The Exchanges

Every argument clarity score on this site is built from rows on this page. Each question and answer was assessed with names hidden, the host's own answers included, on four things from 1 to 5: directness (does it answer the question asked), coherence (do the ideas follow), precision (concrete details and clear references), compression (says a lot per word). The weighted mix (30/30/25/15) is the exchange score. A person's published score averages their exchange scores on raw tape only, at least 8 of them, shrunk toward the cohort mean. Full method →

Jackson Greathouse Fall no published score: only 6 usable exchanges on raw tape, and a fair score needs 8+ · coarse estimate ≈4.0/5 from 6 raw tape exchanges record → ← everyone

Every exchange below was scored with names hidden, four dimensions each from 1 to 5. An exchange's score is 0.30·directness + 0.30·coherence + 0.25·precision + 0.15·compression. The published score averages the raw tape exchange scores and shrinks small samples toward the cohort mean, so five great answers can't beat twenty good ones. Produced feed rows count only toward coarse estimates, never toward a full score.

clear all ✕
6exchanges match
6on raw tape
0redirected or not addressed
Answered raw tape D 5 · C 5 · P 5 · Cm 4 4.85

Q What, what do you wish, you know, GPT-IV had, but doesn't have? Like what would, what would, what would take it, you know, hustle GPT to the next level?

A Uh, the ability to take action on a human input, on a human user's behalf. That's to say, if I want it to, right now, for here's an example, it can, uh, code a pretty good landing page for something, or it can write a really, you know, solid API integration to What I used it recently to write a Spotify API integration to a Webflow website I was working on. Boom, boom, boom. It was done in 10 minutes. Amazing. Way better than like if I was just Google's like searching Stack Overflow or something like that. But what it can do is actually go log into my Webflow account, put that code in, run it, debug it, publish it live. Like, you know, like it can, it can give you the bones, but you've still got to go take that human action. I think that's the thing that I keep kind of running into. As soon as it can like, You know, if I say, you know, find me the, the, uh, plan, plan me a trip to Aruba next month. Um, I want, you know, the, the best airline deal and, uh, you know, cheapest, cheapest airfare, uh, with the shortest travel time, but I want to stay in the nicest hotel possible. You kind of like plan it out and then it can actually go and take that action on your behalf and say, okay, I found you this, I booked it. I have your card on file. I, I bought this, this, and the other thing. And I, and you have an open table reservation at this Cool restaurant for Tuesday night, like when …

AI assessment note: “the ability to take action on a human input, on a human user's behalf.”

Answered raw tape D 5 · C 5 · P 5 · Cm 4 4.85

Q Hustle GT, and I loved how you talked about your, like, as a branding designer, it's, it's taking everything in me not to tell it this is a bad idea, but here we are, and it produced this, like, Kind of like stock looking image. Yeah. Um, and it was like good enough to get going. Right. But like this was, you would never launch an MVP with that. Correct?

A No. And the other thing is, I think when I use, ah, Dali or Midjourney to generate ideas for actual projects that I'm, that I'm running for clients or personal projects or whatever, I think I take a different, more kind of targeted approach to the prompt engineering that goes on there as well. In this case for the hustle GPT, uh, green gadget guru website, I asked it to generate a prompt and put that in verbatim. And I just picked the first one that it made. So I, I tried to have as little, uh, human intervention as possible. I said, you make the prompt, you, uh, you know, are in control of saying what it should be. And it's generally things like, you know, A, a gear for gadgets and like a leaf for eco-friendliness and like green and blue and white colors. And so it's like, okay, you kind of have like a decent idea. The way it was phrased though, and this is funny, was totally not the way you would normally write at a Dolly prompt. It was verbose and it was long and I didn't want to correct it because Dolly is an open AI product that I feel like the GPT language models should understand how to talk to Dolly. And I guess that's just not here yet. Um, which, fine, it, it still generated something. Uh, it knew it wanted a logo, and it understood everything, but it was just too verbose, and it was a little too, like, you know, unnecessarily descriptive to get a concise, uh, concept…

AI assessment note: “No. And the other thing is, I think when I use, ah, Dali”

Answered raw tape D 5 · C 5 · P 4 · Cm 4 4.60

Q And, um, what do you think, like, what was about this hustle GPT concept that really struck a chord with people?

A Yeah, I think it's the idea that, you know, so on one hand, there's been this kind of sci-fi fear narrative of like, oh, the robots are going to take our jobs. Like the AI is going to, is going to take all our jobs. And I think, you know, showing an alternative narrative to that, which is like, why don't we work together? Like we can, we can augment our own, um, productivity and our creativity, uh, with this technology at our, you know, at our side. Um, I took it to a little bit of an extreme, I think for the sake of this experiment, which is saying like, you know, I've got this whole spiel now. It's like, oh, like, let me see what my boss says. Like, you know, when originally we thought, oh, these AIs are going to be great virtual assistants. And now it's like, oh, what if we are the assistant to it? That's the flip.

AI assessment note: “showing an alternative narrative to that, which is like, why don't we work together?”

Answered raw tape D 5 · C 4 · P 4 · Cm 4 4.30

Q Yeah. I think the other, the other cool part about this whole thing is the interactivity of it all. Right. The fact that people could, you know, and in your initial tweet, you actually had the words, I think, follow along. And right. Did you have those words?

A I did. And Greg, I think that's a huge reason that like, that was not, I wasn't, I wasn't planning this. This was not like an engineered, like, oh, if I say this this way, then it's going to go super viral and all that. It was like, I just had to follow along to my friends that I, that I have on it, that I had on Twitter. And, uh, what I think happened was people saw it. They saw the thread developing in real time. They saw the words follow along, not in like a pushy, like salesy way, but they're like, all right, I'm literally going to Follow along. That's why I've, I've never seen, not to like toot my own horn too much, but like, I've never seen this kind of follower conversion off of a single tweet thread. I mean, it's nuts, right?

AI assessment note: “I did. And Greg, I think that's a huge reason”

Answered raw tape D 4 · C 3 · P 3 · Cm 3 3.30

Q What are the, what are the jobs that have the highest number of repetitive tasks?

A It's well, I think the hardest part about this that, okay. Okay. They're, they're entry level jobs, regardless of category or industry that are given usually to younger people who are just entering the workforce. They're the internships. They're the paralegal for an example. Isn't it interesting? That those same young people who are just entering the workforce are also the ones who are going to be the best versed and the most native to these AI tools. Isn't it almost a little bit ironic that those people who would be up for those jobs are the ones who are going to be able to manipulate the AI models and prompt engineer the best on a large scale. Of course, there are exceptions on both sides. Um, I think the people who are really scared about AI taking their jobs are like, Older people who are just unfamiliar with what, what it means or what it is. And I think that's okay. I, I don't, I don't know though. I don't, I, none of us could, could you have predicted this even a year ago? I don't think I could have. And I stayed pretty like up to date on this stuff.

AI assessment note: “They're, they're entry level jobs... internships. They're the paralegal for an example.”

Answered raw tape D 3 · C 3 · P 4 · Cm 3 3.25

Q Yeah. Amazing guy. But I think like, so Actual takeaway for people. What do you think that is? Like, what is it about your background and your story that prepped you for this moment?

A Look, I, I'm, I'm just gonna say it now, because like, I'm here, like, this is happening. Um, you know, I've always known that this is kind of my, my, I've always known that this is kind of my intended, uh, I don't want to say final form, because that implies like, you know, some, some, some completion of something. That's not at all what I mean to say. When I had a conversation with my Incredible friend, Garrett Scott, Pipe Dream Labs, shout out. We sat down for a cocktail in Oklahoma City about a year, year ago, year and a half ago. Um, and I asked him, you know, what is the one thing that you think that I'm failing to do that I could be doing better? And because I trust him a lot as a person and he, without missing a beat, said making video content, growing an audience online, sharing your story with people. You're meant to do this. This is something that you Haven't given a true shot at, uh, for whatever reason, and you owe it to yourself and everyone else to do that. And that really sat with me a lot. And, you know, we can make excuses all day for why not to do something, you know, don't have enough time. Quality's not there. Not good enough. Embarrassed. People are going to judge me. People are going to, I remember when I was 12 years old, starting that first video podcast when I was, you know, Talking to Gary Vaynerchuk after he dropped his first ever book in 2009, I mea…

AI assessment note: “I remember when I was 12 years old, starting that first video podcast”

page 1
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 300 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.