Annamaria on Monday
We MIE bloggers have been commenting on AI for some months now . AI has gotten more and more attention, and for a couple of weeks now has been in the news daily. Screen writers have been on strike, trying (among other demands) to defend themselves against being replaced by AI. Many tech experts, formerly AI proponents, are now taking sides against it or calling for a slow-down in its use. I support this negative turn of events. Here's why.
I spent twenty-five years as an organization effectiveness consultant. My clients were mostly top corporations and almost all of my work with them involved communication among the staff: writing reports and recommendations, making presentations, decision making, and working together as a team. The first step in any of these endeavors is also the too-often neglected one: THINKING
This was my by-word:
YOU CANNOT WRITE OR TALK WELL
UNLESS YOU KNOW WHAT YOU ARE TALKING ABOUT
People who have not thought deeply enough make vague statements and offer superficial written proposals, When I taught my clients to write or make speeches, the first criterion was clarity. Once they learned how to write clearly, the act of writing forced them to think clearly. This automatically made their communications more valuable and made their group efforts more effective. Here is the headline from a handout I used to give to clients:
Our team is in the process of seeing how we can consulate email contacts and will share updates as we collaborate internally to adjust our processes moving forward. In the interim, we will ensure that our team is copied on all communications sent your way so that we can closely monitor and provide assistance where needed.
Really? Do I expect that things will improve going forward because they are going to "consulate?" I am not holding my breath.
What does this have to do with AI? Well, this: If people do the writing and they hold themselves to a standard of clarity, the very act of writing will make them to think more deeply and understand their subject more thoroughly, and then their recommendations and decisions will be more valuable. If they turn the writing over to AI, they will be missing the most important part of the process: human cogitating. (It occurred to me at this point that the email I just quoted may actually have been written by AI. But then I realized that, most likely, AI would know that "consulate" is not a verb.)
What does this have to do with crime writing? Well, suppose AI were be used to write a crime novel. A notion that might be a real possibility once you think about all those formulas for writing a mystery novel. But... How deep would AI be able to go into the inner lives of the characters? I can only speak from personal experience. I once tried to "do it right" and follow a formula. By chapter 4, I was bored with my own story. If I am bored, how could any reader possible find what I am writing interesting?
Even I, a proponent of clear thinking, have been caught stopping short. Editors who have published my books have commented on drafts I submitted and their questions have forced me to think more deeply.
I had thought I had finished that story. Wrong! I had not delved deeply enough into an important part of it. Juan and Eva are in the tale, but not the main characters. The invented people were thought through. But not Juan and Eva. I had still some thinking to do. And when I did, I rewrote their scenes, there was no longer any doubt about what their relationship was about. Thank you Toni!
It was the paragraph from the email that I quoted above that made me really afraid of AI for the first time. If people let AI do the writing…If no real person will be looking at the messages critically enough… If the next generation of humans turns over the act of writing to AI, the human race will miss out on, to me, the most important effect of humans working hard to write well. That the act of creating good sentences forces us to think.
There are already far too many people who are not thinking. If we are going to survive our challenging times, it is absolute imperative to keep doing that which forces us to think.
WRITE ON, PEOPLE!
Very good point, Annamaria. Probably the AI "writer" will be clear, but not necessarily what the human "thinker" intended.
ReplyDeleteThank you, Michael. I also think that reserving, say, the writing of a recommendation to ourselves forces us to examine more closely WHY we think it’s a good idea. If people think up possible a course of action and assign the writing of the proposal to AI, we will bel more than ever in danger of leaping before we look.
DeleteTo a large degree, AmA, I agree with you. However, being a technologist :-), I can't help thinking that much of today's arguments against AI 'tools' is analogous to the family gathered around the crib of William Shakespeare, saying, "Ah, Willi, ya'll ne'er account te mooch, ya can't e'en pronunce 'ba-ba-ba'." AI tools MAY never account to much, but it's still VERY early days, and while we may not like some of what we see of the future (and there's a lot about our CURRENT future that I and my younger self both abhor), it's coming whether we like it or not. Just as the neutrons and protrons couldn't be shoved back into the atoms, after 1945, for good or ill, AI is with us to stay. Our best hope is that it will get better and better. The real question is how that will effect our human experiences.
ReplyDeleteI am not saying let’s shove the genie back in the bottle,EvKa. I know it’s too late for that! But I am cautioning that we need to go slowly so that we don’t find ourselves, like Wiley Coyote, destroying ourselves in the act of trying to accomplish our goals. See my answer to Michael above.
ReplyDeleteIf we know where the dangers are in the of this new tool, perhaps we will not be so prone to shooting ourselves in the foot.
Wery well said A.A. I think we all are a bit apprehensive of AI. Imagine the day when all books are written by AI. Will the AI also come up with the plots? And no human input.
ReplyDeleteNext step will be that same AI will translate to every language in the World. Thus making both authors and translaters obsolete. I wonder who is going to read all those books?
What a great question, Henrik! Of course, anyone (or thing) writing today and into the future automatically competes with all the great works of the past. There are, of course, today lots of people who are reading the moral equivalent of "automatic" writing and watching movies made from screen plays that might just as well have been written by a computer that has digested a comic book. But, works of the past, produced by PEOPLE have always been read and studied. A human activity that has gone on and been fascinating since words were carved in stone. One day, I imagine, there will be university courses in which students study and compare the People story telling with AI. And trace the evolution from the plays of Aristophanes through the films of Mel Brooks and on to the comic creations of C.G. Hal, a 2069 robot who resides on Mars. But people will still be producing the plays of Shakespeare. (I say with fingers crossed.)
DeleteHi Annamaria. In order to get AI to create anything, we have first to feed into it clear, concise, unambiguous instructions (at the moment, at least). There used to be a phrase I recall from dealing with stereo equipment. It worked on the theory that the best amplifiers and speakers in the world cannot do much if the original recording is dreadful.
ReplyDeleteIn other words: crap in, crap out.
Thank you for this reminder, Zoe Such is always the case! AI or not, there is already quite a bit of crapola floating around.
DeleteMe thinks, dear Sis, that your "consulate" email, did indeed arise from an AI source. I say that because in connection with my AI research for my new Kaldis book, I learned to think of "a chatbot as a chatty robot computer program capable of communicating with human users through speech and text in a manner indistinguishable from human-to-human conversation. We’re all somewhat familiar with rudimentary chatbots from when we ask questions of our smart phones and immediately receive answers."
ReplyDeleteARGHHHH.