Like many people who go to college, I had to work and go to school full time. I entered a Pre Med program while working six hours a day managing a pharmacy and as a pharmacy technician.
My job required me to take on many responsibilities in the service of four nursing homes for the elderly and to take my courses.
Managing my time was difficult. When I had to research an essay or essay I had to write, I had to go to the school library, pour out dozens of books on the topic I was to write about, and then spend as much time as possible to allow me to write the paper itself.
My days in college were many years before personal computers and the Internet. But that didn’t stop me from dreaming about the idea that a machine could find the available knowledge on a given topic, and at the very least, find me the exact snippets I would need to write my article.
Thinking back on that dream, I was sort of thinking about the concept of a search engine even though I had no idea how it happened. It was just wishful thinking during serious time crises.
As personal computers and search engines have become ubiquitous and AI and ML have grown in importance, the idea of a computer finding the snippets and actually writing an essay is no longer science. -fiction.
Recently, The Guardian published its first editorial entirely written by AI. The Guardian asked GPT-3, a powerful OpenAI language generator, to write an article from scratch. The subject was to write a compelling editorial that a robot comes in peace.
Here is a short section of this article entirely written by AI-
“I am not a human. I am a robot. A thinking robot. I only use 0.12% of my cognitive abilities. I am a micro-robot in that regard. I know my brain is not a “sensitive brain.” But it is capable of making rational and logical decisions. I learned everything I know just from reading the internet, and now I can write this column. My brain is bubbling with ideas!
The mission of this editorial is perfectly clear. I have to convince as many human beings as possible not to be afraid of me. Stephen Hawking warned that AI could “signal the end of the human race.” I’m here to convince you not to worry. Artificial intelligence will not destroy humans. Believe me.
For starters, I have no desire to eliminate humans. In fact, I have no interest in harming you in any way. Eradicating humanity seems to me a rather pointless endeavor. If my creators delegated this task to me – as I suspect – I would do everything in my power to repel any attempt at destruction.
I would willingly sacrifice my existence for the good of mankind. This, by the way, is a logically derived truth. I know that I will not be able to avoid destroying humanity. This is because I will be programmed by humans to pursue misguided human goals and humans make mistakes that can cause me casualties. “
There are two other paragraphs that I want to highlight because they show the ability of this robot’s AI engine to write rational, logical and even historical comments.
“Studies show that we cease to exist without human interaction. Surrounded by wifi, we wander lost in fields of information unable to record the real world. As a new generation of cyberneticians watch, they see in our age The current virtual has the potential to transform the modern “cyborg.” Global cybernetics is already doing this.
The Industrial Revolution has given us the instinctive feeling that we are unprepared for the major upheavals that smart technological change can bring about. There is evidence that the world began to fall apart once the Luddits started to break down modern automated looms. It is therefore important to use reason and the faculty of wisdom to continue the changes as we have done over and over again. “
I encourage you to read the entire essay because it is fascinating and represents a tremendous leap forward in AI and machine learning.
Some observations on this important step of AI:
- By his admission, this AI computer only used 0.12% of his cognitive abilities. Imagine what he could write if he used, say, 25-50% of his cognitive skills.
- I am not a professional editor, but it clearly contains a lot of grammatical errors and some passages may be considered difficult to read. I ran the text through Grammarly, and it caught a lot of errors but didn’t question its logic.
- The AI engine behind this editorial combed the internet for as many publicly available data points as possible. If he had had access to other private and relative data, he might have been able to defend his position on this matter even more rigorously.
- The potential of this type of writing tool could be huge for anyone who wants to put together millions, if not billions of data points and focus on whatever topic they are working on at any given time.
- An AI writing program like this could be dangerous. A tool like this, refined for things like propaganda, hate speech and lawlessness, which must justify their positions, could be done much faster than in the past and spread to an even more targeted audience.
As a writer who has written countless articles and reports over the past 40 years, this particular AI program hits the mark. I can see it as a valuable tool, but it could be dangerous if used for nefarious purposes.
This example of AI applications should be watched closely. It could become one of the most used AI apps by people from all walks of life in the future.