Connect with us

Hi, what are you looking for?

Tech & Science

Op-Ed: Real time memory nano neural networks work out with University of Sydney and UCLA

It’s a long way to Tipperary and to sentience, but this is where it can happen.

Neuroscience -Tomography. — © AFP
Neuroscience -Tomography. — © AFP

Neuroscience is aptly named. It’s a science that gets on your nerves. The latest effort is pretty interesting, though. University of Sydney and UCLA have put together a network of overlapping nanowires similar to the layout of human synapses. This is extremely important core science.

What’s super-significant about this is a new real time dynamic.  Acquisition and retention of information are virtually simultaneous. That makes real time functions a lot more agile and situationally “aware”. In comparison to the ancient “input–process–output” of traditional data management, that’s a huge leap in the right direction.

That situational awareness is the key. Real time includes peripheral awareness of complex environments, and you can’t do that without a trustworthy system of information acquisition.

This is an absolutely basic step in the development of high-performance systems. It’s not avoidable. I’m wondering how this would work with organic systems like Dishbrain. I think it could. They’re not in conflict. They’re not “alternatives to each other”.  Organic systems being plugged into electromechanical systems does make sense.

The point here is that they are very different systems. This new system is far more attuned to organic processes. Any organic neural element like a neuron would sooner or later have to interact with electromechanical systems, and vice versa. Best of both worlds, maybe.

Note: Duplicating synapses doesn’t mean “sensate”. Neurons are far more complex on nano levels. What’s the bet that electronic systems are developed to mimic organic systems? Even money, I’d say.

Speaking of interactions, it’s long overdue that nano got a word in on these neural networks. In terms of spatial efficiency alone, it’s essential. In terms of energy, it’s a no-brainer. The human brain works on a tiny fraction of the energy of neural networks. They’re comparatively physically inefficient and clumsy by comparison. The minimum possible result of this new neural network layout is drastically improved efficiency in functionality.

The new nanosystem could also make hardware infinitely easier to work with. In lieu of a superconductor, this is a workaround in so many ways. You could also ask what’s the point of a superconductor if the system then slows things to a crawl, too, but that can wait.

It’s expected to improve memory efficiencies and usage dynamics, probably with a bit of custom tweaking for specifics. This is the beginning of a lot of hard work, but it will be worth it.

The obvious ramification for this tech is merging with AI. More efficient systems will create much stronger AI. It’s a long way to Tipperary and to sentience, but this is where it can happen.

____________________________________________________________

Disclaimer
The opinions expressed in this Op-Ed are those of the author. They do not purport to reflect the opinions or views of the Digital Journal or its members.

Digital Journal
Written By

Editor-at-Large based in Sydney, Australia.

You may also like:

Social Media

But like TikTok, ByteDance's AI services could face trouble in overseas markets owing to issues from data privacy.

Entertainment

It is another setback for the world's biggest live televised music event, after five countries pulled out of this year's contest.

World

A SpaceX Falcon 9 rocket with the company's Dragon spacecraft on top launches from Cape Canaveral, en route to the International Space Station -...

World

The cigar festival typically raises millions of dollars from auction sales that are then funneled into the country's health care system.