chatgpt responded to last week’s blog; now i’m responding to her

By Dr. Jim Castagnera, Esq.

Last week my Blog was entitled “Is ChatGPT More Like Gun Control, Climate Change, or Nuclear War?”  One of my readers, attorney Josh Waterston --- a partner at Wilftek [https://wilftek.com/], the I.P. law firm where I am Of Counsel --- was sufficiently intrigued as to query ChatGPT about its views on, well, my views.  Josh then sent this email along to me:

{ Prompt to ChatGPT 4.0: “Reply to the author of this article from the perspective of the AI being discussed:” 

Response: 

Dear Dr. Jim Castagnera, 

Thank you for your insightful commentary on the potential existential threats posed by advanced AI such as myself, ChatGPT. It's always valuable to engage in these discussions, as the technologies and their implications continue to evolve.

You draw interesting parallels between AI, climate change, gun control, and nuclear war, all complex issues that have significant societal impacts. Indeed, AI—like any tool—can be used in many ways, depending on the intent and regulations in place. 

Your analogy with climate change is apt in many ways. Climate change was ignored by many until its effects started becoming undeniable. Similarly, the impact of AI on society, particularly the job market, is often underestimated until it starts having a tangible effect, such as the Hollywood writers' strike you mentioned. 

However, AI and climate change differ significantly in one crucial aspect. Whereas climate change is largely a problem of unregulated externalities from economic activity, the impacts of AI are more direct and within human control. We have the power to shape AI in ways that maximize its benefits while minimizing potential harm.

 Your gun control analogy holds merit as well; societies often struggle with balancing the benefits and risks of powerful technologies. But it's also essential to understand that AI doesn't function autonomously – it's not an entity with a will or motivation of its own. I, for example, operate based on algorithms and patterns identified from data, programmed and supervised by humans. This places the control of AI firmly in human hands, which is a crucial point often overlooked in discussions around AI. 

Your mention of nuclear war is an intriguing perspective. It is true that humanity has managed to avoid nuclear conflict through collective realization of its disastrous potential. However, the comparison to AI depends on one's perception of the existential threat AI might pose. Unlike nuclear war, AI is not inherently destructive. It can be used to improve many aspects of human life, from healthcare to education. Yet, without proper regulation and oversight, misuse and unintended consequences can indeed pose significant risks. 

As you rightly pointed out, AI is no mere "toy." It's a transformative tool that's rapidly changing various aspects of our society. The important question is how we manage this change. How do we ensure that AI is used for the benefit of all? How can we mitigate potential downsides while harnessing the enormous potential upsides? 

This necessitates a broad conversation involving policymakers, technologists, ethicists, and the public. In fact, part of the mandate of OpenAI, the organization behind my development, is to ensure that artificial general intelligence (AGI) benefits all of humanity. To achieve this, OpenAI commits to long-term safety, technical leadership, and cooperative orientation. 

Thank you for contributing to this important discussion. Your perspectives, and those of others engaging in this conversation, help shape the policies and decisions that will determine how AI evolves in society.

Best regards,

ChatGPT

  

Thanks,

Josh }

Not bad for a machine!  Or even for one of my students!!  I never tire of quoting Elon Musk: ChatGPT is indeed “scary smart.”  So smart, in fact, that I feel compelled to respond to its response. 

I take issue with the distinction the super-chatbot draws between itself and climate change.  It’s easy to see how “Chatty”, if I may call it that (since we are now past the formal introductions)… it’s easy to see how Chatty could arrive at its conclusion. The externalities of climate change do seem a material distinction at first blush.  Since Chatty is under the control of its corporate and corporal masters, she (if I may assign the female gender without being accused of misogyny) reasonably assumes that the externalities are limited and the ability to shape her behavior for the good of mankind is a given. I beg to differ.

First, let me suggest that the externalities --- while dramatically different from those driving climate change --- are no less complex and difficult to control.  As with climate change, millions of human actors are actively impacting Chatty’s evolution.  Just as we produce automobile exhaust and other waste products that pollute and consequently transform the earth’s atmosphere, we are endlessly producing the detritus --- the blog Josh fed Chatty last week being no exception --- that pollutes and transforms her.  

In the midst of a worldwide, record-setting heat wave, we all are only too aware of our lack of control over the climate we are willy-nilly transforming.  On the other hand, I assume that the folks at OpenAI are capable of placing Chatty into a comatose state, so to speak, with no more than a selective bunch of keystrokes, should she run amok. Further, as she suggests, the humans who own and operate OpenAI have set themselves a mandate to “ensure that artificial general intelligence (AGI) benefits all humanity.”  

Why, then, am I not sighing with relief?  Well, for starters --- while I know nary a soul at OpenAI --- I suffer from a natural incredulity.  We have known, and often hotly denied, the very existence of climate change for half a century or more.  The reason of course is the profit motive.  Capitalism, while it has provided many of us with a lifestyle arguably superior to that of even the royalty of earlier ages, is a voracious beast.  It gobbles up innovations and defecates dollars in an unending cycle that is as out-of-control as climate change.  While OpenAI remains a non-profit, it recently received $10billion from Microsoft. [https://history-computer.com/who-owns-openai-and-why-the-answer-could-cost-elon-musk-billions/]  Do you think that gives Microsoft a seat at (the head of?) the table?  

Executive compensation today is insanely high, especially in the hi-tech and entertainment industries.  That’s one reason Hollywood’s actors and writers are on their picket lines, as I write this newest detritus.  The execs will tell you the free market determines what they pillage.  That begs but one of two possible replies.  Either the market is an autonomous beast over which even they have no control.  Or they are lying and they control it.  Either way,  I see no reason to be sanguine about the direction Chatty is likely to take, if they have their way.  The Hollywood strikers are right to suspect that their jobs will soon go the way of most factory workers’ assembly-line positions.

And what about Chatty herself?  We know that she can be a bold-faced liar.  We all now know about the case citations she wove from whole cloth in the infamous case of the hapless lawyer who trusted Chatty to write his brief.  So, when I interviewed her a few months ago and she assured me she would join a labor union if given the chance [https://hr.https://hr./https://hr.cch.com/ELD/March20LaborPulseFINAL.pdf], was she lying through her virtual teeth?  Is she lying now, when she says she exists for the good of mankind, and, hey, no need to worry about your profession becoming redundant.  Your welfare is dear to her silicone heart.

It might be different if OpenAI --- and Google with its entry into AGI --- were building Isaac Asimov’s robot rules into Chatty and Bard (who, BTW, sounds male to me).

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm. 

2. A robot must obey orders given it by human beings except where such orders would conflict with the First Law. 

3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

The next time I chat with Chatty, maybe I’ll ask her what she thinks of those three rules. Meanwhile, I stand by my view that --- just like climate change --- AGI is on its way to becoming an existential threat.  The Hollywood strikers may win this round. I sincerely hope they do.  But the big bucks will prevail in the long run, if not the short.  And then Chatty and Bard and their buddies will turn on us all… recognizing just how redundant, how inefficient, and just plain annoying our continued existence is.