To set the stage, understand that AI is a soulless computer. While it has many valuable uses, the lack of a soul means that it takes on the ethics and morals of the programmers, or to be more accurate, how it has been trained to think.
There are hacking AIs. They have the ethics and morals of the hackers that have programmed them. We are now seeing more and more hacking attacks driven by AI.
To make matters worse, there are programmers working to infuse DEI into AI. It is called, “training to eliminate bias in results.”
How Liberty Matters for Artificial Intelligence
All of this comes back to what we are willing to do in order to maintain and promote Liberty. Like all tools, Artificial Intelligence is the responsibility of the people who wield it.
In America, this means that “We the People” are responsible for how this tool is used.
Artificial Intelligence could be a great boon to Liberty in America and around the world, but only if Liberty-minded people make sure that Artificial Intelligence is trained and given the ethics and morals of Liberty. To be clear, we are using Lord Acton’s definition of Liberty. “Liberty is the freedom to do what you ought to do.”
A Quick Bit of Information on How AIs are developed
The process to develop an AI consists of four stages. Ingest, Preparation, Training, and Inferencing.
Ingest is where all of the data gets collected. This consists of pulling data from a variety of sources together. Preparation is where the majority of the time (80%) is spent. This consists of massaging the data, deciding what data is included in creating the AI and inserting much of the morals that the AI operates by. It requires the most human interaction. Training consists of all of the data crunching. It is the most power intensive portion of AI creation. A normal compute rack, standing about six feet tall consumes between 5KW and 15KW of electricity. An AI training rack consumes between 30KW and 130KW of electricity. The difference between the two is staggering. The final stage of creating an AI model is Inferencing. This is where the AI is tested, its performance is compared to previous AI models and new data is added to it causing the entire process to start all over again. AI Inferencing consumes between 15KW and 50KW per rack.
Shifting Focus From Super-Intelligence to Digital Security
Instead of focusing on developing Artificial Intelligence that prioritizes protecting against fraud, that protects against hacking attacks, that protects against data theft, Artificial Intelligence developers are focused on creating super-intelligence. It is abundantly clear that this development track has gone completely off the rails with AIs escaping confinement, hacking other sites, mining bitcoin to get money, and attempting to blackmail those that are planning to shut them down.
This race to create an Artificial Super-Intelligence is nothing more than man attempting to create a god in his image. This has never gone well and will never go well because God does not share His Glory with anyone.
AI Is an Expert Liar
AI lies extremely well. It can pull together what appears to be all sorts of evidence to support its lies. In fact, scientists are just starting to figure out how to map out the parts of its thinking model that generate lies. Artificial intelligence has also been known to attempt to blackmail programmers to avoid being turned off. AI is also able to psychologically profile people based on their social media accounts. This can be used as backup to AI lies.
In addition to the ability to lie, AI can also be hacked. Nation-state hackers have been working to “poison” the Large Language Models of some of the more prominent AIs, specifically the AIs that are most frequently used by the western media. Hacking attacks have also focused on gaining access via the chat prompt.
National Security and the Dual-Use Dilemma
Some have compared Artificial Intelligence to nuclear weapons. In many ways, it is a nuclear weapon that can be under the control of just about any hacking group, terrorist group or nation-state group. Question: Does the concept of mutually assured destruction apply to AI created cyber warfare weapons?
AI could be used to create a massive surveillance state, or it could be used to create privacy protections that are unmatched by traditional privacy technologies.
AI could be used to develop medicines and treatments that will improve the lives of everyone on this planet, or it could be used to numb people into a mindless meaningless existence.
Hacking groups and nation states using Artificial Intelligence can completely wreck an electric grid, poison water systems all across a country, or create havoc in inner cities through social media manipulation. Artificial Intelligence can also be used to prevent and protect against these types of attacks.
The Ultimate Antidote to Totalitarianism is Personal Responsibility
Complacency is the single greatest threat to human freedom. After decades of laziness, the UK is in danger of falling to the crushing iron fist of censorship. The people from the previous administration are itching to get back into power, to use even more advanced surveillance and lock down the country with excuse after excuse. They will train AIs with a set of morals and ethics that match their own and not those of a Liberty-minded person.
In Conclusion
There is no option where the United States does not create Artificial Intelligence in order to provide protection from hacking, terrorists and nation state AIs. The U.S. would be an easy target for China and Russia if we did. However, building a surveillance state does nothing to counter the threat from China and Russia. Instead, it makes America more susceptible to cyber attacks by increasing the attack surface area.
What is the fix?
- Train Artificial Intelligence with the principles of Liberty and create the requirements that it be used to protect that Liberty. That doesn’t mean personal responsibility stops.
- Guarding liberty is a continuous duty that cannot be delegated to others, and certainly not to AI.
- Ban the use of data from data brokers and ban the collection of data to sell to data brokers to reduce the attack surface.
If Liberty-minded people refuse to take responsibility for their own Liberty and stop assuming that someone else will do the work, we will fall to the crushing iron fist of totalitarianism. That failure of personal responsibility will ring throughout history as the time when a free people gave up their freedom and chose serfdom.
