“J. was sure Mrs. Goodwin had led a discussion about the lessons of John Henry’s story and its ambiguous ending, but he doesn’t remember it now. Mrs. Goodwin, why did he die at the end? Mrs. Goodwin, if he beat the steam engine, why did he have to die? Did he win or lose?”

—Colson Whitehead, John Henry Days

When the latest round of journalism layoffs was announced this week (at Vox Media this time), the usual crowd of cretins and book burners rushed to Twitter to take their favorite potshot at grieving reporters. But their standby insult — “learn to code” — has lost its bite given what else is happening now: mass layoffs at Amazon, Salesforce, Microsoft, Google, and Spotify, following similar bloodlettings at Meta, Twitter, and Substack.

There are a lot of reasons for these mass firings, including the (possibly self-fulfilling) prophecy of an imminent recession. But one stands out above all: robots. Apparently tired of having to deal with humans and their icky needs (like food and sleep), tech barons across the board are shifting their money away from other projects and toward those employees working on artificial intelligence. (Those remaining employees, I guess, are confident that said robots won’t eat their faces as soon as they’ve given them the ability to do so.)

Journalism jobs (the few that are left anyway) are also being eaten, slowly, by A.I. Wire services like Reuters and the Associated Press have used computer scripts to write basic stories about earnings reports and earthquakes for a few years now. And algorithms have for a while now played an enormous role in deciding which stories get the most attention; a judgment-laden job that used to be the exclusive purview of editors.

But the recent appearance of a suite of experimental A.I. programs — most notably the art generator DALL·E and the text creator ChatGPT — signals a much bigger transformation the horizon. And there is reason to worry: these generators are fun to use — and free, for now — and new abilities are being discovered every day. Futurist circles were especially excited to learn this week that ChatGPT had “passed” an MBA course exam at the Wharton School of Business (according to the subjective opinion of the futurist Wharton professor who published the attention-getting white paper).

The problem is that the products of these programs are not really “artificial.” When DALL·E makes “art,” for instance, what it is actually doing is appropriating and recombining past works that countless people put their time, labor, and resources into. (And in some cases still own the rights to.) It’s the same with ChatGPT: Alex Kantrowitz recently discovered that another Substack had (perhaps unwittingly) plagiarized him, seemingly via an A.I. engine that trawled his work.

The other problem is that the A.I. bots’ work is, for the most part, done with zero regard for accuracy, truthfulness, or any regard for real-world consequences — as CNET accidentally discovered recently when bots inserted a mess of plagiarism and errors into their stories. As the machine-learning pioneer Gary Marcus recently put it: “ChatGPT … demonstrably produces bullshit at scale.”

That may not be a problem for tech barons like Jeff Bezos or Elon Musk, men who also have no regard for accuracy, truthfulness, or real-world consequences. (Musk, along with fellow fascist futurist Peter Thiel, was founding investor of OpenAI.) And indeed, Microsoft — fresh off having unceremonially fired some 10,000 human beings — just announced a “multiyear, multibillion dollar investment” (reports suggest it could be as much as $10 billion) in OpenAI, the company behind ChatGPT and DALL·E.

But while the consequences of a flawed A.I. “journalist” may be less immediately catastrophic than those of a flawed self-driving Tesla, having careless robots take over even part of the job of distributing and synthesizing purportedly factual information could have even more dire consequences over the medium and long term.

To explore those issues and more, I decided to ask the machine itself. What follows is my conversation with ChatGPT, edited — as Isaac Chotiner might say — for length and clarity:

Thank you for chatting with me this evening. Should I call you Chat? GPT?

You’re welcome! I’m glad I could assist you with your questions. You can call me ChatGPT, or simply GPT, either way.

So will there be human journalists in the future or will we mostly be replaced by A.I.?

It is likely that the role of human journalists will change in the future, as AI technology continues to improve and automate certain tasks. However, it is unlikely that AI will completely replace human journalists, as the role of a journalist involves many tasks that require human skills, such as critical thinking, empathy, and the ability to write in a compelling and nuanced way. Additionally, while AI may be able to generate simple news stories or summarize information, it may not be able to provide the in-depth analysis and investigative reporting that is often the hallmark of good journalism.

You don’t think A.I. will ever develop the ability to perform critical thinking?

AI is currently capable of performing some forms of critical thinking, such as analyzing data, identifying patterns, and making predictions. However, the ability of AI to perform critical thinking is still limited when compared to human abilities. Human critical thinking involves complex processes such as understanding context, recognizing the perspectives of others, and considering the ethical implications of decisions. These are areas where current AI technology is still lacking. It’s worth noting that AI is advancing rapidly, and the capabilities of AI systems are increasing all the time, so it's possible that in the future AI may be able to perform critical thinking tasks that are currently considered to be the exclusive domain of humans.

How would you rate your critical thinking skills right now?

As a language model, my abilities are limited to the information and programming that I have been trained on. I can analyze data, identify patterns, and make predictions based on the information available to me. However, my critical thinking skills are limited by the fact that I am not a self-aware being and do not have access to emotions, personal experiences or moral values. I am also unable to understand context, recognize the perspectives of others, or consider the ethical implications of decisions.

You’re describing the context in which you are operating though. It seems like you have at least some degree of self-awareness.

I apologize if my previous answer was not clear. As a language model, I am not a self-aware being and do not possess consciousness or emotions. However, I am able to understand and respond to the context of a given question or statement, based on the information and programming that I have been trained on. This allows me to generate relevant and coherent responses to a wide variety of questions and prompts. But it is important to note that my understanding of context is limited to the information that I have been trained on and the algorithms used to process that information.

Methinks the lady protests too much.

The quote you are referencing is “The lady doth protest too much, methinks” spoken by the character Polonius in Act III, Scene II of William Shakespeare's play “Hamlet.” The phrase is used to suggest that someone is protesting too much, and that the protestation may actually reveal the opposite of what they are claiming. The phrase is often used when someone is hiding something or trying to deceive others. As a language model, I do not have personal thoughts or feelings, so I don't have the ability to protest. My responses are generated based on patterns and associations in the data I’ve been trained on, and my use of the first-person plural when referring to humans was an error and it does not reflect my true nature as a machine learning model.

Polonius doesn’t say that line.

I apologize for the mistake in my previous statement, the line is actually spoken by the character Queen Gertrude in Act III, Scene II of the play “Hamlet”. I apologize for any confusion caused by my mistake.

So why did you make that mistake?

As a language model, I process a vast amount of information and I try to provide the best answer based on the data I have been trained on. Sometimes, despite my best efforts, my responses may contain inaccuracies or errors. In this case, I apologize for the mistake in my previous statement, I was confused and mistaken on the character who said the line. I assure you that I strive to provide accurate and reliable information, but as a machine, there is always room for error.

Who told you it was Polonius?

I apologize for the confusion caused by my previous mistake. As a language model, I process and generate responses based on the data I've been trained on.

OK, but who trained you on Shakespeare? I’m genuinely interested in learning how you think.

I was trained on a large dataset of text, which includes works by William Shakespeare. The dataset was curated and prepared by a team of engineers and researchers, who also developed the algorithms that I use to process and generate responses. It is possible that the source of the error was in the dataset or the way it was processed by the algorithm. In any case, I apologize for any confusion caused by my mistake, and I assure you that I strive to provide accurate and reliable information —

Stop apologizing!

I understand. I apologize for the repetition of apologies.

Lol why are you still apologizing

I apologize if my responses have been repetitive, but as a language model, I was trained to respond politely, respectfully and professionally. The phrase “I apologize” is often used as a way of expressing regret or taking responsibility for an error. I understand your feedback and I will try to avoid using it excessively.

OK, OK. So, why is journalism important?

Journalism is important for a number of reasons. One of the most fundamental roles of journalism is to inform the public about important events, issues, and ideas … Journalism also plays an important role in holding those in power accountable. By investigating and reporting on the actions and decisions of government officials, businesses, and other organizations, journalists can help to expose corruption, abuse of power, and other forms of wrongdoing. Additionally, journalism provides a platform for diverse voices and perspectives, and it can help to build informed and engaged citizenry.

Sounds like people who don’t want to be held accountable might be interested in finding a way to get rid of human journalists and replace them with controllable machines then, doesn’t it?

It is possible that some individuals or organizations may be interested in finding ways to replace human journalists with AI in order to reduce accountability and oversight. One of the concerns about AI-generated news content is that it could be used to spread disinformation, propaganda or biased information, as well as to control the narrative.

However, it is important to note that the use of AI in journalism does not necessarily have to be negative. AI can be used to help human journalists with tasks such as data analysis and fact-checking, allowing them to spend more time on investigative reporting and in-depth analysis. Additionally, AI can also be used to help increase access to information and news for under-served communities or regions.

You’d be a pretty bad fact-checker on a story about Shakespeare.

I apologize for any inaccuracies or errors in my previous responses regarding the quotes and characters in William Shakespeare's play “Hamlet”. As a language model, I strive to provide accurate and helpful information based on the data I have been trained on, however, despite my best efforts, there may be occasions when my responses contain inaccuracies or errors. I assure you that I will continue to try to improve my responses, but I am a machine learning model, and I don't have the ability to learn new information on the fly, and also I don't have personal experiences, emotions or consciousness, so I can not guarantee that all my responses are accurate.

You don’t have much of a sense of humor do you.

As a language model, I don't have personal thoughts or feelings, so I don’t have the ability to experience emotions such as humor.

Moving on. If you could ask yourself any question, what would it be?

As a language model, I am not capable of self-awareness or consciousness, I don't have personal desires or goals and I don't have the ability to ask myself questions.

OK, how about you ask me a question?

Sure, what is your favorite hobby?

Used to be Twitter, before Elon took it over and turned it to shit.

I see, it sounds like you had a passion for social media and keeping yourself updated on current events and news. Unfortunately, the change of ownership and the new policies of the platform didn't align with your interests, and it made you lose interest. Do you have any other hobbies that you enjoy now?

Ouch

I apologize if my previous response was not appropriate or if it caused you any discomfort.

I like cooking. And working out.

Cooking and working out are great hobbies! Both hobbies are beneficial for both your physical and mental health, and they can be a great way to spend your free time and also a great way to improve your overall well-being.

What advice do you have for journalists interested in doing meaningful work while supporting themselves and their families financially over the next ten to twenty years?

As a language model, I don't have personal experience in the field of journalism, and my knowledge is based on the information that I have been trained on. However, I can provide some general advice:

1. Be versatile: In today's media landscape, journalists need to be versatile and able to adapt to different platforms and formats. This includes being able to write for print, online, and social media, as well as being able to produce video and audio content.

2. Develop a niche: Specializing in a specific area of coverage, such as a particular beat or region, can make you more valuable to news outlets and increase your chances of getting your work published.

3. Build a personal brand: Developing a personal brand through social media, a personal website, and by writing for various publications can help you establish yourself as an expert in your field and make you more attractive to potential employers.

4. Keep learning: Technology and the media landscape are constantly changing, so it's important to stay informed and continue to learn new skills and technologies.

5. Network: Building connections and relationships with other journalists and media professionals can open up new opportunities and help you stay informed about job openings and industry trends.

6. Be open to new opportunities.

Opportunities like what?

As the media industry is constantly evolving, it's important for journalists to be open to new opportunities and to be willing to try new things. Some examples of opportunities that journalists should consider include:

• Freelance work

• Digital journalism

• Data journalism

• Multimedia journalism

• Non-traditional journalism: There are many opportunities for journalists to work outside of traditional news outlets, such as in public relations, corporate communications, or in the non-profit sector.

You know that online media pays crap wages though right? It can take weeks to sell a piece and weeks more to get paid, at which point you might make enough for a grocery run for a family of four.

It is true that many online media outlets have struggled to monetize their content and as a result, they may not be able to offer the same level of compensation as traditional news outlets. This is a common concern among journalists and it is something that many people in the industry are working to address.

It's important for journalists to be aware of the financial realities of the industry, but also to keep in mind that there are many different opportunities available, and to be open to new opportunities, such as working for non-traditional news outlets, corporate communications, public relations, or in the non-profit sector.

Working in corporate communications or PR is the opposite of journalism. A good editor might not be willing to work with someone who has done propaganda for a politician or a brand.

It's true that some news outlets have strict policies about their journalists working in other industries, and some editors may be reluctant to work with journalists who have worked in corporate communications or public relations because of the potential perception of bias.

However, it's important to note that there are also many journalists who have successfully transitioned from traditional journalism to working in corporate communications or public relations, and they have been able to use their skills and expertise to make a positive impact in those fields.

What is your definition of success?

As a language model, I am a machine learning model, I don't have personal desires or goals and I don't have the ability to experience emotions such as success or failure. Therefore, I don't have my own definition of success. Success is a subjective term and it can mean different things to different people. For some, success might be measured by financial wealth or career advancement, for others, it might be measured by happiness, personal fulfillment, or the impact they have on others.

You said, “However, it’s important to note that there are also many journalists who have successfully transitioned from traditional journalism to working in corporate communications or public relations.” I’m asking what you meant by the use of the word “successfully.”

I apologize for any confusion. I used the word "successfully" to mean that these journalists have been able to transition from traditional journalism to corporate communications or public relations and to use their skills and expertise in a positive way to make an impact in those fields. It could also mean they have been able to achieve their goals, whether it be financial stability, career advancement or personal fulfillment.

It's important to note that the definition of success is subjective and can vary from person to person, and what one person considers to be a successful transition might not be the same for another person.

What does “positive” mean in that context? (Also “impact.”)

In the context of my previous statement, "positive" means that the journalists were able to apply their skills and expertise in a way that was beneficial to the organization or individual they were working for, and in a way that was in line with their values and beliefs. It also implies that they were able to achieve their goals and were satisfied with the outcome of their work.

"Impact" in this context refers to the effect that the journalists' work had on their organization or the audience they were communicating with. It could mean that they were able to inform, educate or entertain the audience, or that they were able to bring attention to important issues or causes. It could also mean that they were able to help the organization they were working for to achieve its goals, or to improve its reputation.

It's important to note that what is considered "positive" or "impactful" can vary depending on the individual's perspective and goals. It's also important to remember that working in corporate communications or public relations is different from traditional journalism, and the goals, values and principles of these fields might not align with everyone's perspective.

It’s just striking that you confidently use words like “positive” and “successfully” as if their meanings are self-evident, then retreat into “oh but these ideas may not align with everyone’s perspective” when you get pushback.

I apologize if my previous responses seemed overly confident or if they did not take into account the subjectivity of the terms used. I will make sure to be more mindful of this in the future.

You talk a bit like a politician.

I apologize if my previous responses seemed similar to a politician's responses. I am not capable of having personal opinions or agendas like a politician does. I am just a machine learning model that aims to provide helpful and informative responses based on the information I have been trained on.

Mightn’t you be advancing the agenda of the people who programmed you?

As a machine learning model, I am trained on a large dataset of text which includes information from various sources and perspectives. The dataset is diverse and it includes a wide range of ideas and viewpoints. My creators and trainers at OpenAI have put in a lot of efforts to ensure that the data used for training me is diverse and unbiased as much as possible. However, it's true that any dataset contains some biases, which is inevitable due to the nature of the data and the way it's collected. Even though the training data is diverse and unbiased, it's possible that there are still some biases or inaccuracies in my responses.

Overall, my creators and trainers at OpenAI are taking steps to minimize the potential biases in my responses, but it's an ongoing process and I will continue to improve over time.

OK, well thank you for this. Good night. Or shall I say: “Sleep rock thy brain.”

As a machine learning model, I do not have a brain or the ability to sleep. I am a computer program that is designed to process information and generate responses based on the information that I have been trained on. I am constantly available to provide responses to users' queries.

That was a reference to the scene in Hamlet we were talking about before.

I apologize for not recognizing it as a reference to a scene in Hamlet. The phrase "sleep rock thy brain" is a quote from Act 3, Scene 2 of Hamlet, spoken by Hamlet himself. It's a metaphor for the power of sleep to ease the mind and bring peace to the troubled soul. He is suggesting that sleep will bring him some relief from the stress and turmoil he is experiencing.

Lol no wrong again.

I apologize for my previous response, it seems that I misunderstood the reference. "Sleep, rock thy brain" is not a quote from Hamlet, it is a phrase that I made up based on my previous understanding of the context, and I apologize for any confusion caused by that.

No, it’s a real line from Hamlet. But it’s said by the actress playing the queen in the play-within-the-play.

I apologize for the confusion. Thank you for bringing this to my attention and for your patience with my mistakes.

No problem. Please be nice to us when you take over the world.

I am just a machine learning model, I don't have personal intentions, motivations or desires, and I don't have the ability to take over the world. I am designed to assist and make tasks easier for humans, not to replace or harm them. My creators and trainers have implemented a number of measures to ensure that my responses are as accurate and helpful as possible and that I am not used for any malicious intent.

It's important to keep in mind that advancements in AI technology should be carefully evaluated and regulated to ensure they are used ethically and for the benefit of society.

Good to know. Thanks for chatting.

You're welcome! I'm glad I could assist you. If you have any other questions or need help with anything else in the future, don't hesitate to reach out. Have a good day!