The problem with AI is this: It’s not intelligent.

    The problem with AI is this: It’s not intelligent. 

"Opaque AI systems risk undermining human rights and dignity. Global cooperation is needed to ensure protection." (ScitechDaily, “AI Is Not Intelligent at All” – Expert Warns of Worldwide Threat to Human Dignity)

The AI is not as intelligent as we know intelligence. The AI utilizes specific parameters to gather information. If AI collects information about a person, it might use things like photographs. And it recognizes that object. Then it searches texts that are written or films that are connected to that image. Then it creates a data matrix about that person. The problem with that is this. The AI doesn’t actually know. Do those texts and other data have any connections with a real person? 

The AI is a tool that can connect data from multiple sources. And it can make many impressive things. But AI doesn’t know. Does the person really have a connection with the data that AI uses? This is the problem with the AI. The AI connects data and makes medians, and then it creates a summary about that text. An interesting thing about AI is that. Every single data unit can be introduced as a numeric value. We can give as an example a ca value of 6. Or something like that. 

This makes the system easier for programmers because numeric data is easier to handle. But the problem is that. AI doesn't have a natural ability to be suspicious of data. That it uses. This means the AI is easy to cheat. The AI can have prohibitions that prevent users from using the AI for certain types of purposes. But the problem is that the AI programmers and engineers are not always experts in law. Another problem is this: what if  the AI software order comes outside the EU region? 

The programmer can operate using remote connections. That means the person who orders software might be different from the person whose name is in the paper. This means the real customers can sit far away from the person who is marked as a customer. The problem with algorithms is this. Things like malware are easy to make using the AI. And there are lots of actors like China and North Korea who want to create AI to control their society. 

Those actors have no limits for recruiting programmers to create AI that can generate spy programs. People forget this aspect too often. They think that the entire world is like the EU, where people follow the same rules. We forget that the Internet allows spies to operate from thousands of kilometers away from their target. The spy who operates for Chinese intelligence can sit in an office and use the internet and virtual spying tools to steal secrets. 

https://scitechdaily.com/ai-is-not-intelligent-at-all-expert-warns-of-worldwide-threat-to-human-dignity/