Chart, page-122

For those that are having second thoughts about BrainChip, ie whether to hold'em or fold'em here are some comments by  Sam Volkering from Southbank Investment Research that he made about BRN earlier this year

1 " Perhaps the most important company involved in AI right now in my view is BrainChip Holdings [ASX:BRN].  This company, one of the most exciting companies in the world."

2  BrainChip is developing proprietary technology that is the foundation of true artificial intelligence.

3  Its technology is called the Spiking Neuron Adaptive Processor (SNAP). What’s important to note about the SNAP is that it is hardware. It is not advanced software, or just a bunch of complex algorithms. It is not deep learning. And it is not machine learning.

4  The SNAP is like the neurons in your brain. It is the foundation of a neural network and can learn unsupervised. It is the beginning of  real artificial intelligence.

5 Like Turing theorized about a child-machine, the SNAP learns by experience. It can receive input and then process the input itself, figuring out what information it is receiving and then categorizing that information without human programming or any other input.
This might sound a little complicated. So let me explain how this works with a real example used by BrainChip to prove that the SNAP can learn unsupervised.

6 That’s right. BrainChip and the SNAP achieved unsupervised learning. The SNAP was hooked up to an image sensor that was set over a highway overpass in California. This sensor looked straight down at the highway below. As you’d expect, the sensor
fed vision data back to the SNAP neural network.

7 Now here’s the critical part – the SNAP had no previous knowledge of the information it was receiving. There was no previous input to say, “This is a camera over a freeway.” It was not programmed to identify cars. It had nothing by way of human input to help it learn.
BrainChip called this development Autonomous Visual Feature Extraction (AVFE). The video feed ran for 78.5 seconds, recording approximately 66,100 events per second. That’s over 5.18 million events the SNAP had to process in this short test.

8 According to BrainChip’s market announcement: The SNAP spiking neural network learned the features of vehicles passing by the sensor within seconds. It detected and started counting cars in real time.

9 As the company’s chief technology officer (and inventor of the SNAP) Peter van der Made said, “It shows that BrainChips neural processor SNAP acquires information and learns without human supervision from visual input.”

10 I can’t stress enough how significant the development of unsupervised learning is. The time is right to buy into the only company in the world that’s on the brink of real AI."

​