Mind Matters Natural and Artificial Intelligence News and Analysis
humans vs
Human vs Robots concept. Technological revolution. Unemployment in the digital world. Symbol of future cooperation, technology advance, innovation. Businessman flips wood cubes human to robot symbols.

GPT-3 Versus the Writers at Mind Matters

How does the AI fare when it is asked to write on topics covered in Mind Matters articles?
Share
Facebook
Twitter
LinkedIn
Flipboard
Print
Email

In order to give a real-world comparison of the output of GPT-3 to human-written writing, I decided it would be a fun activity to see how OpenAI’s GPT-3 compares to Mind Matters on a variety of topics that we cover.  Here, we are using OpenAI’s direct API, not ChatGPT, as there is a lot of evidence that ChatGPT responses have a human-in-the-loop.  Therefore, we are going to focus on the outputs from their API directly.

I used several criteria for article selection in order to even the playing field as much as possible.  For instance, I only chose articles that did not depend on recent events.  This way, GPT-3 is not disadvantaged for not having up-to-date material.  However, I also chose recent articles so the articles themselves were less likely to be part of the training data.  I also avoided articles that dealt with chatbots themselves, as it might be likely that OpenAI has preloaded GPT-3 with better responses about chatbots and other AI topics.  To eliminate investigator bias, I took the first output.  Occasionally I had to remove nonsense just to make the article coherent.

For those who want technical details, I also set a few model parameters in order to get reasonable articles.  First, I set max_tokens to 3500 to allow it to write more extensively. I also set the temperature to 0.8 so that it has fairly free reign, but has to maintain some amount of fidelity to the topic.  Finally, I used the text-davinci-003 model, which is supposed to be their most advanced model (which also costs more). 

I also tested different ways of forming responses on a few articles (not in the list below) in order to get the best results, but, once I got the format down, I went with the first response from each prompt.  I found out that if I was too specific, I would either get very short or very repetitive responses.  If I force it to talk longer or forced it to not be repetitive (both are settings in the API), I would sometimes wind up with nonsense responses, like endlessly repetitions of the word “the”.  However, the settings above with a “general topic” prompt seemed to produce the best output, which is given below.

For each article below, I’ll give a link to the Mind Matters article, the prompt I used, and the GPT-3 output. The results seem to be a fair summary of Wikipedia-like knowledge. The one where the lack of understanding of AI really showed through was in the second one below, which, in a prompt about spiders, intermixed data about spiders “the animal” with spiders “the technology used to search the web.”  Anyway, the results are below—you be the judge!

Are Extra Dimension of the Universe Real or Imaginary?

Mind Matters Article: here.

Prompt: Write a blog post for a technical audience about whether extra dimensions of the universe are real or imaginary.

For centuries, humanity has been fascinated by the mysteries of the universe and the idea that there may be more to reality than just the three physical dimensions we can observe. This has led to the notion of extra dimensions – that the universe may contain physical places that exist beyond our three-dimensional world. But are extra dimensions actually real, or are they just imaginary?                            

To answer this question, we must first understand what extra dimensions are. Extra dimensions are theoretical spatial dimensions that exist outside of our conventional three-dimensional space. They are often thought of as being “folded up” and too small to be observed in our everyday lives. The concept of extra dimensions has been postulated in many theories of physics, including string theory and M-theory.

The idea of extra dimensions may seem like a far-fetched concept, but there is evidence that these extra dimensions could actually exist. For example, the properties of elementary particles could be explained by the existence of multiple spatial dimensions. Additionally, some theories of physics have predicted the existence of “wormholes”, which could act as portals to connect different regions of space-time.

At the same time, some physicists argue that extra dimensions are purely imaginary. They point out that, although some theories of physics predict the existence of extra dimensions, no direct evidence of their existence has been found. Additionally, some argue that extra dimensions would have a significant effect on the universe, yet no such effects have been observed.

Ultimately, the debate over whether extra dimensions are real or imaginary is ongoing. While the evidence points to their existence, the lack of direct evidence makes it difficult to definitively say whether they are real or not. We may never know for sure whether extra dimensions exist, but the questions they raise are part of what makes the universe so fascinating.

Spiders are Smart: Be Glad They Are Small

Mind Matters Article: here.

Prompt: Write a blog post for a technical audience about strategies that spiders use to deceive prey and other spiders.

Search engine spiders crawl the web in order to provide users with the most relevant and up-to-date information possible. But did you know that they also use various strategies to deceive and outsmart their prey, and even other spiders?

Spiders have evolved a variety of tactics to outwit their victims. One of the most common is camouflage. This involves blending into the environment, perhaps by changing the color of the spider’s body to match the bark of a tree, or the color of the sand. By hiding in plain sight, the spider can sneak up on its prey.

Another strategy that spiders use is mimicry. This involves imitating the behaviors of other species. For example, some spiders mimic the movements of ants, beetles and other small animals in order to gain access to their food or territory.    

Spiders also employ ambush tactics to catch their prey. This involves waiting in a concealed location and pouncing on unsuspecting prey as they pass. Some spiders build elaborate webs with sticky threads, and others hide in burrows and crevices, waiting for the right time to strike.

Finally, spiders can also deceive other spiders by using special signals and scents. Some spiders, such as the orb-weaver, emit a strong smell that other spiders interpret as a warning. Other spiders use a special type of sound, known as “stridulation”, to deceive other spiders into thinking they are aggressive and not to be messed with.

These are just a few of the fascinating strategies that spiders use to deceive their prey and other spiders. By learning about these strategies, we can better understand the complex behavior of these amazing creatures.

Can Animals Be Held Morally Responsible for Their Acts?

Mind Matters Article: here.

Prompt: Write a blog post for a technical audience about whether animals are moral beings and can be held criminally liable for their acts.

In recent years, more and more studies have been conducted to try and answer the question of whether animals are moral beings and can be held criminally liable for their actions. Despite the fact that animals have been known to exhibit behaviors that could be considered ethical or immoral, the consensus among researchers is that it is unlikely animals can be held criminally liable for their acts.

Animals have been observed exhibiting a range of behaviors that could be regarded as ethical or immoral, depending on the context. Examples of this range from chimpanzees that have been seen helping others in need, as well as dolphins that have been observed engaging in infanticide. Despite these observations, scientists have yet to conclusively prove that animals are moral beings in the same way humans are.      

The most significant obstacle to holding animals criminally liable for their acts is the lack of legal recognition. The legal system currently recognizes only human beings as beings that can be held criminally liable, and animals are excluded from this system. This is because the legal system is primarily concerned with protecting the interests of human beings and punishing those that violate these interests. Until the legal system is modified to recognize animals as moral beings and capable of criminal liability, animals will remain outside the jurisdiction of the law.

At the same time, there are a few avenues available for addressing the ethical implications of animals’ actions. Animal welfare laws exist in many countries, and those laws can be used to prosecute individuals who have abused or neglected animals. Additionally, international organizations such as the World Society for the Protection of Animals have been established to promote the humane treatment of animals.

In conclusion, while animals have been observed exhibiting behaviors that could be considered ethical or immoral, the consensus among researchers is that animals are unlikely to be held criminally liable for their actions. This is primarily due to the lack of legal recognition of animals as moral beings, as well as the lack of a legal framework for prosecuting individuals who have abused or neglected animals. However, there are some avenues available through which animal welfare laws can be used to address the ethical implications of animals’ actions.

How San Francisco’s Gun Fears Prevented Lifesaving Innovation

Mind Matters Article: here

Prompt: Write a blog post for a technical audience about the use of lethal robots for police enforcement actions and the public response it has generated.

Robots are becoming increasingly prevalent in modern life – from robotic vacuums to self-driving cars. But could robots soon be used for law enforcement actions? This is a question that has been discussed in scientific and legal circles, and it has sparked intense debate in recent years.

The use of robots for law enforcement is becoming a reality, as police departments in some countries are now utilizing robots for certain tasks, such as bomb disposal. But there is debate about using lethal robots for enforcement actions. The idea is controversial, as there are many ethical considerations, such as the potential for robots to make mistakes, or even to act in a biased manner.

The public response to this idea has been mixed. Some feel that robots could be a valuable asset to law enforcement due to their precision and efficiency, while others are concerned that they may lack the necessary judgment and humanity. A number of civil society organizations have raised concerns about the use of lethal robots, and have called for more robust regulations to ensure that they are not used inappropriately.

At the same time, there is a growing acceptance of the use of robots in law enforcement, and the technology is becoming increasingly sophisticated. As a result, we may soon see robots being used for a variety of tasks, including lethal enforcement actions.

Whether or not robots should be used for law enforcement is a complex issue that requires careful consideration. The public response will no doubt continue to be divided, as people weigh the potential benefits and risks of this technology. Ultimately, it will be up to policymakers to decide how and when robots should be used for law enforcement.

The Prophecies of Mary Shelley’s Frankenstein

Mind Matters Article: here

Prompt: Write a blog post for a technical audience about Mary Shelley’s Frankenstein and what it represents for humanity and its relationship to both the material and the divine.

Mary Shelley’s Frankenstein is one of the most iconic novels in literature. Published almost two centuries ago, the gothic horror story still resonates with readers to this day. While it’s primarily known for its horrific elements, upon closer examination, Shelley’s novel reveals a nuanced exploration of the relationship between humanity and technology.        

At the heart of Frankenstein lies the idea of the “divine spark.” This is the belief that humanity is imbued with a unique and sacred energy that separates us from the rest of the natural world. Shelley’s novel deals with the morality of using technology to meddle with this divine spark. Victor Frankenstein’s creation of the monster is an attempt to usurp God by creating life without divine help.

In addition to exploring the “divine spark”, Shelley’s story is also a commentary on the dangers of unchecked scientific ambition. Frankenstein’s ambition is driven by a desire to prove his own worth and to master the secrets of life. While his ambition is initially noble, it quickly moves into hubris and overreaching. Shelley suggests that unchecked ambition and pride can lead to disastrous consequences.

The implications of Shelley’s tale remain relevant in our current technological landscape. Frankenstein serves as a warning to be cautious when dealing with both the material and the divine. We must be careful to not overreach and must consider the ethical implications of our technological pursuits. Mary Shelley’s tale of Frankenstein is a timeless reminder of the importance of tempering ambition with morality.


Jonathan Bartlett

Senior Fellow, Walter Bradley Center for Natural & Artificial Intelligence
Jonathan Bartlett is a senior software engineer at McElroy Manufacturing. Jonathan is part of McElroy's Digital Products group, where he writes software for their next generation of tablet-controlled machines as well as develops mobile applications for customers to better manage their equipment and jobs. He also offers his time as the Director of The Blyth Institute, focusing on the interplay between mathematics, philosophy, engineering, and science. Jonathan is the author of several textbooks and edited volumes which have been used by universities as diverse as Princeton and DeVry.

GPT-3 Versus the Writers at Mind Matters