The Last Twenty Percent

The value of humans in a machine-run world

Hello, Beautiful People!

Today is the start of a new occasional series I’m calling Voice Memos. It’s not a podcast, because I don’t want to be that formal. Instead, I’m sending out occasional voice memos that take you inside my weird brain. These periodic recordings will show you what I’m researching and thinking about. I will also talk a little about my process and the art I’m enjoying. You can find a transcript below the recording.

One last piece of business before I let you read the comic and listen to the voice memo. I am moving away from a formal schedule. I will be posting at least once a week, and most weeks, I will be posting two or three times a week. I will post a comic and haiku prompt every Thursday. But, I’m going to let the flow of the work dictate when other things get posted.

the algorithmic
bounty hunter only takes
crypto currency

And now for the first Voice Memo:

Push Play ⬇️

Please let me know in the comments what you think of the new series!

Transcript

Note, I used Otter AI to transcribe this audio.

Welcome to the very first voice memo. My voice is a little bit rough. It's because we just got back from the Taylor Swift concert in Seattle yesterday. And I did a lot of singing, hooting and hollering. Today I want to talk about this idea of the last 20%. It's something thinking about in terms of what role do humans have in an AI world. And it's also related, in my mind, at least, to music producer, Rick Rubin and his podcast. So AI has been on my mind a lot as has been on everybody else's mind.

I've been a copywriter for like 11 years, maybe 11 years ago, I started copywriting, it'd be a better way to put it because I took pretty much this last year away from doing any copywriting. I was really burned out, had some family things that we were working on my wife's health, and I wanted to explore some other things.

And AI was just starting to come onto the scene. And I just had this it was pretty obvious to me that I was going to fundamentally change copywriting and I wasn't sure if I wanted to deal with that. Now, fast forward a year later, I am copywriting again. Because mostly because we have a financial need. Turns out poetry doesn't really do as good a job as paying the bills as writing marketing copy does. But also, I've had a chance to see how AI is going to play out. And in terms of copywriting, I really think it's just replacing the bottom end of the market. And that's just not really where I work.

But there's some other things that I kind of have to be aware of, for example, some clients really want to run things through an AI checker. But it turns out the AI checkers are completely garbage. They vary as much as like 30% as to their score, they're gonna get they're going to give you I found out that AI checkers, you know, this is a an algorithmic driven check on whether AI was used to write the piece of content, right. And so I did some testing and some of them, if you copy and paste from Word, you get different score than if you copy and paste the same text from Google Docs, and so on Google Docs gets a worse AI scores more likely to be generated, you can use the exact same text, if you use Grammerly to help you edit the text, Grammarly is also an algorithmic tool or an AI driven tool. You actually that you tend to get a worse score as being more likely to be aI generated.

So it's really fascinating to me. The way that it's affecting the business, but like I say, the kind of copywriting is specialized, but my specialty is I really, I generally work for more expensive clients. So I'm not like, you know, writing blog posts or $5, or anything like that. In fact, I really almost never write blog posts. What I mostly write is I write marketing emails, and I write video scripts for companies. And those least at this point, aren't things that AI is actually very good at.

And then, of course, I am an artist, I'm a visual artist and poet, essayist, and, you know, it is troubling to me the way AI is being rolled out without really any thought and poorly clearly was been created unethically. I think all the image generators were scraped pictures off the internet, they did not have the copyright to and I think open AI did the same thing for for texts. It doesn't it's, I think really clearly violated copyright. And so I think that's, you know, technically illegal but it's also immoral, in terms of not paying creators for using their work as fuel for these for these digital tools.

Sarah Silverman, comedian, writer is suing open AI for copyright violation. So I'm curious to see how that plays out. But despite I think that The fact that I think these these tools were unethically created, they're here to stay in one way or another, you know that they they're not going away. And so I think it's incumbent upon people to decide what do you make of that fact. So for me, I can see AI being valuable tools. Potentially, they could be used to write the first draft of some of a blog post or some type of copywriting thing. Right? I am not super convinced that they actually going to be great for writing a first draft of a novel or anything like that. I just don't. In the test that I've done and the work that I've seen, I'm, I am, I am not convinced of that. But potentially, maybe they'll get there, I guess, right, they're growing exponentially.

But the problem with AI, is that it has to have human fuel to continue growing. If you feed in an artificial intelligence system, the work product developed by it or another artificial intelligence system, you break it. Because they're essentially the one of the ways to think about these large language models is the kind of like really fancy text predictors. And so if they're not getting human input, constantly, they're not going to get closer to writing, like humans, or drawing like humans. Instead, they use this, it gets this kind of feedback loop. If you continue to feed it AI gobbly gook. And you just the output, you're gonna get as strange. You'll increase the number of what AI experts call hallucinations, which are not really big fan of that term. For essentially, lies, right, like if you ask the AI to create something with sources or to create some very convincing sources that don't exist. And the technical term for that, that we've decided on, I guess, is hallucination. Which I think is a quite the sugarcoat quite the whitewash of of what's actually happening there. But like I said, I think AI is still here.

So what does that mean for humans? I think that in terms of like content, right, like, blog posts, and things like that, humans are always going to be needed to get it to the last 20% at least right? To bring it from like a 75% C product to a 95% A level product. Because AI just lacks empathy, it can get the grammar, the grammar can be perfect, but that's not really, grammar is not what makes something compelling to read.

I think the human touch excellent human writers are always going to be needed even no matter how far AI goes to get some kind of work product, fiction, nonfiction, you know, content, art, whatever, to the last over that last 20% Where, where it's really going to be a solid, great reading experience. But a lot of companies are going to be happy with a C level work product, right? They're gonna be happy with that 75% And so a lot of writers I think you're gonna find themselves in need of a finding other work.

And then I was thinking this idea of 20%, the last 20% of something, I'm really falling in love with the idea that last 20% being what makes something from good to great or something from mediocre to good. And so there's this music producer named Rick Rubin. He's very famous, in lots of circles. He kind of helped discover the Beastie Boys. He's done some legendary recordings with Johnny Cash and the last ones that Johnny Cash ever did. He's worked with all kinds of of artists, Tom Petty, all kinds of artists love to work with him. He's kind of a kind of a Zen vibe. He's also somebody that I'm fascinated with music by the way, if you don't know I am obsessed with music, listen to it all the time. I love learning how the sausage is made. So I've known a lot about Rick Rubin for a long time. I've read a lot of stuff from him watched a lot of interviews with him and he's just somebody that I don't know exactly what to how to say it but he just somebody that has never I don't really fully like him and I don't really have a rational basis for that.

The only way I can really describe it, he's the feels a little too Libertarian for me. But he definitely does some great work. He has some really interesting thoughts on on creativity and he has a new podcast out. And I'm going to forget the name of it Tetragrammaton or something along those lines. And I have had mustard for a long time, just because like I say he if something about him that just, I find a little bit off putting I think also like he's a little bit too willing to his touch false humility he feels to me and maybe too willing to accept his genius reputation.

But again, that's, you know, I don't know the guy. So I had avoided podcast, but it had been recommended enough times by people that I whose opinion I respect, and then other podcasts, I listen to talk about greatest podcast. So I decided I listen to it. And it's really good. He is a an excellent interviewer. And one of the things he succeeds at is he doesn't talk too much. But there's something that makes his podcast, to me at least a particularly delightful experience.

And it's the last 20% of what he's done what he does. So it's a podcast, it has ads. They are the best ads I have ever heard. I listen to a lot of podcasts in my car. And if I am at a stoplight, and a commercials on or ads on the podcast, I will skip through it. And chance I get I skipped through the ads, especially ads in the middle of the podcast drives me nuts. But on Rick Rubin's podcast, he has produced the commercials to sound like they're an old timey radio, commercial advert. But there'll be for something modern, like some kind of nutritional supplement or whatever. But like the audio quality is very, very much old timey radio. And old timey here I'm talking like 50s 60s sound to them. And I mean, the audio quality is phenomenal in terms of its replication of that sound, down to the music, the speed of the narration, the structure of the ads, even it's, it's, I actually enjoyed listening to the ads, it was an enjoyable experience. And you know, I am a professional marketer, I probably am more interested in advertising than the average person, but to get me to sit and listen through podcast ads, because that almost never for things that I have any desire to buy is a is a real feet.

And but not only is that great for him for Rick Rubin's advertisers, it's actually makes the whole podcast in general more enjoyable, it makes me more likely to listen to it, there's less of a barrier for me to listen to that podcast. And to me, it's like kind of the last 20%, right. That that takes something that he did that would be have been good, and really makes it a great podcast. And so I'm trying to apply that to that last 20% notion to my work and to what I'm doing.

I'm starting something a little it's a little bit scary for me, I'm I'm actually writing a novel that I am going to indie publish. It's, of course, it's weird because I am writing it. So that idea I've had for a long time. I wrote my first outline for this novel last year, and just kind of let it cook for a little bit in my brain. And I'm about 10, 10, 15% done with the first draft started writing it this week. After kind of revising my outline and thinking about it a little bit better, and it's gonna be it's called "Attack on Gorgon Station", it's going to be a space fantasy. It's this idea. The idea for this novel is essentially, imagine a world where you have elves, dwarves, humans, dragons, wizards, but they continued, they didn't get stuck in the middle ages, right? They proceeded and reached the space age and they colonized the solar system. So it's a it's a novel set in that kind of a setting. And I, I've been thinking a lot about the last 20% because there's there's lots of books out there, although I don't think there's a lot of space fantasy books, but there should be. So I've been thinking about what can I bring to this, this project?

That that gets it from a 75 to a 95 gets from 80 to 100. What can I do that can really make this great. And so that's something I'm trying to be conscious of as I write this first draft. Am I putting enough of myself into it? Because I think I want to. This is where it's going to tie back into this this again, thing of AI. I think that in a world of AI, there's going to be a large number of people who really want human made things.

And what can I do to put the most of myself to make this work that I'm doing as human as possible? And to me, that's what the last 20% is, at least for this novel that I'm working on. What what makes it human? What makes it the best work that this human, me can produce? So that those are kind of just my thoughts right now.

What I'm doing what I'm working on, I'd love to hear from you. This is a brand new thing for me. I'd love to hear from you what you thought about this voice memo. If it was too rambley if it was too long, if it was just right, yeah, definitely, please feel free to give me some feedback. In the comments, this is fun, like kind of I'd like to do this semi-regularly. Until next time, I'll just remind you to be the weird you want to see in the world. Bye.

Cheers,

Jason