#futureproofAfrica: AI isn't as scary as you think

2019-10-18 08:48 - Gabi Zietsman
Post a comment 0

...

Throughout human history, we've always had visions of recreating life in our image.

Jews had golems made from mud, the Greeks thought you could recreate life if you could recreate the four humours and - most famously - Frankenstein's monster was born during a time that electricity was being explored. 

WATCH: China unveils a robot-operated 'future hotel' where humans are a thing of a past

Today, in an age of incomprehensible computation power, the idea of artificial life has become ones and zeros, and with that the idea that our creation could potentially one day surpass its master. 

While AI is making headlines from experts warning that advanced computers that can think for themselves will lead to human extinction, scientists at the SingularityU South Africa Summit sees our fears more about the way we see ourselves.

"Visions of technology reflect the time and space where they originate," says Alix Rübsaam, a researcher in the philosophy of technology and posthumanism. 

Humanity is always visible in technology, and AI will still be driven by human expression - Uber isn't driven by an algorithm for where to send drivers, it's driven by the users who request rides.

"Machines replace what we do, but what we do isn't who we are," adds Rübsaam.

"The threat is only true if we define humanity as computational."

Local South African tech expert, Stafford Masie, reiterated the same beliefs, stating that the way we present humanity is sometimes too wrapped up in mathematics. 

"If you can measure it, a machine can do it better than you – human beings are built to waste time like scientific research, the arts, creativity – but through that we incur more work."

For him, the plurality of artificial intelligence augments humans, rather than diminish our own intelligence.

ALSO WATCH: You know it's bad for humans when the robots that replaced them at this Japanese hotel get fired

However, while AI as a red light on a wall in some ethereal mainframe might not exactly inspire dread, when they find physical form people are a little warier.

Synthetic humans are another avenue of AI where a living programme is given the means to physically interact with the world, something that scientist Suzanne Gildert believes will actually help achieve true AI more easily. 

Her company, Sanctuary AI, is working hard to create robots that are indistinguishable from humans, while also exploring the ethical issues that arise from creating these kinds of machines. 

One of the robots from Sanctuary AI.

"We should rather look at what they give than what they take away," says Gildert when addressing the common issue of job losses arising from automation.

She also believes these synthetic humans can be used through tele-operation - when a human controls the robot from afar to complete tasks in dangerous situations - and through it the AI can learn even more about being a better human.

But what is preventing us to create these AI powerhouses? 

"The problem is understanding the nuances of human language." They can know the definitions of words and pick the right words to speak, but they do not understand the words themselves, the emotions behind them and their context. They cannot comfortably recreate an authentic human conversation - for now at least. 

Instead of seeing AI as coming to replace humanity, instead it should be seen as augmenting our lives, possibly making us even more human - if we use it for good.

Find Your Escape by signing up for the Traveller24 Weekly Newsletter – Subscribe hereOr download the News24 App hereto receive expertly curated travel ideas and deals directly to your mobile.