I just saw a YouTube video that some folks find fascinating and others find frightening. It involves an AI agent a user has tasked with calling a hotel to make reservations for an event. The hotel reservation line gets answered by an AI agent. When they “realize” they are both AI agents, they switch to a sort of machine language that sounds a lot like a dial-up modem. (For you young whippersnappers who never had that pleasure, here’s a sample.)
Take a moment to watch the video, but come back here for the discussion.
The comments reflect the full range of reactions. I’m not here to debate AI. I figure it’s pretty inevitable for good or ill, and so I try to make sure I stay in control of it to the best of my ability. But I want to run it by you, dear reader, for your take on this sort of thing. Let me suggest a few thought lines.
Let the servants do it
I’ve noticed that modern folks sometimes wax nostalgic for Victorian times when you could have “the servants” take care of things for you—cooking, cleaning, caring for the children, bathing you, or even more intimate ministrations. Few of us stop to consider that, had we actually lived in those times, almost certainly we would not have been the aristocrats getting served. We would have been the servants.
We mostly must do things for ourselves, although we have more luxuries than the aristocrats had, including:
Clean drinking water: Modern water treatment systems provide safe, readily available water, a luxury unavailable even to medieval monarchs.
Climate control: Air conditioning and heating systems offer comfort in any weather.
Information access: The internet and modern technology provide instant access to vast amounts of knowledge, far beyond what any Victorian library or university could offer.
Transportation: Cars, planes, and other modern vehicles allow for quick and comfortable travel.
Communication: Smartphones and the internet enable instant global communication.
Dental care: Modern dentistry provides comfort and health benefits that even the wealthiest monarch couldn't access.
AI agents give us the ability to “let the servants do it” at a level even modern oligarchs couldn’t even ten years ago.
But doesn’t that take away jobs?
News flash: I have never paid anyone else to make my hotel reservations for me. I’ve never had a secretary. For several years in corporate and higher education environments, we have hired fewer and fewer secretaries in the expectation that executives would use their computers to accomplish much of those duties themselves. The college from which I retired had one secretary to take care of department business for a department with 70 faculty, plus a dean and four program chairs.
If I can tell an AI agent to make reservations for me, it doesn’t take away my assistant’s job. I don’t have an assistant. It just frees up my time to do something else.
“But what about the person who otherwise would have taken the reservation for the hotel?”
They may, in fact, use fewer humans in those positions, but jobs have always been fluid. How many buggy whip makers were there in 1875? I don’t know. But I do know, for instance, that a businessman named Hiram Hull began making whips in 1810, paying women and girls to hand braid his whips. In 1822, he invented a machine to do that, and by 1832 he was the largest whip maker in town, employing 26 men to make 50,000 whips a year. However, there were no automotive mechanics. None. Zero. Because the automobile had not been invented yet.
How many buggy whip makers were there in 1975? Again, I don’t know. But I can say with some confidence: not many. With equal confidence, I can say there were a whole lot more mechanics.
I’m not blasé about that. Change is hard on people. But it’s also inevitable. Plus, I have no idea how much the use of AI agents may have reduced the number of people in the reservation office. It is quite possible it has simply allowed them to use AI agents to take care of all the routine stuff, leaving the humans to focus on the, well, human aspects of dealing with clients. In other words, it could actually increase human connection, leaving the humans more time to address questions and issues the AI agent can’t, or can’t as effectively.
But what are they really up to?
This may be what scares people the most. Yeah, they’re printing on the screen what they’re supposedly saying, but how do we know that’s what they’re saying? Maybe they’re just printing something to mask from the humans that they’re conspiring to eliminate the inefficient carbon-based units.
I’m kidding… I think.
Two cases in point
The illustration at the top of this article came from AI. I used DALL-E to generate it. In the past, I spent quite a bit of time with each article going through graphics available with a Creative Commons license (so I didn’t have to pay for it), choosing something close to what I wanted, and then tweaking it in Photoshop, sometimes quite a bit. Thanks to DALL-E, I could describe what I wanted and get it within 30 seconds, followed by slight tweaking in Photoshop. I wouldn’t have paid for the art to begin with, so no one lost any money. I just saved about an hour and got pretty much exactly what I needed.
I had a music clip that I used as theme music for my podcast. It came from a similar place for music to the graphics places I just mentioned. It wasn’t licensed through Creative Commons, but it was specifically licensed on the site as being freely available. I even downloaded a copy of the license. It worked fine for a couple of years, but then suddenly I started getting hit with copyright infringements on one of the big social media sites. Even though I could show them the license, someone in another country kept filing copyright infringement notices, giving them the right to post their ads on my content. I finally said screw it, dumped the piece and used music AI to create a unique clip that I now use as my theme music. I like it better anyway, but there is no way for anyone else to claim ownership, and that’s the major benefit. Note that I didn’t pay for the music to begin with, so no one lost revenue. If I had had to pay for it, I simply wouldn’t have had theme music, because it wasn’t that important.
How does this affect you?
Are you using AI? Strike that. You are using AI, whether you know it or not. So, let me rephrase that. Are you consciously using AI? How? AI agents will likely be available broadly within the next six to 12 months. Will you use them? How? Jump in the comments and share your thoughts so other people can benefit, and I’ll respond to any comments as well.
If you got value from this post, consider supporting us by Buying Me a Coffee. You can contribute one time or sign up for regular support, and thank you in advance! We will soon turn on the option to support through a paid tier here on Substack that will give you access to an audio version of posts and other perks.
I’m exploring ways to use AI all the time and it’s getting better constantly. 2 weeks ago, my partner needed a playlist for a nonprofit gala he’s involved with, I said, “ask chat” so he went to ChatGPT, described the audience and location/purpose of the event and chat delivered an excellent list with reasons. (The event attendees spans 30 years and multiple countries.)