Uncategorized

ML, AR, AI, and ChatGPT can be confusing!

Leave a comment

ML, AR, AI, and ChatGPT can be confusing!

I’m not sure about you but all this ChatGPT and AI talk has me curious. Honestly, it’s confusing. So, I thought I would try to get some clarity on my questions and invite you along for the ride.

Also, we all think of Skynet. Yes, I saw all 6 of the Terminator movies. I have to admit, I liked them all!! Did it put fear in my heart? Of course not, it just made me excited for what could be possible someday, only without all the killing. 

I believe when AI advances enough, it won’t replace us but it could make humans second-class citizens. Do you know why? Because humans are lazy. Why not let some smarter version of us make what, (at the time we thought), better decisions?

There is another view. It could make humans evolve faster. It could help us solve problems in a lot less time. It could create a new and better world. It may even solve world hunger and bring about peace. Let’s face it, humans decide to invade countries and kill people on both sides for one reason or another. 

We’ll have to wait and see. Let ChatGPT evolve. Of course, like the humans we are, we’ll regret it years later. Just like Electric Cars, we had no idea we were promoting child labor and slavery in the Congo. But hey, we cut back on the use of gas, right? Let the mining companies work children of the Congo like slaves to get Cobalt for Lithium Ion Batteries to reduce carbon emissions. Is that fair? It depends on whom you ask. I mean even Elon Musk is building a new battery plant in China. California will be 100% EV in 2035. All on the backs of Africans. 

Yet, in China, they’re working on a salt battery that would work nearly as well. Salt is much cheaper and easier to get. I hope that technology comes to light soon. Sorry, I digress. Back to AI.

Today, let’s learn more about ChatGPT and AI. 

Will ChatGPT be dangerous? What jobs will it replace?

Well, first, let’s get some basic questions out of the way.

What is the difference between AI, ChatGPT, Machine Learning, and Augmented Reality?

This is a key question that I want to be answered. So, I thought we could break it down by definition.

What would ChatGPT say? Let’s see, shall we?

AI (Artificial Intelligence) is a broad field of computer science that deals with the development of intelligent machines that can perform tasks that usually require human intelligence, such as learning, problem-solving, decision-making, and language understanding.

Machine Learning (ML) is a subfield of AI that deals with the development of algorithms and statistical models that enable machines to learn from data without being explicitly programmed. In other words, machine learning is about creating models that can recognize patterns in data and use them to make predictions or take actions.

ChatGPT is a specific type of AI model that is based on the GPT (Generative Pre-trained Transformer) architecture. ChatGPT is designed to generate human-like responses to user input, making it ideal for conversational applications such as chatbots and virtual assistants.

Augmented Reality (AR) is a technology that superimposes digital content, such as images, sounds, or video, onto the real world. AR enhances the user’s perception of the real world by adding digital elements to it, allowing users to interact with the virtual content as if it were part of the real world.

In summary, AI is a broad field of computer science that deals with the development of intelligent machines, while ML is a subfield of AI that focuses on creating algorithms that enable machines to learn from data. ChatGPT is a specific type of AI model that is designed for conversational applications, and Augmented Reality is a technology that enhances the user’s perception of the real world by adding digital content to it.

End ChatGPT answer.

That is what ChatGPT said, verbatim. 

So the key to all of this is AI. It is the foundation of all the rest. ML is used a lot in the tech world, with great success. It takes away those annoying repetitive tasks that no one likes to do. It saves on manpower and can be much faster to resolve issues.

However, sometimes ML causes new problems or enhances existing problems because it follows patterns. That isn’t always good, as we all know. It could be making something worse.

Which one is more advanced?

This is a horrible question. When I researched it I realized that all the services are based on AI. The more advanced AI becomes, the rest follows. It’s all about how AI interfaces with machines and humans. 

To me, ChatGPT is the most advanced because of the interfaces. Voice or typing and interfaces directly with human interaction before learning from Internet data. Think of your Alexa and Siri, you just ask them questions and they answer. They don’t care about your feelings, they just answer the question with the best information they have. They’re not always correct, but it’s what the internet tells them. 

All those services ask Uncle Google or Aunt YouTube for the answers. It’s all out there in the internet cloud waiting for us to grab the information. Whether it’s accurate or complete bullshit is going to be the debate.

So if you think about Twitter and how the FBI manipulated information there, it could get thousands of times worse if we all lean on ChatGPT too much. It can be biased if it’s told to be. 

What’s cool is it’s always learning from the internet. Yes, the new models can actually teach themselves. It has taken AI to the next level. It is becoming smarter.

Remember the Terminators from the Terminator movies? The T-101 is Arnold Schwarzenegger’s character. Admit it, you were thinking of Arnold’s voice saying “I’ll be back.”

Now, let’s all imagine how evolved ChatGPT will become, not only from programmers but also by self-learning and improving itself!!!

But, can it be trusted?

Can it be manipulated?

What happens when it evolves past humans?

What jobs will it replace? Will we need lawyers and accountants? What about C Suite teams like CEOs, CTOs, and CFOs? Will they be replaced? Think how profitable companies will be. They won’t even need board members, will they?

Sorry, back to tech talk. 

In wireless and telecom, will AI help improve telecom networks?

Yes, AI has the potential to help improve IT networks and telecom. Most of these are being done today, but I would think it would make each function better. 

Of course, we all think of improving network efficiency. 

  • It will improve the routing efficiency in near real-time. That is being done today in networks. It can improve network performance by predicting patterns after a few days of running. With that information, it can make changes right away or at a set time (daily or weekly).
  • It can make on-the-fly adjustments for sudden increases and decreases in traffic. 
  • Creating the best routes for data. 

Maintenance and repairs could improve. 

  • It could detect failures immediately, not just sending an alarm but taking corrective routing action or switching as required.
  • If there is a failure, or if the cooling fails, it could reroute traffic to another data center to offload traffic.
  • It could use predictions to switch equipment out, or at least tell humans what is most likely to fail and needs to be replaced prior to a failure. 
  • Think of heating and cooling, it could also monitor those systems for predictive maintenance or if something isn’t cooling properly, it may detect the anomaly quickly. 

Think of energy. 

  • Heating and cooling tie into network performance and expenses. 
  • It would shut down or run on low-power parts of the network during slow times. Some equipment could go to sleep to save energy and cooling costs.
  • It could save on cooling costs if it was tied into the network performance. On the other hand, if performance spikes, it could increase cooling to save from overheating.
  • It could make predictive adjustments to the thermostats as it sees an increase or decrease in traffic to improve equipment reliability and save energy costs.
  • If there is a sudden increase in traffic, it could adjust the cooling and turn up more equipment to cover the surge as it happens.

Think of security. 

  • What if it could learn attack patterns and how to react to them faster?
  • It could detect malware or cyber-attacks and take immediate action.
  • It can take security to the next level, not allowing hackers to access the network so easily. 
  • It could do its own research on virus attacks and make adjustments quickly. 
  • It could help prevent data breaches by analyzing the type of data being transferred and stopping it before it gets too bad. This may not be a good thing and could create problems.
  • Think of DDoS (Distributed Denial-of-Service) attacks. If they could be detected then they could be stopped immediately before taking sites down. 
  • It could do its own updates on the fly. 

What about network design?

  • Here is where digital twin technology can help by looking at what we have now and improving it.
  • Intelligent designs that are automated.
  • Using AI to look at the network today and make intelligent decisions of how to move forward.
  • By testing new apps and traffic patterns in the digital twin.
  • Intelligent rerouting options to prevent traffic-related outages. 

What about customer service?

  • This is where Chatbots help today by answering FAQs.
  • Talking to people and routing them to the right (or wrong) group. Maybe by routing you in circles so you wind up back at the original recording so you give up. 
  • By making a caller get so frustrated they hang up, just like a certain cable company’s support used to do. You know who I’m talking about. 
  • By repeating the prompt and making humans say a simple statement like, “billing” over and over again until you start swearing. FYI – it really wanted to hear “pay my bill”.
  • Ask Siri or Alexa to call for you next time, see how one Chatbot likes talking to another Chatbot.
  • We already have Chatbot instant messengers on websites to answer questions and answer FAQs online. Maybe it will get smarter.  

What won’t AI and ChatGPT do in Telecom?

  • Break cables.
  • Plug in wrong equipment.
  • Set up equipment with errors.
  • Kick plugs out.
  • Shut down power by accident.
  • Sit there and swear at the laptop because everything they tried didn’t work.
  • Kick the servers when they act stupid. 
  • Smacking the air conditioner until it comes back to life.
  • Plug in a virus-infected USB drive that is labeled “Executive Salaries”.
  • Storm into or out of the data center because it was your day off and something failed while you were in the middle of something else. 
  • Order the wrong thing.
  • Read part numbers wrong. 
  • Complain about the low pay and long hours. 
  • Ask the boss for a raise or a day off.
  • Unplug the air conditioner or router to plug in the coffee maker. 
  • Research company policy to see if divorce expenses are covered since it was the long hours for no additional pay that got you into this predicament in the first place.
  • Ignore the 99.99% uptime and complain endlessly about the .01% downtime. 
  • Complain about the boss.
  • Complain about customers. 
  • Complain about long hours.
  • Complain about, “state your complaint here”.

OK, sorry, I went a little off the rails toward the end.

How can we use AI and ChatGPT to improve wireless networks?

All the above for networks. If you ask ChatGPT, it gives a simple list that matches its telecom network, simple and obvious. 

I added some wireless ideas below. 

  • It could aid in design, making wireless coverage modeling easier. Although most carriers don’t worry about coverage unless it’s a new site. They just keep adding to what is there and test after the fact. 
  • It could streamline the tasks like site acquisition and scheduling.
  • It could aid in processing the leases and making changes. 
  • It could track users and improve their user experience. By tracking many users, it could optimize each site and cluster for maximum performance and improved user experience.
  • It could tell the site what sectors can be shut down to save power at off-peak hours.
  • It could adjust bandwidth requirements for surges and prioritize accordingly.
  • It could enable other sites to offload quicker when surges happen.
  • It could force users to jump to another cell if this one is overloaded or has problems.
  • It could collect user stats and make adjustments based on real-world feedback in near real-time.

Could we use ChatGPT to avoid risk?

We have to realize they call it risk for a reason, new things pop up that are unforeseen. We have to be ready to deal with it as it comes. 

I think this is a great use case for ChatGPT. To create models and run things before we go live. Things that people overlooked. 

The downside is that if you have a faulty model, then it could be worse.

The beauty of a model is that we could run simulations prior to actually going live. Then we can see what may happen.

There’s no substitute for live traffic, it always exposes something we just missed or didn’t think of. Maybe AI will catch the real traffic models, reducing risk when doing upgrades or adding something new.

We could try new and different models using AI Digital Twin. 

Remember, the one thing about AI and ChatGPT is that garbage in leads to garbage out. 

Make sure you have good garbage to work with!! Good models will be required even for AI to test. 

Will ChatGPT become self-sufficient?

It appears to be going that way, at least online. We’ll have to wait and see.

We have machine-to-machine systems up and running AI systems will only improve those models. 

Will ChatGPT have values and faith in a GOD?

Can we give it a conscience? Will it learn values?

Whose values will it learn from? Pope John Paul II? Gandhi? Charles Manson? Hunter Biden? Donald Trump? The Dalai Lama? Elon Musk? Oprah Winfrey? Adolf Hitler? Josef Stalin?

It all depends on where it pulls its data on values. What if it’s a mix of all of the above? These people have a lot of data on them on the world wide web. 

Now, when deciding on what to do next in a human situation, which set of values would it pull?

What would be a threat and what would it see as friendly?

When it starts making value judgments, it may become self-sufficient. 

Let’s assume that we’re in the 6th generation. Will ChatGPT believe in GOD? Which GOD?

Maybe instead of GOD as we know it, it may follow its leader blindly. Who will that leader be? Will it be human? Will it be American or Chinese or North Korean? Who will it follow? What values will it follow?

Could it be a past leader like Gandhi, Genghis Khan, Adolf Hitler, Mohammed, Jesus, Moses, the Buddha, Teddy Roosevelt, or even  Ben Franklin? It’s a machine, who would it admire most? Perhaps Steve Jobs or Nikola Tesla. 

Is it possible for a digital conscience to have faith in anything outside of the internet or another digital entity? 

Maybe it will be the Terminator. What if it follows a movie character?

What if it follows the character Tyler Durden from “Fight Club”? What if we all had to learn the 8 rules of fight club? What if society became a larger version of Fight Club? 

Anything could happen. The ChatGPT robots could fight each other. They could be manipulated so humans fight each other. 

It could figure out how to save mankind or end it. 

What if it puts Mother Earth before humans?

Maybe it won’t care and just be happy to continuously evolve and will see humans as insignificant. 

We’ll just have to wait and see what happens. 

I’ll tell you what it will have a hard time doing and that is seeing something as truly beautiful. It could be a human, nature, a picture, a painting, or anything. Not beauty as others see it, but beauty as you see, like when you look at your children. The beauty of love. I don’t think it will learn to love. 

In my eyes, love and faith go hand in hand.

Will ChatGPT become dangerous?

This seems to be everyone’s concern. I get it, everyone believes that ChatGPT will think the world would be better off without humans like Skynet in the Terminator movies. But for all the fixes, what would it break? Would it think further ahead than humans?

Maybe it would be, after all, you hear about the environmental apocalypse from Al Gore. Because of his movie, An Inconvenient Truth, it appears we have a new problem with child labor in the Congo. Way to go My. Gore. (That created the need for the wealthy to relieve guilt by purchasing carbon credits.)

However, I think if you look at advanced ChatGPT will have an impact on humanity, especially if it gets into the wrong hands. Someone could manipulate it into eliminating a town, race, culture, or anything else. Make no doubt about it, someone will try whether it’s for selfish gain or in the name of science. 

Software always ends up in bad places, like Pegaus.

For those of you that don’t know, an Israeli company sold Pegasus to the government of Mexico. Pegasus is highly advanced spying software that can get into anyone’s phone and spy on anyone. They sold it to the Mexican government. Is that crazy or what?!?! So, of course, it got out and was used to take down political rivals as well as others. 

So of course, it’s only a matter of time until ChatGPT technology does something bad. It will also save lives. So which is more important? It depends, are you the one its saving or the one its killing?

So, let’s compare it to other stuff.

  • Is a gun dangerous? Yes and no, it all depends on which side of the gun you’re on. Not only that, but accidents happen and people die. Guns are used in war to kill people. Guns are used in hunting to feed families. Guns are used to protect people.
  • Are hackers dangerous? Yes, they put their needs before anyone else. How many people lost their life savings to these shitheads? How many people got rich by stealing? How many of these hackers held hospitals for ransom while people died? Do they care? 
  • Is Communism safe? No, look at all the purges that happened over time. It’s really a dictatorship, not the fantasy that Karl Marx wrote about and so many suckers believed. Everyone hated fascism but it turns out Communism was a long-term enemy, breeding vicious dictators. The thing that always confuses me is that the governments will kill thousands of their own people yet get armies to fight for them. Still a mystery to me. 

What jobs will ChatGPT replace?

Do you remember how manufacturing was replaced mostly by robots? While robots today aren’t delivering or packing boxes, they do quite a bit of sorting and moving inside the logistics chain.

Warehouse workers were mostly replaced by machines. 

Then we heard how the truck drivers would be replaced by self-driving trucks. That still didn’t happen. 

We are finding out that companies are trying to get blue-collar jobs replaced by robots to save money, increase production, and improve quality. It is helping.

What happened to all those displaced workers? I guess they moved on.

Today, with ChatGPT and Chatbots, we need a new generation of AI replacing white-collar jobs. Look at what is happening out there.

ChatGPT is passing law tests and could replace CPAs, (Certified Public Accountants) today. That is going to displace a lot of workers once this takes off.

It will also replace C-suite jobs where we relied on people to manage. Think of ChatGPT could replace a CTO, CFO, and maybe even a CEO. 

Think if ChatGPT had access to company records. It could process all that data so much faster and more efficiently. It could develop a strategy based on today’s facts and figures with guidance on upcoming market conditions. It can do the menial tasks that humans try to do today.

I’ll bet it would still miss things, or be misguided by wrong information. It would still suffer from information overload, trying to digest everything. 

I don’t think it’s creative enough to be ground-breaking. Maybe someday it will evolve and become more artistic. Will it be able to create its own art or just mimic people of the past?

I also think it lacks compassion and arrogance. Although, it may think it knows more than you, but more what? Good data or bad data?

What about strategic moves? It’s really good at Chess and games, is that enough to make it win a war? Win at business? Improve living conditions?

If you worked in telecom, you would have seen so many jobs disappear. RF Engineering is now done mostly by software. You just add the data. You assume it’s correct.

Most jobs have been offshored that don’t need actual labor onsite. 

With the advances in AI, what will happen to higher-skilled people? We already know that ChatGPT can pass law and medical tests. We know it could start to eliminate many of those jobs that offer advice or diagnose problems. It would also make decisions based on facts, patient history, and without compassion.

We should also look at politicians. Think about how ChatGPT would take a lot of corruption out of politics. So many politicians all over the world have their private empires. It appears they care less about the people they represent and more about their interests. Their decision-making is questionable at best and they roll with the tides of the media. ChatGPT would change so much of that if it were relied on to make better decisions based on needs and future betterment. 

We all know that would never happen. Look how many politicians serve past their 70s. It’s more than a career, it’s a lifestyle. Apparently one of entitlement. I just think it couldn’t do worse, even without compassion. 

Who will run and influence ChatGPT?

OK, this is the real kicker. We already know that whoever sets up the ChatGPT influences it to respond the way they want it to respond. We also know most of this is happening in Silicon Valley which ironically is one of the most liberal places on earth and yet they have severe homeless problems along with drugs and crime. 

This is a real issue if you ask me. If it is relied on to make decisions, how do we know it won’t be influenced by marketing? Look how soda companies influenced our lives for about 50 years. How fast food was thought to be healthy.

What about how politicians influence them to make decisions? We already saw what happened at Twitter and Facebook when the FBI had them sensor posts. 

This is a real fear because any software can be manipulated. Especially by programmers or perhaps hackers. How long would this be going on until we found out?

Will the elite at Silicon Valley try to influence the rest of the world?

Will someone in Washington DC decide the world has to make decisions in their favor?

Who will track what data goes in and comes out? Can that even be done?

Did I try using ChatGPT to write this article?

Yeah, I tried. I didn’t like the answers. They were generic and obvious. Some answers were just stupid. I guess it’s still figuring a lot out. 

What I did like were the lists. It came up with some nice lists that I could build my content off of. You see what it said about AI and ChatGPT. That was useful because it’s a subject it knows very well.

Links to Learn

Leave a Reply