applications and others, we are getting more and more dependency on AI in our day to day operations. However, the most sophisticated AI or Super AI (SAI) still needs to rely on its two other integrated sub-sets of components namely, Machine Learning (ML) and Deep Learning (DL). However, there certain limitation and flaws that exists within DL component of AI or SAI that will cause and error to grow way beyond control and will impact its main master component namely AI and SAI for its final processing of data and information in a trusted way of process of the precise decision making and prediction of event including any forecasting as part of its Use Case (UC) and Service Level Agreement (SLA) assigned as task to the AI or SAI. This article points few of these limitations and flaws that presently are concerns of the scientists and engineers behind the artificial intelligence technologies momentum. These sorts of limitation and flaws also would impact the Business Resilience System (BRS) from perspective of resiliency built into your daily business operations that is also pointed it out in this article.

Searching for the ground state of complex spin-ice systems using deep learning techniques

“…Our method shares general limitations and drawbacks existing in common deep learning algorithms, such as the requirement of a big dataset and difficulty in explaining how the trained machine obtains the solution 38 . Nevertheless, we believe that this method can transcend the performance of conventional methods in the problems of searching for the ground state of various spin-ice systems, particularly those with novel complex tiling patterns, intractable interaction terms, and large sizes.…”

AI in medical research

“…Det skyldes, at AI er baseret på eksisterende data, og svarene er bundet til disse datasaet. Selvom AI kan udvikle sig med input af nye data, er teknologien ikke fejlfri [25]. Ud over dette oplyser skaberne af chatbotten ChatGPT, at botten har begraensninger.…”

“…Desuden forstår brugeren ikke fuldt ud, hvordan AI genererer sine svar pga. det komplekse samspil mellem de forskellige lag i systemet[25]. Dette kan også ses som en begraensning, fordi vi ikke ved, om botten genererer nye og originale svar, hver gang et spørgsmål stilles, eller om den blot gengiver tidligere laert information.…”

Mobile Networks and Internet of Things Infrastructures to Characterize Smart Human Mobility

“…Even when they are available, labeling efforts require massive human resources. Furthermore, it can be difficult to know how a mathematical model trained with deep learning arrives at a prediction, recommendation, or decision [60]. Therefore, we present other two emerging solutions for scholars of human mobility prediction: gravity model (GM) and spatial equilibrium model (SEM).…”