To begin, I must state that AI is the new wave that everyone should get on immediately. It is no longer an option. Yes, it is not flawless and needs to be explainable, but it is extremely powerful and amazing in a magical way.
Since Generative AI first came out, there has been a lot of talk about biases, trust, accuracy, hallucinations, and the need for, AI, that is responsible and ethical.
The question is, whose values and ethical principles should we follow? Do humans really agree on what is ethical and what is not? I am not discussing the general definition of ethical AI here. You will see if you look at current international conflicts that what some people deem unethical, others may consider to be purely ethical. I won’t provide any examples in this post to avoid potential blocking.
Now, when it comes to governments, how can they agree to expose their citizens to generative AI solutions that disregard their ideology or ethical principles, especially when these solutions cope with the values of their rivals that they disagree with?
It is true that they can’t afford not to embrace the “new” AI, but they also need to safeguard what they consider to be their values and ethics. For them, it might even be a national security issue.
So, what might governments do?
To be safe, they might either block their citizens from accessing what they don’t like them to access or use their own data to train “clean” generative AI models that only return “acceptable” results. This is exactly what I meant by “AI Silos” (you can call it “Gen AI Silos” as well).
What will happen if this really does happen? A disaster, I believe.
Allow me to explain; let’s assume that each of the US and its allies, China, Russia, India, the Middle East, and Europe, develops a Gen AI of its own. For the last two regions, some countries may not agree with their neighbors on all topics, causing them to build “sub silos” to meet their needs, assuming they have the resources to do that.
And to make it a successful scheme, some countries may require their population to use these artificial intelligence(s) instead of traditional search engines to retrieve information, which they (the citizens) may agree to do mostly due to our natural tendency as humans for convenience and speed, like what we do with fast food. If this happens, each nation or group of people will have one controlled, fully biased source of information, where access to the internet will be exclusively through these silos. As this violates the Web’s fundamental principles, I considered it a disaster. And if it happens, I think mankind could enter a new dark age.
Is it possible that humans can come up with a compromise to avoid this “dark” scenario in a way that satisfies everybody?
I hope so.
