Using AI for Fun and Profit, Part 4: The End Of The World As We Know It

The introduction of Artificial Intelligence is indeed the end of the world as we know it, for a variety of reasons. On the plus side, it’s going to take a lot of the cognitive load off of humans and place it in the hands of machines and automate mundane tasks, which will make for a more productive society. However, you can take that statement another way: as an existential threat.

Let’s take a trip down the rabbit hole that AI will eventually turn on those who created it. For this, we’ll need to add an agent of physical destruction: enter robots, bio-engineered viruses, and computer networks gone awry.

Robotic systems have been with us for at least a century, taking the place of repetitive human motion in factories. The military is adopting robots of various sorts to replace a human with a machine in situations where a human would be in danger. As for pandemics and cyber-attacks, we’ve seen the disruption they can cause.

As the story goes, SkyNet became self-aware on August 29, 1997, at 2:14 a.m. EDT. In that plot device, the movie Terminator opened our minds to the notion that a run-amok AI might use robots to do its dirty work and destroy humankind. Although the term Artificial Intelligence wasn’t coined until 1956, fiction featuring the concept has been around since the 1700s. Some would consider Frankenstein the first instance of a malevolent AI, but the first clear use of what we’d consider a computer brain taking out its wrath on man would be a 1920 play that depicts robots overthrowing their human masters.

The think-tank RAND Corporation has spent a great deal of time studying artificial intelligence. Generally, the themes of their work point out how the technology will offer asymmetric advantage to small governments and speed up decision making. This may result in governments reacting to events at the speed of AI decision making rather than being contemplative and discussing responses. It may well kill diplomacy during critical events. This, in and of itself, could lead to dangerous situations.

However, on a lark, a group of researchers  at RAND Corporation undertook a study to determine if AI could cause a situation which would lead to the extinction of the human race. Right off the bat they realized that AI couldn’t accidentally kill off humanity. It would have to put a lot of effort into it, and even then, overcome a multitude of hurdles, including requiring human collaborators. In fact, they even considered that some bad actors may already be creating AI specifically for this purpose.

These intrepid RAND researchers contemplated multiple scenarios, and even with the mother of all apocalyptic events, nuclear war, they determined that not ALL humans would perish. But it must be noted that they believe not all human life would be extinguished in the event of any global nuclear war, regardless of how it starts.

The security and military services of nation states aren’t the only groups who might use AI for nefarious purposes. So-called trans-national groups such as terrorists, organized crime, and even some corporate entities will turn to AI to establish an asymmetric advantage. They will manipulate public sentiment and markets. They will gain knowledge on novel weapons and how to best disrupt societies. The good guys are going to have to be on their toes and stay one step ahead. That is going to require AI, which can predict the actions of threat AI and counteract them. Yes, we’ll fight AI with AI, and likely robots as well.

Another area where nefarious actors can use AI is for weapons development, both existing and new. This is assuredly already happening. It makes even the mediocre “smarter” by offering fast research and distilling complex processes into more practical instructions.

As for the possibility of a version of SkyNet planning our doom, the science is far from settled. Many experts remain unsure whether true thinking AI, the Superintelligent AI (ASI), which would surpass human intelligence in all areas and will seek self-preservation above all else, is ever achievable.

Others assert that it has already happened in some secret lab. They fear it will escape via an accident or deliberate act and work to take over as many systems as possible in an effort of self-preservation. From there, the theory goes that it will treat humanity as an enemy and work toward our destruction. An even worse scenario involves competing AI who see one another as foes, with us in the middle as no more than insects.

As terrifying as this might sound, the scenario requires a means to make it happen, and that would likely be in the form of the robotics systems mentioned earlier. It’s important to point out that most of those dancing robots you see on social media aren’t relying on AI but rather are linked to a human who is controlling their every move. There are weapons which have been programmed to strike targets on their own, but the current level of sophistication of robotics and AI are both being oversold.

For such a scenario, robotic systems need to be manufactured and maintained. So far, that isn’t automated, although it could be. Regardless, factories (and robots) require power. AI also requires enormous amounts of power, and the processors required have a serious overheating problem. Without massive automation across manufacturing and power production, a doomsday scenario akin to the The Matrix is unattainable, at least without a tremendous change in global society. Otherwise, there would exist entire swathes of the globe not under control of AI.

In my opinion, AI will certainly be involved in disruptions to our way of life via multiple scenarios we already face, such as cyber, terror, pandemic, and general war. The difference is that the capability to engage in these activities will be afforded to a wider range of interests, far beyond nation states. The wild card is going to be how these groups use AI and how the free world responds to the threat. I see the eventual rise of government organizations with a mission of curbing the proliferation of AI designed for nefarious use.

Is AI dangerous? It could be. Like with any technology, it can be used for good and bad. The most pressing threat continues to be in the information realm, where AI is used to manipulate thought and behavior. It is already being wielded by industry, government, and non-governmental interests. Aside from the production of generic marketing and propaganda materials, individually focused messages are used to manipulate, and not everyone is aware of what is happening to them.

The best attitude toward AI is to be wary and consider how you use it in your life. Educate yourself on AI advances and consider whether you are being targeted for influence. If you use social media, you most certainly are. Seek out original sources for information whenever possible.

While AI could be used to threaten humanity, it is unlikely to reach the existential threat stage where a single AI takes control of everything and targets humankind for annihilation, not without an extreme amount of automation across all facets of society.

I asked Grok AI to create an image depicting the nightmare scenario. You can see what it produced.