Accidental impacts of emerging tech could be ‘profound’, Davos panel warns
Image credit: Dreamstime
Leading figures in tech and innovation gathered at the World Economic Forum (WEF) in Davos, Switzerland, concluded that allowing immature technologies out in the real world before they are fully understood could have devastating consequences.
At the Davos panel discussion, Future Shocks: Rogue Technology, the five speakers discussed their greatest environmental concerns, including plastic pollution of the oceans and the role that emerging technologies had in aggravating or reversing this destruction. This followed the publication of a WEF report which warned of the potential environmental damage that could be caused by bioengineering and other rapid technological advances.
Technologies such as CRISPR gene editing technology – which allows for engineers to precisely edit DNA – could be used to introduce adaptations to help threatened species survive in challenging conditions. This could be extended to entire environments, such as coral reefs. Meanwhile, drones are extensively used in conservation efforts.
While these technologies hold enormous promise for reversing some of the damage done to the environment by humans – and by humans with the help of some technologies – they could also be exploited to cause further destruction.
Marc Benioff, chairman and CEO of Salesforce, commented that autonomous deep sea mining was an example of how sophisticated computer vision and machine-learning software could be used to knowingly cause damage to the environment. However, Professor Mary Cummings, director of the humans and autonomy lab at Duke University, said that the unintended consequences of letting loose immature technologies could be far more damaging than we realise.
“My concerns aren’t really with the malevolent use of these technologies – the intentional malevolent use. It’s with the unintentional malevolent use,” Cummings said. “AI is definitely opening up a Pandora’s Box. Most applications of AI, particularly when it comes to autonomous vehicles, we do not understand how the algorithms work, so I’m much more concerned about the accidental damage that may be profound, as opposed to malevolent damage.”
Only recently, it was discovered that driverless cars could be fooled by placing simple stickers on road signs, which may not trick humans. For instance, a “STOP” sign could be subtly made to appear like a speed limit sign.
“As a researcher, what I worry about is [that] we’re still finding about the emergent properties of these technologies – CRISPR, AI – yet there are many companies and agencies that want to take these technologies and start deploying them in the real world, but it’s still so nascent that we’re not really sure what we’re doing,” she said.
“I do think [there] needs to be more of a collaborative arrangement between academics and government and companies to understand what’s really mature and what’s very experimental.”
Professor Feng Zhang of MIT – who played a central role in the development of CRISPR technology – agreed with Cummings, stating that in bioengineering, it is crucial to understand the potentially vast consequences of even the smallest changes being made to a population.
“When we’re engineering organisms […] I think we have to be very careful and proceed with a lot of caution,” he said, adding that researchers must suggest containment mechanisms to recall a strategy if it turns out to have dangerous consequences.
According to Benioff, even with far more “mature” technologies, we are seeing unintended consequences emerge long after the genie has left the bottle. In recent months, he commented, there has been a backlash against social media companies for failing to prevent their platforms being used to spread misinformation and abuse.
Governments are caught in a bind when it comes to regulating the products of the fourth industrial revolution, said Marcos Souza, the Brazilian Secretary of Innovation, due to the sheer speed with which they are developing. For instance, Brazilian regulators are struggling to review and approve new biotechnology products quickly enough. As a result, the adoption of new technologies is being delayed as regulators try to understand the potential impacts of the technologies.
“The characteristic of this fourth industrial revolution is the speed of the technology advancing and, as you know, the government regulation is always behind this […] speed, so it’s a challenge for us. The previous revolutions took longer so we could prepare those regulations properly, but this is going too fast,” said Souza.
An important part of the solution is to have international oversight, particularly with regards to ethical guidance, said Peter Thomson, UN Special Envoy for the Oceans. This will prove valuable in ensuring that technological development does not aggravate inequality or destroy shared environments, such as the oceans and space.
“Globally, I think this is definitely lacking at the minute and is needed: a global discussion on the ethics of AI, genetic manipulation and so on,” Thomson said.