AI processes text based on probability, predicting the next word based on the preceding ones to form coherent sentences, which makes it more challenging
FinTech BizNews Service
Mumbai, July 18, 2024: Which type of data do you think is easiest for AI to process: text, images, video, or audio?
Surprisingly, the answer is images. Each pixel of an image can be converted into a colour code and flattened and processed, allowing AI to read and interpret each pixel to understand the entire image. Videos, being a series of images in rapid succession, are processed in a similar way.
These insights were part of an engaging AI Masterclass recently hosted for the Reserve Bank Innovation Hub (RBIH) team. RBIH firmly believes that sharing knowledge is key to fostering a culture of collaborative innovation, and what better place to start than within its own team!
For audio, AI converts it into a spectrogram—a visual representation of sound frequencies—making it easier to analyse.
Text, on the other hand, is more complex. AI processes text based on probability, predicting the next word based on the preceding ones to form coherent sentences, which makes it more challenging.
