You can hardly turn on any of your various devices these days without encountering yet another hyperventilating article or program about ChatGPT or, more generally, about artificial intelligence (AI) chat bots. Time magazine recently ran a cover story about ChatGPT and the AI chat bot race. As if that were not enough to signal that ChatGPT has arrived as the social phenomenon du jour, the Wall Street Journal recently ran a long self-consciously serious op-ed column in which Henry Kissinger and two other public intellectuals called ChatGPT “a new technology [that] bids to transform the human cognitive process.” Wow! OK, then.

Despite these and many other recent signs and signals of impending doom, I was fully prepared to ignore ChatGPT and carry on with my life – that is, until one of my colleagues recently asked me whether, given that advent of ChatGPT, my days as a blogger are about to come to an end? The question, in its simplest form, is whether I am about to be replaced by a machine. If you listen to the current ChatGPT hype, it may be a question that all of us need to be asking ourselves.

First of all, and for starters, what exactly is ChatGPT? This seemed to me to be exactly the kind of question one might profitably ask ChatGPT itself. So I logged on to the ChatGPT site, set up an account, and went straight to work. What, I asked my new technological acquaintance, is ChatGPT? Here is what ChatGPT responded:

ChatGPT is a large language model developed by OpenAI, designed to generate human-like responses to user input. It is part of a family of deep learning models called Generative Pre-trained Transformers (GPT), which are trained on massive amounts of text data and can generate natural language text that is often difficult to distinguish from text written by humans.

ChatGPT works by analyzing the input provided by the user, processing it through a complex series of neural networks, and generating a response based on patterns and correlations in the vast amounts of text it has been trained on. The model is constantly learning and adapting to new data, allowing it to generate increasingly sophisticated and contextually relevant responses over time.

To use ChatGPT, users typically enter their input into a chat interface or other text-based communication channel. ChatGPT then analyzes the input, generates a response, and outputs it back to the user. The responses can be tailored to specific use cases, such as customer support, education, or entertainment, and can be integrated into a wide range of applications and platforms.

With all of those skills and capabilities, and in light of all of the hype surrounding the new tool, I decided I would go straight to my ultimate question – can ChatGPT write a blog post worthy of inclusion on The D&O Diary? Is writing a blog one of the “use cases” to which the tool can be tailored?

The test I contrived was to ask ChatGPT to write a post on a current topic that is actually something I considered writing about myself. I asked the bot to write a blog post about the SEC’s recent amendments to Rule 10b5-1 and what the implications are for executives’ adoption of Rule 10b5-1 trading plans. Here is what the bot came up with:

Hello and welcome to our blog, where we bring you the latest developments in securities regulation and compliance. Today, we’re discussing the recent adoption by the SEC of amendments to Rule 10b5-1, and the implications these changes may have for the adoption of Rule 10b5-1 trading plans.

First, some background on Rule 10b5-1. This rule was introduced by the SEC in 2000 as a way to provide insiders with a safe harbor from insider trading liability. Under the rule, insiders could establish a trading plan that specified certain parameters for buying or selling company stock. By doing so, insiders could trade company stock without fear of liability, even if they later became aware of material nonpublic information.

However, the use of Rule 10b5-1 plans has been controversial, with some critics arguing that insiders have used them to engage in insider trading. In response to these concerns, the SEC recently adopted amendments to Rule 10b5-1. The new amendments seek to address some of the concerns raised by critics of the rule, while still allowing insiders to use Rule 10b5-1 plans as a safe harbor.

One significant change introduced by the amendments is a requirement that insiders must wait a minimum of two months after establishing a trading plan before making any trades. This waiting period is designed to prevent insiders from using Rule 10b5-1 plans to engage in short-term trading that could be seen as suspicious.

The amendments also require insiders to disclose their trading plans to the public, including the specific parameters of the plan. This disclosure requirement is designed to provide greater transparency and help deter insiders from using Rule 10b5-1 plans to engage in insider trading.

Another change introduced by the amendments is a requirement that insiders must not enter into multiple, overlapping Rule 10b5-1 plans. This is intended to prevent insiders from using multiple trading plans to evade insider trading liability.

So, what are the implications of these changes for the adoption of Rule 10b5-1 trading plans? Overall, the amendments should provide greater clarity and transparency around the use of Rule 10b5-1 plans, which could make insiders more comfortable using them. However, the waiting period requirement may deter some insiders from using the plans, as it could make it more difficult for them to execute trades at the optimal time.

In conclusion, the recent amendments to Rule 10b5-1 represent a significant shift in the regulatory landscape around insider trading. While the changes may deter some insiders from using the plans, they should ultimately help restore public confidence in the integrity of the markets. As always, we’ll continue to monitor developments in this area and keep you updated on any further changes or developments.

I have to say that I found this to be a surprisingly strong response. Although I don’t like the contrived introduction at the beginning of the bot’s blog post, the rest of the post covered the topic in several interesting ways. It is particularly noteworthy that on its own, the bot began its discussion by introducing the Rule itself and its purposes and then identified the controversies that have arisen concerning the Rule, before then talking about the amendments . While there is a lot more than could and should be said about the Rule’s amendments, the bots draft blog post is not off base. It is not obviously junk. Other than the cheezy intro, it reads like a serious commentary.

Having put the chat bot through the paces on a sort of standard blog topic, I decided to try to throw the bot a more challenging test. One of the things I particularly enjoy about having a blog is being able to write blog posts about my travel adventures. It is perhaps the thing I enjoy most about having the blog. What would the bot come up with if I asked it to write a blog post about a trip to Paris, and to do so in the voice of The D&O Diary? Here is what it came up with:

Bonjour, mes amis! I recently had the pleasure of visiting the City of Light – Paris, and what a magnificent experience it was. From its breathtaking architecture to its world-famous cuisine, Paris truly lives up to its reputation as one of the most beautiful cities in the world.

My first stop was the iconic Eiffel Tower, and I must say, it was an awe-inspiring sight. As I made my way to the top of the tower, the panoramic view of Paris left me speechless. The River Seine winding through the city, the grandeur of the Louvre Museum, and the quaint, charming rooftops of the Montmartre neighborhood – it was all simply stunning.

Next, I made my way to the Louvre Museum, which houses one of the largest collections of art in the world. From the Mona Lisa to the Winged Victory of Samothrace, the Louvre is a treasure trove of priceless masterpieces that left me feeling humbled and amazed. The sheer size of the museum is overwhelming, and I could easily have spent days exploring its halls and galleries.

Of course, no trip to Paris would be complete without indulging in the city’s world-renowned cuisine. From croissants and macarons to escargots and foie gras, Paris is a food lover’s paradise. I savored every bite of my baguette sandwich at a sidewalk café, sipped on a rich cup of hot chocolate at Angelina, and had a memorable dining experience at Le Jules Verne, a Michelin-starred restaurant nestled inside the Eiffel Tower.

But what truly made my trip to Paris special were the people. The locals were warm and welcoming, and I loved practicing my French with them. Whether it was chatting with my Airbnb host, admiring the art of street performers, or making new friends at a neighborhood bistro, I felt a true sense of community and connection during my time in Paris.

In conclusion, Paris is a city that deserves to be experienced at least once in a lifetime. Its beauty, culture, and charm are simply unparalleled. So, if you have the chance to visit this magical city, I highly recommend taking it. You won’t regret it!

While I feel obliged to protest that this draft is really not in the voice of The D&O Diary (for example, I would not start a post with an overly cute introductory phrase in French), this was otherwise an interesting effort. I liked the details, such as bot describing chatting with its AirBNB host. While the bot did plunge deep into its cache of stock phrases (“magnificent experience,” “panoramic view,” “treasure trove”), other sections do again have an interesting writerly air (I kind of liked the reference to the quaint charming rooftops of Montmartre and the comment that visiting the Louvre left the bot feeling “humbled and amazed,” because visiting the Louvre does in fact leave me feeling humbled and amazed). There is one particularly nice little detail smuggled into this that brought a smile to my face. The reference to the “rich cup of hot chocolate” at Angelina’s is actually something I might well recommend to others. Again, the bot’s blog post was not obviously junk. Yes, some might say, it is so formulaic — which of course it is, that is the whole point.

I don’t think the bot’s blog post could pass as my work but I could easily imagine someone trying to pass it off as their work – which does start to raise all sorts of questions about how someone (or many someones) might try to use ChatGPT not just as a shortcut but as a complete substitution for effort, creativity, and in lieu of actual accomplishment. I fear that need to get ready for an onslaught of ChatGPT imposters and poseurs.  

I continued to experiment with the tool, trying to understand how one might best use its capabilities. One thing I tried that I have to say produced a truly magnificent performance from the tool was to ask the bot to help plan a trip to Seville, in Spain (which is a trip my wife and I are considering taking this Spring). I asked the bot to come up with a proposed travel itinerary for a three-day visit to Seville. Here is what the bot came up with:

Sure, here is a three-day itinerary for a visit to Seville, Spain:

Day 1:

Morning: Begin your first day in Seville by visiting the Royal Alcázar of Seville, one of the most beautiful examples of Mudéjar architecture in Spain. The palace was originally built by the Muslim rulers of Seville in the 10th century, but it was expanded and renovated by subsequent Christian kings, resulting in a fascinating blend of architectural styles.

Afternoon: After exploring the Royal Alcázar, head to the nearby Seville Cathedral, which is the largest Gothic cathedral in the world and a UNESCO World Heritage site. The cathedral was built on the site of a former mosque and contains many notable features, including the tomb of Christopher Columbus.

Evening: For dinner, head to the Triana neighborhood across the Guadalquivir River from the main tourist areas. Triana is known for its flamenco culture and excellent seafood, so you can enjoy a traditional Spanish meal while immersing yourself in the local culture.

Day 2:

Morning: Start your second day in Seville by visiting the Plaza de España, a stunningly beautiful square that was built for the 1929 Ibero-American Exposition. The square features a large central fountain, a moat, and a series of buildings with intricate tilework that represent different Spanish provinces.

Afternoon: After exploring the Plaza de España, take a stroll through the historic Barrio Santa Cruz neighborhood, which is famous for its winding streets, colorful houses, and hidden plazas. This neighborhood was once the Jewish quarter of Seville and is full of history and charm.

Evening: In the evening, head to the Metropol Parasol, a large wooden structure that is also known as the “Mushrooms of the Incarnation.” The structure offers incredible views of the city, and you can enjoy a drink or a bite to eat at one of the many bars and restaurants located on the top floor.

Day 3:

Morning: On your final day in Seville, take a day trip to the nearby town of Córdoba, which is known for its stunning mosque-cathedral and its charming old town. You can take a train from Seville to Córdoba, which takes about 45 minutes.

Afternoon: Once you arrive in Córdoba, spend the afternoon exploring the city’s many sights, including the Mezquita-Catedral de Córdoba, which is a UNESCO World Heritage site that combines Islamic and Christian architecture, and the Alcázar de los Reyes Cristianos, a former palace that is now a museum.

Evening: Return to Seville in the evening and head to the lively Triana neighborhood once again for some tapas and drinks. This area is full of small bars and restaurants where you can sample some of the best food and wine in Andalusia, so be sure to try some of the local specialties before you leave.

I have to say the bot’s proposed itinerary is a truly impressive performance. The places and sites the bot proposed are very much the kinds of destinations my wife and I would want to visit. I particularly liked the richness of the bot’s explanation of the sites proposed; there is a vibrancy of description and a use of historical detail that made the proposed itinerary sound like a travel plan that you would really like to try. I also particularly liked the bot’s suggestions of where to go for food and drink in the evening. The bot’s inclusion of a suggested day trip to Cordoba is also a nice detail; my wife and I were already thinking about the possibility of a day trip to Cordoba.

From the short time I spent playing around with the bot, it is clear to me that the tool has extraordinary potential. Sure, I included some criticisms of the bot’s textual responses above, but those criticisms arguably overlook the very impressive output the bot can generate. I think it will take time to figure out what the bot’s best uses are and how best to incorporate bot responses into larger projects, but I think there is every incentive for all of us try to figure those kinds of things out.

Many people have said that the availability of AI bots like ChatGPT are going to change the way we work, and I have to say I think there is some truth to that. Now, I am old fashioned, and it is very hard for me to imagine my substituting chat bot responses in place of my own basic document drafting routines. After all, oftentimes I still pull out a pad of paper and a pen and write out my first drafts by hand. On the other hand, if I need some quick descriptive text (say, an explanation of the Rule Against Perpetuities; a description of the causes of the monsoon rains in southeast Asia; or the problems with Major League Baseball’s new pitch clock), I can definitely see asking ChatGPT to write those things for me, or to at least write up the first draft for me.

So, to return to where we started, is ChatGPT going to replace me as the blogger-in-chief of The D&O Diary? I don’t think so, at least not anytime soon. If you compare ChatGPT’s Paris travel post to one of my Paris posts (say, for example, this one) you will clearly see that the absence of a human personality in the machine version is a discernable difference.

My view is that ChatGPT responses are a tool for writing, drafting, and creating, but they are not a substitute for those exercises. That said, and as I noted above, there are going to be people, perhaps many people, who may be all too happy to try to pass of a chat bot’s responses as their own. There may even be many people who are persuaded to accept the machine version as “good enough.” We are all about to live a massive experiment testing the meaning of some very basic things, such as, for example, “creativity” and “expression” and even perhaps “human communication.”

But in the meantime, it is going to be a lot easier to come up with superlative travel itineraries.

One final note. As a last exercise as part of preparing this blog post, I asked ChatGPT what it thought the social implications are of the advent of AI chat bots. Here is what the bot itself said:

The advent of AI chatbots has several social implications, including:

  1. Job Displacement: One of the most significant social implications of AI chatbots is job displacement. As AI chatbots become more advanced, they can perform many tasks previously done by humans, leading to job losses in fields such as customer service, data entry, and other related fields.
  2. Privacy and Security Concerns: AI chatbots have access to a vast amount of data, including personal and sensitive information, which can be a major concern for privacy and security. The potential for data breaches, hacking, and misuse of data can be a significant concern.
  3. Impact on Communication and Relationships: The increased use of AI chatbots in communication can potentially have an impact on human-to-human relationships. For example, if individuals rely too heavily on chatbots for communication, they may lose important social skills and miss out on the benefits of face-to-face interaction.
  4. Bias and Discrimination: AI chatbots are only as unbiased as the data they are trained on. If the data used to train them contains biases or discriminatory attitudes, these biases can be perpetuated and even amplified by the chatbot’s responses, leading to social implications.
  5. Ethical Considerations: As AI chatbots become more advanced, ethical considerations become increasingly important. For example, there may be concerns around the use of AI chatbots in fields such as healthcare and law, where human judgment and decision-making are critical.

Overall, while AI chatbots can provide many benefits, there are also significant social implications to consider as their use becomes more widespread.

Well, there you have it. Yes, my friends, we do have a lot to worry about, ChatGPT itself said so.