We don’t know for sure if this is going to end up being the “Skynet” moment for the human race  (obviously in reference to the Terminator franchise). Only time can answer that. Nonetheless, this is potentially a big leap forward in the development of artificial intelligence (AI), or more specifically, artificial general intelligence.

AI using new artificial synapses

An international group of scientists has announced that they invented a new type of synthetic synapse for AI systems that use the neural network model. In an average artificial neural network, sophisticated computer systems are developed to emulate the functions of the human brain. To accomplish its objective, the system uses digital neurons and synapses whose modus operandi resembles that of their biological counterparts.

Simply put, the synapses operate as gateways for neurons. Be they biological or synthetic, the purpose of synapses is to transmit information to one another. According to some estimates, the average human brain contains as many as 100 trillion synapses.

While scientists have had managed to achieve remarkable success with AI networks over the past decade or so, today’s AI systems still operate at half throttle. That’s because unlike the mammalian brain in which synapses accommodate two distinct varieties of signals – inhibitory and excitatory – artificial synapses so far can process only a single type of signal at a given time.

But the new invention alters that status quo.

The team comprising scientists from the US and China created a unique type of synthetic synapses that handles both varieties of signals. The study, originally published in the journal ACS Nano, further states that because of their ability to process both the signals simultaneously, the AI “brain” can reconfigure it much more efficiently.

“These new artificial synapses allow the same synapse to be reconfigured into either excitatory or inhibitory modes, which was not previously possible in solid state artificial synaptic devices,”  Han Wang, co-author of the study, explained.

The team now hopes that this breakthrough invention will play a critical role in the emergence of heightened cognitive abilities in next-generation AI systems.

“This new functional flexibility is important for enabling more complex artificial neural network that can also dynamically reconfigure just like our brain does.”

What’s your take on this new development? Do we think scientists are playing with fire by developing increasingly sophisticated AI systems without any safety protocols officially in place? Let us know in the comments below.

  • Greff CallCo

    Hype. AI will always be fragile and barely useful.