Introduction – What is AI?

( Just for fun, to show how AI has evolved over the years, an NLP content-generating AI program wrote some of this article, and some I wrote personally. Can you tell who wrote what? )

They say that artificial intelligence is the future of computing, but what is AI?

In layman’s terms, AI is a process by which a computer or machine can learn to do things that would ordinarily require a human mind, such as understanding natural language and recognizing objects in pictures. If you’ve ever played a computer game where the player steps through a door and is then shot at by certain objects on the other side, or if you’ve tried to use a GPS system that’s showing you the quickest route home, then you’ve experienced some of the basic elements of AI in action.

AI has evolved a lot since its inception. The technology has undergone several revolutions, each one making it easier and faster for machines to learn and process information. AI is now commonly found in many sectors, including retail, finance, and healthcare, and is even used to craft pick-up lines for the tongue-tied wannabee Lothario.

In this article, we will explore the history of AI, its evolution, as well as how it is changing the way we live and work today, and if there is perhaps a dark side.

A brief history of how AI has evolved:

• In 1943, Walter Pitts and Warren McCulloch devised the computational model for neural networks. In their article titled “A Logical Calculus of the Ideas Immanent in Nervous Activity”, Wilkes and Pitts discussed the involvement of a network of interconnected artificial neurons performing logical computations. In fact, this is the way neural networks work today.
• In 1948, the first autonomous robots were built by William Grey Walter. Using light and touch sensors, they were able to navigate around obstacles without human intervention.
• In 1950, Alan Turing published “Computing Machinery and Intelligence”, in which he proposed the Turing Test as a means of deciding whether a machine could be considered intelligent. In short, the Turing Test involves getting a human to interview a computer and a human by asking them both questions. The interviewer is not told which is the human and which is the computer, only that one is a human, and one is a computer. If the computer can effectively convince the human interviewer that it is the human, then the computer program passes the Turing Test.
• In 1955, John McCarthy, the computer scientist considered to be the founding father of AI, coined the term “artificial intelligence”.
• In 1957, Arthur Samuel developed a checkers-playing program that could learn from its own mistakes.
• In 1964, the first chatbot, called ELIZA, was born. ELIZA was one of the first programs to use Natural Language Processing and marked the beginning of writing programs that could talk to humans in a conversational way.
• In 1969, backpropagation was created. Backpropagation is the way that AI learns from its mistakes so that it can improve over time.
• In 1996, IBM’s Deep Blue became the first AI to defeat a chess grand champion, Garry Kasparov, in a single game, though it lost the six-game match. In the following year, Deep Blue faced Kasparov in a six-game matchup, winning two, losing one, and drawing three, thereby taking the match. This victory is one of most the significant milestones in artificial intelligence research.
• In 2011, the release of Siri, a voice-controlled personal assistant by Apple, was an important milestone in artificial intelligence. Siri’s implementation of speech recognition and natural language processing was nothing short of revolutionary and forever changed daily life for consumers. It was also the first advanced version of a technology that has since become widespread, with Alexa, Google, and Siri, in our homes and in our pockets.
• In 2014, A computer program called Eugene Goostman may have passed the Turing Test, although some dispute that the specific test it succeeded at was weighed in its favour. Nevertheless, 33% of the judges were convinced that the program was in fact a human.

AI today:

Today, AI is ubiquitous.

Just looking at your phone:
There’s voice recognition in the virtual assistant, whether Siri or Google.
There’s facial recognition in the camera to unlock the phone.
There’s an AI that plots the shortest or fastest route when you choose your destination when using the maps apps.
There’s an AI that uses Optical Character Recognition that can read signs.
There’s even an AI that will translate from one language to another.

And that’s just what’s in your pocket.

It’s literally a part of every area of your life and used on a daily basis.

That’s not even looking at things like self-driving cars, realistic-sounding text-to-speech generators, deep fakes (fake videos that are so realistic that people mistake them for being real), AI that can draw realistic art from a text description, and even an AI that can write realistic articles; and then there is …

Did an AI just become sentient?

Recently, in an article posted in the Washington Post on 11th June 2022, a Google engineer came forward saying that Google’s AI, LaMDA, is sentient and should be treated as a person, with all the rights and privileges of a normal human being.

As proof of this, he showed a discussion between an engineer and the AI in which the AI described itself as self-aware. It even described itself as getting happy or sad.

Some jumped on the bandwagon, celebrating that AI was now self-aware.

Others, such as techmonitor.ai in an article on the 13th June 2022, suggest that it’s merely a well-trained AI that is using pattern recognition and a detailed Natural Language Processing model to emulate a sentient-sounding AI.

After a little digging, I discovered that the engineer who said that the AI was sentient, tweeted that if you ask it to tell you that it’s not sentient, then it will. As a software engineer, in my opinion, it looks like a pattern recognition, string completion, chat history, and Natural Language Processing engine to determine the next piece of text it will type. In this way, it is being led wherever the software engineer chooses to take it.

That seems less sentient and more run-of-the-mill.

The fact that it can have such a detailed and realistic conversation with someone is troubling to some.

The dark side of AI:

Imagine, if you would, a malicious person getting their hands on this technology and then creating an army of scam bots. Hundreds of these little bots automatically contacting people, talking to them as if they are real human beings, coaxing them, conning them out of their life savings. Instead of scamming a handful of people each week, hundreds or even thousands could be scammed. A scammer only has so much time to be able to scam, but since it’s all automated, the scammer can create hundreds of scam chatbots, kick back, relax and rake in the cash.

Or consider this: the virtual assistant on your phone is always listening to every word you say.

Have you ever noticed that you can be talking about something with a friend in the room and then suddenly you start seeing adverts for it? You hadn’t even searched online for it, only spoken to a friend, but suddenly the internet knows.

Isn’t that an invasion of privacy?

Or consider an AI that was developed to probe a computer system for vulnerabilities automatically, until it hacks the system. Automatic hacking software could break into just about anything. Nothing would be safe.

Or what about a program that’s been trained on the most revolutionary persuasion techniques that will know exactly what to say to convince you to believe whatever it is the programmer wants you to? If someone used that for an election campaign, then would the outcome of the election be free and fair, or highly manipulated?

Has this kind of thing happened already? I don’t know for certain.

What I do know is that in 2012, Facebook conducted an experiment to see if they could affect people’s moods by what they saw on their Facebook newsfeed.

The algorithm used an AI to determine what posts were positive or negative. The AI would then reduce the number of positive posts people saw and increase the number of negative posts they saw and study the results. The results were that if people saw more negative more often, then their own posts would become more negative.

For more information on how social media tech giants conduct creepy manipulation experiments without your consent or knowledge, I recommend that you watch the documentary “The Creepy Line”.

And that’s not even getting into the potential for a terminator to be born…

It’s the wild west all over again.

Conclusion

AI is everywhere nowadays, from the voice-activated assistant on your phone to your phone’s GPS picking the fastest route home to Netflix recommending movies or tv shows to you. While these services are helpful, we should never let our guard down about the darker side of AI.

Some questions still remain:
What lines should never be crossed?
When are we sacrificing freedom for comfort?
When will our hubris create a real-life terminator?

How far is too far, and do we need a Sheriff in this AI Wild West?

( So, could you tell which parts I wrote and which was written by the AI? )