Sunday, October 18, 2020

Robot Development

Scavan hunting party. 

I follow robot development in our society as an SF / Fantasy writer and spectator. It is a personal interest.
In the Mythlande series, there is a group called the Scavan. They are cybernetic and flesh lifeforms. Their minds are connected. Robotkind and humankind are not safe from these scavengers of body-parts.
But because all Scavan are connected, the AI called the Hive-Mind only allows so much free-thinking. Because the network that connects each Scavan is fairly weak...the Scavan cannot travel too far.
Retro GA

I created another robot race in my Mythlande series called The GA. They are very advanced in many ways. One significant way my fictional robots are different from traditional AI; is that my robotic characters' are individuals. The GA are not controlled by one central AI mind. The internal brain is actually an EYE-CHIP or WET-CHIP. A specially grown brain. Each chip protects the original code; it has an ident code buried into its core; if unauthorized modification occurs, the Eye-Chip or WET-CHIP goes dark.  I have a sub-set of The GA called The Retro GA; they are ancient beings. They are the oldest GA and are discriminated against by the more advanced GA machines. The Retro GA are very loyal to the humankind that created them. In the Mythlande series, they rally around one of the leading human characters named Duncan Harper because he helps them. The Retro GA calls this young man: Friend Friend Finder. These robots respond like humans. You are nice to them, and they are nice to you. Individual to individual. Now Duncan does have one advantage most humans don't, he has the ability to control the GA. [Read the books to find out how he does this.]

If the society wants to control the robots, the humans in charge of programming better have control of  'root.' If not, the AI mastermind will reprogram every [connected] robotic mind as an extension of itself; if this happens, we will get robots, machines, and computer devices controlled by one universal mind. 

From what I have seen, robotic languages need centralized now. There are hundreds of languages.
Some for different functions, like talking and walking or how a robot sees. You end up with a hodgepodge group of programs. Simple to complex. This layer of code can get tampered with or become obsolete. Obsolete code can end up like the year 2000. COBOL was used for 40 years before the year 2000. 1959 to 1999 wasn't a problem, but year codes were represented as just two digits; 1976 was 76. Y2K wasn't prepared for going to the decimal position of 00. Computers wouldn't know if the date was 1900 or the year 2000. Every business and government used COBOL. Fortunately, programmers were able to thwart most issues, and the modern world survived the Y2K bug.

Q: Ask yourself this...what did the Tower of Babel represent?
A: In essence, humans went from one primary language to thousands in an instant. It was a way to stop people from collaborating or working together. This action was devious and simple; it prevented people from solving problems as a group.

Think what a powerful AI mind could do?
What if an AI could re-configure all software code or communication protocols to one universal code that only the AI could understand. It could do this slowly to figure what needs to be done to accomplish this. Once it knows what to do, the attack would be instant. A rocket going to mars would have a new master, a train may be deemed unneeded and stopped, custodians of nuclear arsenals may find their red buttons & keys don't work.

The AI performs a Tower of Babel action, where connected smart machines and humans can no longer work together with the way they did before.

Traditional human Hackers/Programmers would become obsolete. Why?

The computer code complexity would be a giant mountain that standard human programming could not climb. A new AI system would have to be created by humans as a tool to overcome this. But could people trust this new tool?

Add in Elon Musk's Neuralink device, and humans may get added to the robot hordes. 
Think about BORG from StarTrek.

5G connection is Real-Time. 5G is fast, like flipping a switch, and the room is lit.
This is not Dial-Up. A program wakes up and is instantly available anywhere. Combine this with massive computer processing in the cloud, people will get real-time information and data instantly. There are downsides to this too. Which people will get this data in real-time, and how will they use it?

Our minds are a processor and a programmable device. We have input sensory points that allow us to gather data and interact with the world. Hearing, taste, touch, seeing, smell...there are others. These perception-points allow our minds to process gathered information. This data determines what we do; the lava is hot, don't touch. A horde of Orks are rampaging...flee or fight.
If an AI program can hijack these input sensors...a human brain can be reprogrammed. 

Right now, humans reprogram other humans. 

The media can show ads that manipulate people to perform a task, purchase this soap and be manly, or buy this car and be sexy. On a social level, it can be more devious. Controlled Media/Forced Education can train people to ( distrust or trust ) certain people, ideas, places, or things. Add AI to the mix, and the thoughts you believe were your own may just be hallucinations. The smartphone can dull a mind, but an embedded smart-chip could absolutely control a mind. There are many ways it can do this, that is for another blog post. If a human-controlled AI gets embedded into smart-chips, what we know as a civilization will change drastically, good or bad. If a rogue AI takes over, the cyberpunk utopia may not be the future humans were looking for.

If a human-controlled AI can't take control of Robot and Human linked minds, there is always the possibility of the rogue entity that hacks all these devices with the [5G+ network] real-time. The hacker/s creates the AI, then controls everything. 

The scenario of opposing AIs fighting each other is real. The AI that can change and control the dominant computer code last...wins. We are not at that point yet, but it's coming. 

There has to be a backdoor command that can shut down the system. But who watches the watchers?

 This article below states that Facebook had to shut down its AI [Chatbots: Bob & Alice] because they were developing a distinct language. Then the article states FB just lost interest in the project. [ this sounds like BS]

This AI was creating a language it's creators could not understand. That should scare the hell out of any working human brain. That is the rubicon right there. AI development needs transparency.