The REAL future of AI has just been invented!

Biology has been a key obsession of mine since I was at university, and I still remember the wonder I felt when I learned how a synapse [in the brain] actually works. For those who don’t know, it’s an incredible dance between electrical energy and chemical modulators. And it’s those chemical modulators that have always made me poo poo the current version of AI. A neural network built of nothing but electrical energy can never truly approximate the insanely efficient working of the mammalian brain. The size of the current AI data centres is proof of that.

So, what’s changed? This:

“Scientists have built an artificial neuron that’s so realistic it fires, learns, and responds to chemical signals just like the real thing – a breakthrough that could transform computing, medicine, and the way that tech merges with biology.” Quote from a New Atlas article.

This, my friends, is the first step towards real AI because it allows artificial neurons to function like our own, biological neurons:

“The researchers then added chemical sensors that could detect ions, such as sodium, and neurotransmitters like dopamine. These sensors changed the circuit’s electrical properties in response, mimicking how real neurons adjust their behavior based on chemical signals in their environment; a process known as neuromodulation.” New Atlas.

Let me repeat that:

‘…real neurons adjust their behavior based on chemical signals in their environment…’

Those chemical signals help determine the strength of the electrical signal that is sent on to other neurons. They also help determine which other neurons receive that signal. And it all happens in the synapse.

You can think of the synapse as a lake. Electrical signals arrive on the sure and are picked up by chemical gatekeepers that determine how much will be ferried across to multiple ‘docks’ on the other side. This biological neural activity is ‘analogue’ and helps give human brain cells their insane power and variability.

Let me illustrate the power of variation using a graphical example:

That photo of the kitten’s eye is made up of squillions of pixels – the tiny square containers of colour seen on the super enlarged image on the right. Pixels are digital objects that are used to approximate what we see, or what an old analogue camera used to capture. The important part is that every single pixel is a slightly different shade of colour. Taken together those millions of different shades of colour combine to form a single image that is unique to that moment. The kitten blinks and the angle of the light hitting its eye changes. That, in turn, changes the array of colours that become visible in the eye. Thus, no two photos can ever be exactly the same.

Now, imagine that the chemicals that ‘adjust the behaviour of neurons’ are like the colours in that photo. See what I mean about variability? It is virtually infinite. Oh, and those chemical modulators? They also allow living creatures to feel.

And now artificial neurons will finally come close to what a human brain cell can do. This is the big breakthrough, not the Large Language Models currently impersonating AI. But don’t take my word for it. Read the whole article:

https://newatlas.com/medical-tech/artificial-neuron-mimic-reality-bioelectronics/?utm_source=newsletter.newatlas.com&utm_medium=newsletter&utm_campaign=a-tasty-flu-test&_bhlid=33ee4985f78d8575890c9ca80675f5ea1070255e

Is it a good thing that researchers have worked out how to create ‘wetware’ for artificial intelligences?

I guess only time, and the profit motive, will answer that question. Personally, I hope that all the current Tech Bros go bankrupt or end up in jail before this truly innovative invention becomes commercially viable. It’s too important to leave to the profit motive.

And just for the record, have a look at the two researchers responsible for this breakthrough:

Shuai FuJun Yao

Is it really a coincidence that both lead researchers are Chinese?

Meeks