A few years back, many of the world’s top scientists and industry leaders like Elon Musk and Stephen Hawking came together and wrote a letter warning about the dangers of artificial intelligence.
Among the things that they pointed out was that humans could not possibly compete with AI on so many levels. AI, they said, posed an existential risk to humans.
What Is Artificial General Intelligence (AGI)
The kind of AI that poses the greatest threat to humans and their way of life is what computer scientists like to call artificial general intelligence or AGI. This kind of intelligence is one that is capable of thinking like humans or even surpassing their ability to generate new knowledge. It will be able to have goals, and these goals need not necessarily be aligned with those of humanity. An AGI might even have a warped theory of morality that is totally different from what humans would expect. Worse, it might even consider us as inferior- as beings in need of replacement.
While we are far from creating an AGI, that doesn’t mean that AI has not already acquired capabilities that pose some serious threats to our way of life. Here is a list of 5 AI capabilities that already pose a threat to humanity.
1. Deep Fakes
After the 2016 US presidential election, there was a big fuss about the role that fake news played in the election of Donald Trump. The revelations that Russian bots might have influenced the election made headlines all over the world and showed how vulnerable democracies are to manipulation.
“But the 2016 US presidential election, while worrying was perhaps just a sign of things to come. Deep fakes powered by AI are blurring the lines of what is real and what is not.”
What Is a Deep Fake?
You might have seen videos on YouTube of President Obama making speeches that he never made. These videos are generated using human image and voice synthesis techniques that superimpose existing images and videos onto source images and videos with the help of AI.
In other words, all you need to create a fake Obama video, in this case, is original footage from which one can synthesize what an Obama video looks like. These fake videos can then be widely shared with serious consequences to law and order.
Although deep fakes are yet to be deployed in a massively disruptive way, they are all over the internet now. There is even one of Mark Zuckerberg (Facebook’s founder) that is trending online at the moment showing him saying some disturbing things about privacy.
We will have to wait and see what role deep fake videos and sound bites will play in future wars and propaganda.
2. Autonomous Weapons
One of the things that the letter we were talking about earlier called for was the banning of autonomous weapons. Militaries around the world are deeply involved in the research and development of autonomous weapons. These are weapons that can make assessments and hit targets without any human input.
The danger with this capability as pointed out in the letter is that it will make it easier for politicians to start wars because the human cost (the death of soldiers) will be significantly reduced.
And while it is not a bad thing to save the lives of soldiers and to win wars more quickly, it raises serious moral issues when machines decide who gets to live or not on the battlefield.
There is also the risk that such machines may be hacked and we are not sure that our current anti-malware solutions are up to the task of preventing autonomous weapons from being infiltrated by rogue agents.
3. Privacy Violations
Every day, we hear of some of the big internet companies, be it Google, Facebook, Amazon, or Twitter, being rocked by some privacy scandals. This is because these companies collect so much useful information that they are constantly experimenting with what to do with the data.
“The companies also rely heavily on AI to develop insights on consumers and to sell the same data to marketers who needless to say also rely heavily on AI.”
The privacy violations present an existential threat to humanity because they are just one step from a dystopian future where no one is free from total surveillance.
But it is not just the big corporations that are interested in your data. Governments, hackers, and spy agencies are also looking for ways to tap into the wealth of information that is generated by the millions of people going about their business online.
The only way to keep yourself safe from the incessant privacy violations is to deny the big companies access to your data by using Mac and PC cleaning software to delete cache files or if you are using Android devices to use an Android cleaner to secure your files.
A report by McKinsey & Company notes that 50% of all current work activities are technically automatable by already available technologies.
“The report also notes that that automation could render jobless anywhere between 400 million to 800 million workers by 2030.”
With rising income inequality, coupled with disruptions in democratic systems around the world, the displacement of workers through automation could herald political headwinds the likes of which the world has never seen before.
It may tear down entire societies apart. Hopefully, we will have figured ways to solve the rising issues by then.
5. Adaptive Malware
Even without AI, cybersecurity is an already urgent task. Anti-malware developers are working day and night to stop attacks from crippling anything from electricity grids, schools, and hospitals to entire governments. IBM researchers, in anticipation of future threats, have been able to create AI-powered malware called DeepLocker that is able to adapt and counter any defences thrown at it. Malware entities such as DeepLocker will be commonplace in the coming years, and we will have to wait and see if the PC repair tips and tricks that we have long relied on to thwart malware attacks and other unwelcome disruptions will be of any good in the era of AI.
Artificial intelligence is here with us and how we manage it will help shape our future for the better or for worse. And while articles like this might sound alarmist, they serve to remind us how bad things could get if there are not enough controls to safeguard our tomorrow from the existential threat posed by AI.
Cathy Trimidal a Computer Engineer by degree and a writer by profession, writes for Software Tested and Outbyte.
Also Read: Will AI Rule the world and Take Over Humans