Don’t be a meat proxy: use AI to make your messages shorter

0:37Scott

Quick little story. Yesterday I sent you a Slack and the Slack that I sent you, I sent it to you from when I was inside of Claude. And so I had done this whole thing. There was this whole entire thing with Riverside. Riverside is a platform that allows you to do some recording.

That kind of stuff. It's a thing we're recording this on right now. And I was looking to do some things in it. Colleen was like, "Hey, can you do this?" I was like, "Sure, let me go into Riverside.

Let me get it for her." And that led to one thing, the other, the other, the other. And it was just like, come on, just snags all along the way. It was like 20 things, you know? And I was like, this is so annoying.

But I was like, had this one long chat with Claude about it. And then at the very end of it, I just said, "Can you tell Colleen about this? Don't talk her ear off. Just like tell her briefly, just tell her briefly like what happened and then give her the username and password that she needs." And Claude's like, "Yeah, no problem.

Here's what it'll say. Do you want me to send it?" And I was like, "Uh, change these two things and then you're good." And then it's like, cool, send it. And a couple, like one thing happened to me, like when I did that, I was just like, I need to do this consistently.

There is this communication tax. Like whenever you do something, you now, you've done all the things, you've thought it all through. And now I need to communicate what I've done to other people. That's like a whole other thing. Do you know what I'm saying?

What, I didn't have that with you in this situation with Riverside, because I was just thinking through the thing. And then I said, "Can you tell Colleen about this?" And it's just, "Okay, here's, here's the message Colleen." And I was like, "Okay, cool. Change a couple of things, get done."

And I didn't even have to go into Slack. I didn't have to do, I didn't have to think through what to say. I didn't have to, I didn't have that communication tax. I just had the strategy tax, meaning like I had to figure out the problem. And then I came to a solution and then I was like, "Oh my gosh, now I have to tell everybody about this.

I hate this. This sucks." And then I'm like, "Oh, wait, Claude can do this." You know? And so then I was like, "Oh, just have Colleen, just tell Colleen." And so it did. And then I felt like, "Oh, and then I realized this is, it was nice.

Like I kind of just skipped a step." But then that got me thinking, here's what I want to talk about. There is this new term, this new term in the world just came out, like hit the zeitgeist a couple weeks ago. Have you ever heard of this? It's called meat proxy.

Okay, this is fun. M-E-A-T, meat, like a human being has meat in their bones, you know? And here's what it is. This is a person who pastes AI output into Slack, group chat, anywhere, shoots an email, whatever, without adding anything to it. So it's like this person has just like, "I wonder how many whatever, or like can it do this?"

And then it's like, "Oh, here's a book for you to read." And then that person just copies and pastes it and sends it to you and says, "Claude said this, ChatGPT said this." That's a meat proxy. And then there's this massive gigantic response. And then the meat proxy goes, "Oh, that's a cool response.

I'm going to copy and paste that and then send it to somebody else." So you're just a proxy, but you're meat. So it's called a meat proxy. Isn't that great? It's

3:40Colleen

Perfect. It's exactly right.

4:03Scott

Along information. The other person at the very end is like getting bombarded with this. And it's like, "Bro, I didn't want any of this. What are you doing? Why are you giving me this?"

Like this person's like, "I can talk to Claude myself." Then It got me thinking, how has this always worked? And so the way it's always worked is there's always been this like pile of ideas that you would have in your head about something. And then you would have to do all the work yourself to whittle it down into something.

Like all the brain, your brain does the intelligence of saying like, "Oh, maybe I could do this, whatever." You know, you can throw some things out. You're like, "Oh, that doesn't, that's not going to make it into the final thing." Like, because you're doing the work thinking, you're doing the work, you're doing all the good stuff. And then you have one clear note and you send it to them.

Look how tiny this is. Look how tiny that is that they have to read. You don't have a lot to read. All of the work was done by you prior to. Now it's one clear note that's happened.

So then AI showed up and the shape flipped because now what happened is, is that you have six words. You skip your brain, throw it into AI. Then you go 600 words. This

5:09Colleen

Is so real.

5:14Scott

So this is what the problem with the meat proxy is, is that it's taken a traditionally, I have a million thoughts and I need to condense it into a single note and send it to you. Now it's like, I have a whittle tiny idea of a thought and then explosion of a thing. And then I'm going to put the burden of consuming all of it onto you. There

5:35Colleen

Was a moment last year. I know you got more slides, but I just want to like pause here and say there was a moment last year when you sent me an idea and you sent it as a podcast and infographic. Do you remember this? I don't know. There's like a five page output from ChatGPT at the time.

And I was so pissed. I was like, dude, just talk to me about the idea. I don't want to read all this. I don't want to consume all this. I would much rather spend that 25 minutes just talking about it.

Yeah.

6:01Scott

And it was, it was one of those things that was like, uh, you start playing with it. You're like, this is really cool. But then after a while you realize that this isn't cool anymore. Like this is just normal life. And so, yeah, it doesn't help me, man.

Can you just get to the point? Don't give me more homework to do. Like, that's really what the meat proxy problem is, is that I've given you more work to do when I want you to do the work, man. So if what I just talked about meat proxy, that's bad AI, good AI keeps that old shape, but it just gets there faster. Here's what happens.

So we have the same pile of ideas. You talk it through with your intelligence and artificial intelligence and boom, let's talk, let's figure things out, blah, blah, blah. You throw things out, you throw them out just like you did in the first one. Like, I don't like this idea. This is a good idea.

And it slowly sharpens and it gets smaller and smaller and smaller until what happens. Now it's one sharp note, just like the first one. And now you can send it to them. This is what happened in Slack with you is I had so much, I could talk about this Riverside situation for the whole podcast and bore everybody that's not listening. Do you know what I mean?

And so the thing is, is that that's the problem is that right now we all need to be using AI to shrink things, not expand things. That's good AI uses to shrink it. So when I said to it, Hey, you and I have been talking for an hour, Claude, about this problem with Riverside. I want Colleen to understand it in 20 seconds, give her a message that she can just briefly skim, get the gist of it and move on with her life. Instead of, I need to tell you everything that happened.

No, dude, it's inversing things. You need to make things sharper, not make things bigger. And that's where it's the Albert Einstein quote, everything should be made as simple as possible, but no simpler. This is really about empathy. You think about the other person who's receiving it.

It's not you, the one who sends it expands bad AI expands after you send good AI compresses before the send, you know, that you spend time whittling it down. Now, this is the five things you need to know instead of the 50,000 that I could tell you. And cause it's true, you could. And then the last one is don't use AI for everything. Just write your mom's birthday card yourself.

9:06Colleen

Um, a couple of

9:17Colleen

And matters because it is really important. You know, it is impersonal to let AI send something on your behalf. And so as long as the topic is impersonal, that's totally appropriate. And so like this exact communication string, I wasn't bothered at all. Uh, I was actually, I was impressed with how succinct it was and how much it sounded like you.

Cause I saw that it was sent by Claude. It notified me that it, that it was, I also had the little bit of context of like, as I was leaving yesterday, I was like, Hey, I don't know if you saw, but I'm looking, looking for this. And you're like, yeah, it's been a whole thing. Uh, and so I knew it had been a whole thing and getting that very brief message with a very clear answer was appreciated by me because I appreciated that you'd done the work. Um, but with that, like if you send, I swear to God, if you send me a happy birthday message from Claude, I'm going to punch you in the face.

You

10:12Scott

Just allowed me to automate that. Thank you. Yeah.

10:14Colleen

Now it's a guarantee that that's how every birthday for the rest of my life is. Yeah, no problem. But like, for real, like, I think that those sorts of things, like they matter, you know, is like the people in the same way we've talked about earlier in this conversation, that conversation is our connection to other people. Let's use some of that freed up space to be really intentional with the places where personal interaction does matter. I think just to get back to the main point of what you're talking about with meat proxy versus did you call it, you called the second thing, Silicon proxy.

I am

10:48Scott

Switching it up because I think that it's actually exact opposite of the meat proxy. Meaning like what you're doing is I'm using AI, which is Silicon. I'm using AI to communicate to you a very lengthy thing. Now I'm condensing it into something else. It's exactly what I did with you.

I went and I talked with AI and I was like, no, not this, this, this, this. It, he and I were like, whatever, we were chatting back and forth, not he, but it was just like, we're talking back of, okay, cool. And then at the very end, I was like, tell Colleen, I feel like I was using it as a proxy to communicate something to you in the same way that the other person does as well. There's still a proxy, but it's just the idea that what are you using? Like I am, I'm being a meat proxy that could, to go to another meat, but I'm using the Silicon to meet, to meet I'm, I'm shifting it.

I'm using the Silicon to communicate to you what needs to be done, but in a much smaller way than I've already done all the hard work. If

12:45Colleen

About doing the hard work. Truthfully is I think, I think it's natural for humans when we are met with something as powerful as AI that is very, very good at organizing ideas. I think a lot of people are tempted to just let their brains shut off in that moment and not stay aware and stay curious and continue to like question what they're reading. After you read a lot of AI written copy over time, it can be really tempting, especially if you don't fully understand what you're reading. It can start to be like, oh, I guess this is right.

I mean, I find this in myself sometimes, too. It's like you've got to stop and say, do I actually understand what I'm reading? Do I agree with it? How do I know it's true? Like, this is one of the most important skills for anybody who's using AI regularly to develop, I think, is that critical thinking lens, that active awareness when they're working with AI.

It's actually one of the reasons why I find the concept of an AI employee a really helpful one, because if I'm in front of a person, I have a different level of awareness and presence and mindfulness and curiosity than I do when I'm like scrolling social media. I think that that's something that we need to invite ourselves to do so that our goal is to communicate really well on the other side and find the signal through the noise and then pass it along. It's

14:06Scott

Good. Yeah, I love that. Um, I'm going to, yeah, just take the analogy, just maybe, uh, maybe in a different direction, but it's a little bit like when you walk down a dark alley and somebody is coming down, what do you think in your head? You're like, I am heightened. I am aware on edge. I am ready to go.

It's because you're in a situation that presents a potential opportunity for some scary things to happen and your body knows that and it reacts with adrenaline, with heightened senses, like that kind of stuff. Uh, and then there's other times where you might be with a loved one and you're just watching a show or something and you're just like chilling. There's no sense of like danger. And you're like, cool. And your body reacts in that way too.

I'm just saying a lot of people, when, when it comes to AI, their body naturally almost feels like, well, this thing will do the thing for me. So I don't have to do the thing and I don't have to think it'll just do the thing. It could not be further from the truth. As a matter of fact, like this is actually the biggest problem with AI right now is that people assume that it's omniscient, you know, that it knows everything. So therefore it will know the right answer, what, what we're really coming up with, there's a difference between knowledge and wisdom that sure, it may have a ton of knowledge.

It's the smartest thing on the planet knows everything, but if you ask it for some wisdom, it's almost always going to fail you. It will give you, here's some ways to think about it, but it's not going to be able to give you wisdom. It's just, it's not there. Like it's not sentient. It's just knows a ton of things and they can guess stuff.

15:46Colleen

Yeah. Yeah. I mean, I think an easy way to even break that down in our own minds is to think about like, it's trained on all of the information in the world, but most of that information conflicts, like, like there's a whole lot of disagreement among humans about every single topic. And so if all of that information is inside this robot, how does it choose what to say to you when you have a question? Well, it's like, well, what do they want to hear? Like what, what would be based on the context I have, what's the response that's most likely to give them the answer they're looking for. That's what, how search engines work too.

And we just have more information fueling these chat bots, these LLMs. So we have to, yeah, we have to stay present to it. We have to stay sharp with it, but, but there's things you can do to anchor it. Um, like depending on what you're working on, one of the things like, okay, you're troubleshooting this, this piece of software that we're using yesterday. Uh, I would bet, and if you didn't do this, you should, I typically, one of the first things I prompt, uh, when I'm doing that is go read the live documentation of this software so that you're up to speed on exactly how it works.

Because I don't want you making up answers for how it's going to work. And you, if you've ever tried to use AI to like troubleshoot something, uh, you've probably found yourself down a path and be like, wait a minute, that button doesn't even exist. Yeah,

17:05Scott

All the time. Yeah. Step one's open the app. Step two is to go here. And I'm like, step two isn't there. I'm there and it's not there.

So what do I do? Yeah. And it's like, oh, my bad. Yeah. So yeah, 100%, you always have to say, let's read documentation, get up to speed. One

17:21Colleen

Other thing though, I think is really important about this meat, meat proxy versus Silicon proxy idea is that I think it's important for our brains to do that synthesis work. Like for us to understand something, we actually do need to look at all of the options, understand them, like the information, and then make decisions about what we're keeping and what we're throwing away. That helps wire our brain. It helps improve us. It helps us to understand something.

And there's this perceived expertise that happens when you're using AI, because you have more information accessible to you than you've ever been before. And you're able to create things you've never been able to create before. That it can give us a false perception that we're an expert in a topic that we absolutely are not an expert in. A week

18:23Scott

Just talking as a team and they brought this up. And this is something that is happening to us quite a bit. It's there is a term for when you have so much content that's coming at you from AI that you have to sift through to apply your taste and your decision making, your approvals, denials, that kind of stuff. It's just constant work being thrown at you all the time. There is a term for that, and it's clatterboarding.

Oh, my gosh, this is so funny. I don't know why that strikes such a funny chord with me. But waterboarding, Claude. Oh, my gosh, dude. It

19:01Colleen

Just feels so deeply true, especially when you've been working this way for a long time, because you feel like the one thing you're most tired of at the end of the day is reading. Just feel like I'm reading so many words every day. It was actually really interesting. A few weeks ago, I went on vacation and I was like, I'm going to read a novel. I don't read a lot of like fiction.

I mostly read nonfiction. And I blasted through that thing in like two days. And that's never been me as a reader. I've always been like a slowly chip away at a book, but my brain is so strong in reading right now that it was just like, yeah, easy, easy. Good

19:35Scott

Point. I was I was reflecting on this earlier that it's very bizarre, like just growing up. All right. So me growing up, we weren't like a reading household. We didn't have books littered anywhere, you know, like everywhere. When my wife and I got married, her name's Ann. Ann was what she she'd always come to me and be like, hey, we're buying children's books, like picture books.

And we're like a year married. We didn't have kids till we were four years into marriage. And so I'm thinking the whole time, why are we acquiring books? Because it wasn't part of my culture growing up. And so now I see my kids are just they read all the time.

They're incredible readers. And I was reflecting on this, that in the world, in the age of AI, being a reader is an incredible skill. Because of this onslaught of content that is coming at you consistently all the time that you have to sift through it. So it's like you need to have the basic tool of taking in information. Oh, you know what it's like?

Do you want to have a landline or do you want to have like fiber optic cable coming in? Because it's it's data downloading into your brain. You upload something onto Google Drive to store it there. When you do that, how long does it take you? Depends on your upload speed.

That's reading. Boy,

20:54Colleen

That's really that's a great analogy for it, because then if you're like not a strong reader or if that's not a strong input for you, then you've got to use your AI to adapt to the way it's communicating with you. Like I've been playing around with just truly just out of my own fatigue for like, oh, I'm reading so much that I get tired of all of the reading. I've been playing a little bit with like, hey, communicate this with me differently. So like I'll have it spin up an HTML and a diagram and visualize something when I feel like my brain is getting tired and that will help me like sink my teeth into it. And then it makes it easier to contextualize the like paragraphs I need to read.

There's this programming Language called Mermaid, and I didn't know it. I don't know it. It's it's just something I discovered, especially when I'm working on processes. Like, OK, here's a lot of the clients I'm working with right now and a lot of our team were just rebuilding processes.

And sometimes this like we intuitively understand them. So like I was talking to somebody on our team yesterday and she has been working on the same process for one of our clients since she took it over three years ago. And there's a spreadsheet that predates that that it's been running off of. And it's all been fine. It works, but it's lots of hours per week.

And she doesn't want to do that work anymore. Like it's not it's not life giving work. It's it's just work that has to get done to keep everything organized. So we talked through the process and I was like, just show me how you do it. Just like walk me through your day as if I was taking it over, train me in how to do it.

And we recorded that whole thing. And then we took that to Claude and we said, hey, can you just visualize this using Mermaid? And it drew this incredibly detailed and complex and accurate diagram of every system where information lives, how it flows, where the checkpoints are, where things circle back and all that. We're like, great, you got it right. Oh, you missed this one thing.

So she was able to just quickly say, oh, you missed this right here of like when this revision happens, it needs to get reassigned in our project management tool. And because she could see it, she was able to then approve it and like, great, now let me just go automate that and rebuild the whole thing as an automated database where AI is running the show. But like visual can be really helpful and like overcoming that that like reading thing. I now

23:14Scott

Manage it instead of create it. And so now the thing that is creating it isn't a human. That's not scary. We have a lot of things happen by robots, you know, like or machines and like those kind of things. So it's not it's just the newest tool.

And so it's now doing a thing. And when it returns the results, now you're just the manager of it. You're the approver. You're saying like, this is taste. It's always taste.

Humans have taste. Robots don't. They just do things. And any taste they have is the vanilla of the world. It's the amalgamation of humankind saying most people like this kind of thing, but you aren't that person.

And so it needs to reflect you.