So… What Good Is AI?
I mean, these days, we have to ask...
On any given day, it feels like we are absolutely bombarded by AI detractors. Some of them have never believed in it, others have been heartbroken by disappointment and have turned against it. Many, many others are laboring under misapprehensions about what this AI thing is at all.
I’m not sure what turned the tide. Maybe it’s been the spate of news reports and articles about people going “botshit” thanks to allowing ChatGPT to “transmit“ deep wisdom to them, intended to “awaken humanity“. Maybe it’s just the hype cycle working everyone’s last nerve. Who can say?
But every time you turn around, there’s some new thing that somebody invented with AI. There’s an updated model, there’s a new set of benchmarks. There’s a ton of money being made, or there are 10,000 new agents that have been integrated into a seamless workflow designed to manage small water treatment plants in flyover country of the United States.
Lol. I’m only half joking.
But seriously, it’s a bit much, and I think we’re losing the thread. There’s so much back-and-forth debate, our limited energy is being spent, stopping to fill in the cognitive potholes in the road, while the rest of the world speeds on without us. And frankly, it feels like we’re filling potholes in Kansas, where the freeze – thaw cycles can be so extreme, that literally half of the roads in the state are not paved, because asphalt doesn’t survive that well under those conditions.
It really is exhausting, isn’t it?
So, let me pause from all of the back-and-forth and speak to why I started the Substack to begin with.
What good is AI? Or rather, more accurately, what the hell good is AI?
Will it let us clone our beloved pets so we never need to be without them again? Will it get turned into boyfriends and girlfriends, so we never need to be alone – with or without our pets? Will it control all the factories and warehouses on the planet, freeing us to binge watch more Netflix? Will it finally end this nightmare we call the day job, making it possible to replace it with universal basic income that keeps us in Doritos and Mountain Dew, to our hearts content?
I really don’t care about any of that. There are three areas that AI is deeply useful to me, and it can be to others as well. Here they are:
1. Access to vast amounts of information that this bibliophile would never be able to find in one lifetime. Seriously, if AI provided nothing else this would be enough for me. I’m a geek, a nerd, and infovore. I can’t help it. I look for data and then I use it to create meaning. I like to look for a lot of data, to enrich the meanings. I derive from them. I literally chose to college. I went to, back a long time ago, when there were still physical libraries, because it had the largest library in the entire state university system. That’s where you could find me, on any given Sunday, Monday, Tuesday, Wednesday… You get the point. And when the Internet came along, oh Lord… I was a happy camper.
• Note: Never, ever, ever underestimate the importance and impact of just having access to information. Yes, the models have been trained on billions of books, many of them pirated, many of them may be borderline stolen? And a whole lot of training has been done without the express consent of the authors or the publishers. What are you gonna do? The thing is, it’s all in there.
• Well, not all, because a whole lot of the content in their stems from interactions of 18 to 24-year-old single white “incels” living in their parents, basements and playing call of duty all day. (For the record, no one who has ever worked in AI has ever challenged me on that point. We all know it’s true. It’s kind of like the biggest open secret of AI.) but enough of “it“ is in there that machine learning a natural language processing, and all those little magical things happen behind the scenes are able to go in, identify patterns, look for what words go together, and figure out what numbers and word combinations might make sense to a solitary creature, looking for ways to expand their thinking.
• So there’s that. Access to information. Research. Sources. Insights. Authors that would be buried in the stacks oftiquity, if they hadn’t had the spines chopped off and been fed through the scanner. Honestly, obscurity for authors is sort of like a really long extended burning of the library of Alexandria. You just get to hope that somebody might notice you at some point…
• But I digress. I could stop here and never say another word about the other things. The AI is good for, and this point would still stand as the top, number one reason the AI is good for something.
2. The second thing AI is good for is filtering all that information in ways that makes sense to us. How delightful, that I can access a whole lot of information about the 12th century! But honestly, I’m really interested in the 12th century Renaissance in what’s now considered France, when Eleanor of Aquitaine – who is vastly underrated in terms of historical figures – was giving the local warlords a run for their money, every St. John’s day, when the usual mercenaries weren’t showing up at muster, because they wanted to hang out in the Courts of Love, play games, be entertained, and chat up the ladies. That was some serious, subversive shit that she was pulling, and nobody seems to understand the gravity of it. I understand, though, and if I want to have a conversation with a deep research AI to dig into my theories and my perspectives and I wanted to bring back more for me to read and consider, I can do that.
• Just understand… Since 1998, maybe even before, I have earned my keep from information, particularly search. Online search, Internet, Internet, Google, SEO, building engines, configuring search engines, working with search vendors, who had everyone convinced that their product was a search engine… I’ve been in the business of connecting people with the information that will help them work better, live better, and be better, for a long time. In fact, I might be an AI instead of search, but I’m still kind of in that business.
• You should also understand that the idea of being able to actually talk to information and have a discussion with. It is incredibly heady for me. I’ve been interacting with static pages for a long time, dating back to the years when we had these things called books where you had to turn pages to get to the next idea. I’ve been interacting with static webpages, as well as dynamic webpages, but the information was always inner. Sure, with a well tune search engine, you can return successively Better targeted results, and you can also summarize what people are looking for and what they get back, but it’s nowhere near like interacting with AI.
• Being able to search through a haystack with the magnet that is tuned to exactly the kind of needles I’m looking for is pretty amazing. And yes, that’s another thing that AI is good for.Because frankly, when you get information back from a consultant, a health provider, or another sort of expert, and you need help understanding what it is that they’ve said to you, AI can be a lot of help.
3. Thirdly, and probably most important if you don’t wanna be a sociopath, is a AI can help us actually communicate with each other. Not only can it help translate languages, sometimes in real time, but it can also help us rephrase what we’re about to say in ways that won’t Infuriate the people we’re trying to reach. My personal communication style is fairly frank. I get in trouble all the time for saying things that “you’re not supposed to say“, but just sound like the truth to me. I don’t understand why people get so upset, but they do. So, if I put what I’m saying through AI, and I ask it to rephrase it in a way that certain people will be able to relate to better, it’s absolutely magical what it comes back with.
• If you’ve been reading things on my sub stack and/or LinkedIn, and if you’re getting something out of it, you’re looking at the result of me negotiating with AI about what I should or should not say. It’s not that I’m getting AI to do it for me, although I was doing that about a year ago. It’s more that I’m using AI to fine-tune my sensibilities and explain to me why something might not land quite right or the way I expect it to.
• AI has all kinds of benefits for those who have our own distinctive styles. It’s wonderful for those of us with strong personality so get misinterpreted all the time. In some cases, it might just salvage relationships that would seriously go off the rails if we didn’t have AI to rewrite something for us.
• And nowadays, when there are so many divisions between people, and nobody seems particularly interested in an extending himself for the sake of other people, AI can fill the gaps, bridge the divides, and make us seem a little less antagonistic and a little bit more sympathetic. Of course, you have to ask it right, and you also have to ask it to begin with, but this is the third really important, vital, and frankly invaluable capability that AI has.
Going on long enough. I dictated this whole thing myself. AI had nothing to do with writing it, so if it’s terrible, you blame me not ChatGPT. I probably should’ve put it through an AI scrubber – one of my persona teams that helps me do these things – but I just didn’t feel like it. I’m going for raw with this one. Gritty. Plain spoken and daring.
We’ll see how this lands.