The new AI learns like a human.

The new AI-based machine learning uses technology that mimics human optic view pathways in human brains. This technology is more effective than previous, conventional, convolutional neural networks, CNN-based architecture. "A convolutional neural network (CNN) is a type of feedforward neural network that learns features via filter (or kernel) optimization. This type of deep learning network has been applied to process and make predictions from many different types of data including text, images, and audio.  "

Convolution-based networks are the de-facto standard in deep learning-based approaches to computer vision and image processing, and have only recently been replaced—in some cases—by newer deep learning architectures such as the transformer." (Wikipedia, Convolutional neural network)

The CNN network shares images to squares. And handles it with square-shaped filters. These kinds of systems are effective, but they require a large number of microchips. 

This limits their ability to detect wider patterns in fragmented or variable data. The new technology called visual transformers, ViT, is more effective. It's more flexible and accurate. However, its problem is this. ViT requires more power than CNN. CNN requires an entire data center. So the ViT requires as many data centers as it has layers. 

"In the actual brain’s visual cortex, neurons are connected broadly and smoothly around a central point, with connection strength varying gradually with distance (a, b). This spatial connectivity follows a bell-shaped curve known as a ‘Gaussian distribution,’ enabling the brain to integrate visual information not only from the center but also from the surrounding areas. In contrast, traditional Convolutional Neural Networks (CNNs) process information by having neurons focus on a fixed rectangular region (e.g., 3×3, 5×5, etc.) (c, d). CNN filters move across an image at regular intervals, extracting information in a uniform manner, which limits their ability to capture relationships between distant visual elements or respond selectively based on importance. Credit: Institute for Basic Science" (ScitechDaily, Brain-Inspired AI Learns To See Like Humans in Stunning Vision Breakthrough)

"Lp-Convolution, a novel method that uses a multivariate p-generalized normal distribution (MPND) to reshape CNN filters dynamically. Unlike traditional CNNs, which use fixed square filters, Lp-Convolution allows AI models to adapt their filter shapes, stretching horizontally or vertically based on the task, much like how the human brain selectively focuses on relevant details.

This breakthrough solves a long-standing challenge in AI research, known as the large kernel problem. Simply increasing filter sizes in CNNs (e.g., using 7×7 or larger kernels) usually does not improve performance, despite adding more parameters. Lp-Convolution overcomes this limitation by introducing flexible, biologically inspired connectivity patterns." (ScitechDaily, Brain-Inspired AI Learns To See Like Humans in Stunning Vision Breakthrough)

"Brain Inspired Design of LP Convolution

The brain processes visual information using a Gaussian-shaped connectivity structure that gradually spreads from the center outward, flexibly integrating a wide range of information. In contrast, traditional CNNs face issues where expanding the filter size dilutes information or reduces accuracy (d, e). To overcome these structural limitations, the research team developed Lp-Convolution, inspired by the brain’s connectivity (a–c). This design spatially distributes weights to preserve key information even over large receptive fields, effectively addressing the shortcomings of conventional CNNs. Credit: Institute for Basic Science" (ScitechDaily, Brain-Inspired AI Learns To See Like Humans in Stunning Vision Breakthrough)

And what makes the ViT technology so effective? The ViT means that the signal travels through multiple CNN networks. So the developers use multiple layers of the CNN networks. The system can use the expanding ViT model. There the optical signal travels first in the small CNN layer. Then the CNN layer's size expands. And then it contracts. That makes the CNN layers size or the number of processors. That participates in the operation that looks like the Gauss curve. 

The system can have two CNN layers that play the ping-pong ball with data. Every turn when one of those two layers sends information to the other the other layer uses more power to that problem. Then the system focuses data on one point. Or in the linear model, the system can use multiple layers of the CNNs. That model boosts machine learning but it requires more electricity and enormous data mass. 

The ability to use multiple neural layers to analyze information is the thing that makes ViTs so effective. The thing is that the ViT systems require lots of space. That means they can control robots through the internet. The other version is that millions of compact-size robots can turn them into the ViT network. 

https://scitechdaily.com/brain-inspired-ai-learns-to-see-like-humans-in-stunning-vision-breakthrough/

https://en.wikipedia.org/wiki/Convolutional_neural_network