Tag Archive for: AI

Listen to the full episode right here, or wherever you stream podcasts.

Find all episodes at HelloMergeTag.com or wherever you stream podcasts.

Links

Josh’s LinkedIn

Josh Reports

GOP Texas Primary Tracking

Battery Energy Storage (BESS)

Josh’s free AI tools

AEO Checker

State Leg Zips

AI Topic Explorer

 

Big thanks to our sponsors:

Civic Shout. Learn more about the great work they’re doing to help Democratic campaigns and progressive nonprofits build their lists ethically at civicshout.com/partners.

 

Episode Transcript

This transcript was automatically generated and has not been copy-edited. While it can serve as a helpful guide to our conversation, it may contain errors or omissions. For the most accurate representation, we recommend listening to (or watching) the full episode.

Josh Klemons: Okay. Josh Canter is vice president of digital advocacy at Maguire Woods Consulting where he uses digital tools to
Joshua Canter: yeah.
Josh Klemons: influence legislation and shape public policy. His work spans multinational uh multi-million dollar ad campaigns, interactive platforms, organic content, and full-scale omni channel strategies. all focused on driving real world impact. His work’s been featured in outlets like the New York Times and Wired, and he’s worked on campaigns and advocacy efforts all around the world, hence the multinational instead of multi-million. Uh, from what I can see, he’s also one hell of a vibe coder and experimentter. I caught a LinkedIn post from Josh in my feed where he was talking about using AI to completely rethink how we analyze data from the meta ad library and I of course had to reach out.

Josh Klemons: So Josh, thank you so much for coming on the pod. Um, let’s just start like high level.
Joshua Canter: Thank
Josh Klemons: I am assuming anybody listening to this podcast is familiar with the meta ad library.
Joshua Canter: you.
Josh Klemons: If you’re not, welcome. Like I’m glad to have you here. Uh but why don’t you just very high level, let’s not get into the tool yet. Let’s just give very high level what the meta ad library is for anybody who like doesn’t know it
Joshua Canter: Cool.
Josh Klemons: yet.
Joshua Canter: Yeah. Thank you so much for having me and uh shout out to LinkedIn for algorithm pairing us together.
Josh Klemons: I know LinkedIn knows my audience.
Joshua Canter: That’s exactly what I want to Yeah,
Josh Klemons: I know. It’s
Joshua Canter: that’s exactly what I’m going for.
Josh Klemons: awesome.
Joshua Canter: So, uh, yeah, definitely. So, the meta ads library has been around since 2016, I think,
Josh Klemons: I think 18.
Joshua Canter: is when they started and they basically 18,
Josh Klemons: Right.

Josh Klemons: Yeah.
Joshua Canter: right? I think the drama in 16 started in 18,
Josh Klemons: Right. Caused it.
Joshua Canter: right? It’s snowball effect.
Josh Klemons: Right.
Joshua Canter: So, um, basically they wanted to release a tool and I should note that subsequent, you know, the other ad networks have since released similar tools. They’re just not as breath. Google, Snapchat, etc. But they released a tool where after you verify that you’re American citizen running ads in an American, you know, in in a state or in the country and you are uh verified to run these political ads, all of those ads that are deemed political, social or issue are then stored in an ads library for up to seven years um with information on impression, delivery, um spend and then the actual ad content. A lot of this stuff is ranges. So, I’ll say you spent, you know, a thousand to $5,000. Um, but it’s became really useful for seeing opposition, seeing how people are talking about issues, seeing what kind of ads people are running.

Joshua Canter: Um, a lot of stuff has sprung up from the ads library. There’s a great weekly uh Substack, I believe it is, that talks about kind of different campaigns and spending levels. There’s a leaderboard, I’ll call it, where people are spending, you know, who’s spending the most per week. that Facebook updates and then they released an API that is intended for research purposes as kind of a goodwill gesture to um have you know universities research into
Josh Klemons: I’m
Joshua Canter: how digital ads around political campaigns issues social issues are kind of utilized. I think that’s like the most high level
Josh Klemons: very curious about your ability to get permission because like a lot of the idea behind like the API was
Joshua Canter: overview.
Josh Klemons: to slow down the breaking news like they didn’t want people to actually understand what’s happening so they gave it to researchers knowing it would take
Joshua Canter: Yeah.
Josh Klemons: years before it came out but AI sort of changed the impetus.
Joshua Canter: Yeah,
Josh Klemons: So let’s not I I I want to come back to that.

Joshua Canter: totally.
Josh Klemons: So let’s start with what you built is pretty cool. Like I was digging through some of the reports you put through. So essentially what I understand is you you give you connected claude to the
Joshua Canter: Yeah.
Josh Klemons: AP to the ads library through their through the meta API and then had it burst like spit out reports so that we could actually under like it’s very easy to eyeball the ads library.
Joshua Canter: Yeah,
Josh Klemons: It’s very hard to like really make analysis without spending like 10 hours in there.
Joshua Canter: totally.
Josh Klemons: So why don’t you just walk us through like that’s what the ad library is. What is your tool actually doing on top of it?
Joshua Canter: Totally. So I mean like what I’ve been doing is exactly that for the past when ever since 2018 like re Google or searching within the ads library for advertisers or keyword base taking notes what I see compiling having a calculator that’s what my clients have been asking for right and it’s exactly what you said it’s been really hard and so I kind of took a first step of like I’m a big like you said throw everything in claw like

Josh Klemons: Sure.
Joshua Canter: how can it like take whatever you know nonproprietary I should say but you know whatever in cloud and like it analyze it. You know what? It’s been a godsend for polls and things like that. So, all of that in there. And so, at first I was like, okay, let me like dump my research into Claude and have him help organize it and structure it for whenever I present it to the client. And then I was like, there’s an API here. Like, I wonder if I can like just connect the two. And honestly, what happens every single time I try and build something new on cloud is I just go to cloud and I say like, hey, there’s this API. Can you do this? And every time they’re like, “Yeah.” And they’re like, “It’s going to take three to four days.” And I was like, “No, you’re going to do it.” And he’s like, “Okay, 30 minutes.” I’m like, “Great. This is great.” So, so yeah.

Joshua Canter: So, so start that was the next progression was making the connection. The connection was a little difficulty like you said because the research and Facebook’s APIs like I don’t really know what’s happening over there, but the API interface looked like Facebook from 2011.
Josh Klemons: Oh, interesting.
Joshua Canter: Like it was like the buttons were the old style. It was very odd. But Claude walked me through it all. So I connected it all and then I started just doing queries just saying like what about this?
Josh Klemons: Weird.
Joshua Canter: What about this? Um can you pull these numbers? Can you pull the or these ads or these pages? And then the next step that I progressed to is actually building a skill where I had kind of a repeatable task that I basically told Claude how to become a researcher um and then build reports from it. And I can either have like a published HTML page report on my uh repo on GitHub or I can have like him just return results to me and then I just copy that into an email or just use that for my own

Josh Klemons: First of all, I have to say um many people have talked about AI on this uh podcast.
Joshua Canter: knowledge.
Josh Klemons: You’re the first person to gender AI, so that’s very interesting to me.
Joshua Canter: I just I thought of that like mid-sentence.
Josh Klemons: That’s okay.
Joshua Canter: I was like I called him a he.
Josh Klemons: He they back to he.
Joshua Canter: Well, so me and my friends like clo Yeah.
Josh Klemons: Yeah.
Joshua Canter: Me and my friends like colloquially referred to him as a he.
Josh Klemons: Yeah. anthropic wants us to think of it as a real person.
Joshua Canter: So that’s where it like gets a lot like it,
Josh Klemons: Like they seem to think that it’s a living breathing thing.
Joshua Canter: right?
Josh Klemons: So I’m sure they would be thrilled to know. But yeah,
Joshua Canter: Yeah.
Josh Klemons: I just I had to flag that.
Joshua Canter: Yeah. Yeah.
Josh Klemons: That’s like a fascinating insight into like how you’re thinking about it, which is cool.
Joshua Canter: Totally.
Josh Klemons: Um so the So I mentioned this earlier and I might be wrong.
Joshua Canter: Yeah.
Josh Klemons: I’m not an expert on this, but I have read this. So like essentially Facebook used to have like what was that tool that like essentially re reporters had access to?
Joshua Canter: Yeah.
Josh Klemons: I’m drawing a blank on it, but it was very easy to essentially see what people were doing and then they they killed it because it was bad for them.
Joshua Canter: Yes. Yes. Yep.
Josh Klemons: like they were like there was Twitter I think Kevin Roose set up like a Twitter account that was like sharing the top posts on Facebook and it was
Joshua Canter: Totally.
Josh Klemons: humiliating. So they removed a lot of transparency and replaced it with the API specifically for researchers.
Joshua Canter: Totally.
Josh Klemons: And my understanding this I’m sure I’m just regurgitating Casey Newton here but like essentially by giving it to researchers, academic researchers, they guaranteed that any controversy was three or four years down the road and no longer relevant. Are you worried that once they realize people are doing this kind of thing that they’ll shut it down?

Josh Klemons: Did you have any issues with Meta like allowing you to use the API or it’s just like an open box that anybody can plug into and like toy away
Joshua Canter: Yeah, it it’s definitely an open box.
Josh Klemons: with
Joshua Canter: Um I I would say I haven’t talked to anyone at Meta at it. Um I you know, am I worried about them shutting it down? I’m very much like let’s get in there and break s***
Josh Klemons: Sure.
Joshua Canter: and if they shut it down,
Josh Klemons: That’s like their whole mantra.
Joshua Canter: they shut it down. you know,
Josh Klemons: So
Joshua Canter: I’m right I’m like I’m no I’m no worse off if they shut it down because I just go back to the
Josh Klemons: Sure. Right.
Joshua Canter: old way, right? Um, but I could I I’ll tell you and I kind of hinted at this like just I don’t know a lot about API like cloud was doing the heavy lifting here obviously but like it seems like they are not upkeeping it like it is very you know it’s not like click here a lot of other APIs that I’ve connected to it’s like create a token or create a secret passcode and then boom that’s it like theirs is like what kind of permission what kind of this what kind of
Joshua Canter: it’s it it does it doesn’t seem very user friendly or like anyone has been in there you know engineer on metaite for a while. Um,
Josh Klemons: Like I was going to say,
Joshua Canter: so I think I kind of hope they’ve forgotten about it,
Josh Klemons: do you think they’re doing that intentionally or they’re just doing nothing and that’s the issue?
Joshua Canter: frankly. But I think I mean I think they’re going through their own,
Josh Klemons: Hard to say.
Joshua Canter: you know,
Josh Klemons: Yeah,
Joshua Canter: there’s they’ve pivoted so far to AI or the metaverse that just now shut down and then to AI,
Josh Klemons: I know.
Joshua Canter: you know, that it’s like I think they’re, you know, Yeah. And I think you said this too, like I don’t know if they’re as cognitive as the fact that AI can just I mean, sorry, let me clarify this. AI can just do things right and I think they know that and they’re focused on that for themselves and aren’t thinking about it as it relates to the a to the APIs like you said like at my you know any other company or any other thing nine months ago you would have to hire a developer plug into an API work out an interface right I use claude it’s a built-in interface right and all these kinds of things and now AI has just transformed and leveled that playing field incredibly that I can do it in a matter of few hours and then
Joshua Canter: refine it over a few weeks and here we are two months later I think from when I posted maybe not even maybe just a
Josh Klemons: Yeah.
Joshua Canter: month right and now we’re here you know it’s so far it’s so so much
Josh Klemons: So, yeah. So, so folks, first of all, I will link to at least one or I’ve got two of the reports pulled up and maybe there are more,
Joshua Canter: faster
Josh Klemons: but I saw the one on Bess and the one on the Texas gubinatorial uh Republican Senate primary. Let’s talk about the Senate primary one.
Joshua Canter: Yeah.
Josh Klemons: So, help people understand, and again, you can go in the show notes and find the links,
Joshua Canter: Yeah.
Josh Klemons: but help people understand what the report actually looks like, but also walk us through how much of that was your work versus theirs. Like you obviously did all the work to train Claude and all that,
Joshua Canter: Totally.
Josh Klemons: but did it spit out a final draft of that or it gave you a bunch of data that you then went and like cuz even the formatting like

Joshua Canter: Totally.
Josh Klemons: the UX is beautiful on it.
Joshua Canter: Totally.
Josh Klemons: Like is that you claude or a combination? Like walk us through what people can expect to see there and how did you make it look so digestible if
Joshua Canter: Yeah. Yeah. Yeah. So,
Josh Klemons: you
Joshua Canter: let me dive into the skill, too, because I think that would be really helpful because like I said, the first step I was just quering things with with Claude and and was just like spinning back results. The skill sets a defined amount of process that I could just say like, “Hey, run me a report using this skill.” So, the skill has a few steps to it. So, one is you obviously input something to Claude into the skill and then claude asked clarifying questions. What’s the scope? So, it regional based g uh uh time period, right? Are we going back 90 days? Are we going back a year? Is it just this singular Facebook page or is it just this singular issue or is it all things around it?
Joshua Canter: Clarifying that within questions and then cloud dispatches agents to basically ask a bunch of queries to the API related to that and gets it all back. So if you say like the um perfect example is I want to use the Texas Senate primary, right? And then it’s like is it just the candidates? What about outside groups, right? How far back do we go? Do we want to, you know, are we focused just in the last week or more? After he gets all that back, I actually formatted the skill to then do another pass and do a second research pool. So, it comes up with a bunch more different queries, sends it and gets it back. And then what I do is I have Claude.
Josh Klemons: No, no.
Joshua Canter: Yeah.
Josh Klemons: I just wanted to clarify.
Joshua Canter: You are you going to ask it?
Josh Klemons: You’re talking about skill.
Joshua Canter: No. Oh,
Josh Klemons: This is Claude’s skill, not your skill. Okay. I just want to make sure folks are following.

Joshua Canter: yeah. Correct. This is all claw.
Josh Klemons: Like, it’s not that you developed a skill.
Joshua Canter: This is all claw. No. No. No. Yeah.
Josh Klemons: It’s like you trained the AI and that’s I just want to make sure that I understand the language you’re using.
Joshua Canter: Yeah. Yeah. I think that’s like anthropic’s official language of what a skill is is like a repeatable task that that uh
Josh Klemons: Okay, cool. Okay, cool. Okay, continue though.
Joshua Canter: Claude can do.
Josh Klemons: That’s helpful.
Joshua Canter: So, but I’m not doing any of this. This is all this is all that I get.
Josh Klemons: Okay.
Joshua Canter: So, he does so that he does a second or I sorry, he did the first pass and then in between the first and the second pass, what Claw does is do web search research on whatever the topic is. So, it reads all the articles. It reads the websites of the the candidates, right? It it reads the historical facts about the election and it kind of creates like a context to all of the research that it’s pulling, does that second pass, pulls it all back, then it analyzes everything, and then it puts it into report.
Joshua Canter: Um, and so it’s actually hard coding that HTML report that you see.
Josh Klemons: Oh,
Joshua Canter: Um, and everything that you see like I’m not like here’s an icon like it goes and finds the icon and pulls it back. um and the analysis. And so,
Josh Klemons: Wild.
Joshua Canter: and I I kind of steered it in the direction of like be very visual. Try and use charts and graphs, right? If you’re going to talk about an ad, link to an ad, right? Um don’t make assumptions, right? Or if you do make assumptions, make it clear that you’re making an assumption with your reasoning behind it so someone can help see it. Um at the same I I’ll call out like a funny little thing. So, the Republican Senate primary in Texas, there’s uh a third party, Wesley or a third candidate, Wesley Hunt, right? It was um Paxton and Cornin and Hunt,
Josh Klemons: It was right.
Joshua Canter: right? Um I said pull everything related to the primary and somehow when it pulled back, it pulled that Wesley Hunt had spent like I think it was like $6 million on Facebook or something.

Josh Klemons: Wow.
Joshua Canter: Um and obviously dwarfing Cornin and Paxton and I was like that’s odd, you know? And so there’s like a gut check thing to it. So a lot of times before the reports are finalized or before I like even present it to clients, I’ll just gut check it and do some stuff. It’s not imperfect, right? it’s going to make mistakes, but it’s really helpful to have that like outside contextual window and into the actual data and kind of pair that together.
Josh Klemons: So to clarify, Hunt had not spent $6 million on ads.
Joshua Canter: No,
Josh Klemons: It just like hallucinated that.
Joshua Canter: no, no, no, no, no. I think it was like 300,000 maybe.
Josh Klemons: Okay. Yeah. I mean,
Joshua Canter: Yeah.
Josh Klemons: that makes more sense given the campaign,
Joshua Canter: I don’t I don’t know exactly. I think there was like a Yeah,
Josh Klemons: but
Joshua Canter: I think there was like a um it had it was he ran for primary like he’s a perennial candidate so I think it had somehow like totaled all of his primary spending rather than just
Josh Klemons: got it. Got it.
Joshua Canter: this racist primary. I don’t know. Yeah,
Josh Klemons: So,
Joshua Canter: it was very it was very funny though.
Josh Klemons: it’s good you caught that, but like what other kind like give me an example of something else that like surprised you reviewing the report and also I
Joshua Canter: Yeah.
Josh Klemons: really want you to talk more about this whole idea of like contextualization of like events in real life. Like it’s not just a flat like here’s what they spent, here’s what the ads look like. You’re actually connecting it back to what’s happening in the real world, right?
Joshua Canter: Yeah. And I I’ll start with the contextual. So the contextual I think works really well. It’s harder. So I’ve been toying around this idea of like claw as a campaign manager, right? like campaigns are run as like you know we all know the playbook or more or less the playbook or our playbook or whatever but then there’s real world s*** that happens you know and you have to like pivot on the fly and that’s what makes campaigns fun in my opinion is you never

Josh Klemons: There.
Joshua Canter: know you know what’s going to happen and so I you know with that and I do a lot of issue advocacy and in the legislative process there’s also real world s*** that impacts it right the politics of of of governing and so I found that a lot of contextual will kind of help paint that picture of it and can kind of assign it to real world things. So, I’ve seen a lot of stuff come out of like, oh, well, they’ve announced their plan, you know, their this candidate has announced their 10-point plan for when they get elected and there’s an increase in ad spend. And so, Claude can kind of make those connections where you would have to be a real big expert in the race to know that and involved in the race to know that. And so, that’s something that I think has been really helpful to understand. Um the analysis piece of it is I really tried to have it be like okay what do we know and what can we not assume. So like for example one that I think it did on the Democratic primary was it found that Crockett was running video ads Telerico was not or something similar to that.
Joshua Canter: I think Telerico was at the end of it but there’s like some gap to because all this is also plotted on a timeline. So there’s interesting kind of strategies that you can kind of take from the um from each of the candidates. I don’t know. It’s done some issue based stuff like, oh, here’s top issues, here’s other top issues. It’s pretty good on a high level, but when you get into like objective or strategy of the ad, it knows some it knows like lead generation versus not lead generation, but it I don’t know if it’s been able to figure out persuasion versus awareness or things that are more nuanced like
Josh Klemons: What about like you said lead versus not lead like what about like conversion ads like that are geared
Joshua Canter: that.
Josh Klemons: towards fundraising versus more like awareness and whatnot?
Joshua Canter: I think if there’s yeah if there’s a clear call to action like it definitely knows like it knows that either through the like type of ad that it can see so like a lead genen form instant uh instant form on Facebook or if there’s a clear call to action like

Josh Klemons: Interesting.
Joshua Canter: donation right it like clearly is like this is a donation ad um for that and it also helps I’ll
Josh Klemons: And
Joshua Canter: say too at scale like you know they’re running tons of ads and I can’t see all 5,000 ads or whatever so you know we could see at scale like are most of them donation are most of them lead genen right and that tells us a lot about the campaign
Josh Klemons: Um, what are you picturing this as a tool? Is this an upper research tool or is this for like campaigns to learn to get better? Like who do you think is the prime audience for a tool like this?
Joshua Canter: Totally.
Josh Klemons: I guess I’ll start there.
Joshua Canter: Yeah. So, I definitely see it as oppo research like 100%. Um,
Josh Klemons: Just like keep up with your c with what the other candidates doing kind of thing.
Joshua Canter: you know, like we often totally totally Yeah.
Josh Klemons: Okay.
Joshua Canter: And I’d love to expand it to the FCC reports for TV, right?

Joshua Canter: You know, like somehow get in there and and see that too. Um, as a digital first buyer, it’s very um, like the TV stuff is just a black box to me, right? I can kind of understand it, but as much as I need to, but there’s so much there, too. Um, and I’m sure there’s tools that exist for TV outside of this.
Josh Klemons: Yeah, Andy Bar has something,
Joshua Canter: But yeah.
Josh Klemons: but like it’s not pulling in the it’s not pulling in this kind of like from what I’ve seen and I saw a very early version of it,
Joshua Canter: Yeah.
Josh Klemons: it’s not pulling in this kind of data where it’s actually explaining what things look like. It’s much more based on like high level like spend.
Joshua Canter: Totally.
Josh Klemons: Um,
Joshua Canter: Yeah.
Josh Klemons: but like here’s what they’re spending on is like a whole different kind of story,
Joshua Canter: It’s probably FCC document based.
Josh Klemons: you
Joshua Canter: Yeah. Yeah.
Josh Klemons: know.
Joshua Canter: It’s probably it’s probably more like based on the actual reportings than it is like the content of the ad, you know?

Joshua Canter: Maybe it is. I mean,
Josh Klemons: Sure.
Joshua Canter: maybe you could connect it to like a monitoring and and build it that way. But yeah, but I really see it as like a tool for oppo research, a tool for just campaign strategy adjustment. Um, a tool for research. I mean, like I have no dog in the Texas Senate race, right? Um, and but it’s really cool to me to be like, hey, you know, like let me see what these campaigns are doing more than just me scrolling Twitter and seeing
Josh Klemons: Um, so you are obviously like a pretty technical person like you you know what you know the difference between different APIs
Joshua Canter: it.
Josh Klemons: like so I’m going to say that puts you at like a pretty high level for our industry which cool. Um, is this like is this the type of thing where do you think that like any digital director on a campaign should
Joshua Canter: Yeah.
Josh Klemons: be trying to build something like this using your tool? Like how do you see this being implemented like in real life?

Joshua Canter: Yeah. Yeah. So, I think any digital director anywhere should have a cloud subscription or a chat GBT
Josh Klemons: Sure.
Joshua Canter: or Gemini or something like and try and do more of the vibe coding or the coding basis because there’s so much out there to connect to. Um like the one that I worked on last night is the nation builder connection. So, I know a lot of campaigns are built on nations. I have a trade association built on nation builder. some of the searching. It’s funny to me because like searching and filtering, you know, keyword, boolean, whatever kind of format they come in to me, it’s it’s so much more native to do aic and just in a prompt conversation. And so building those connections I think too are like also something that can help speed up workflows to where instead of the digital director grabbing an intern and saying hey do all this research on this candidate on the Facebook
Josh Klemons: Sure.
Joshua Canter: ads just goes to chat GBT or something prompts it and then it pulls it all back like that’s where I mean if you want to get super AI economy shifting changing like interns are screwed maybe not interns because they’re you know it’s a learning opportunity but like entry level like entry level grunt work
Josh Klemons: We’re we’re breaking the pipeline, right?
Joshua Canter: like that’s change.
Josh Klemons: Like folks are not going to get hired for the lower level jobs so they can work their way to the higher level jobs,
Joshua Canter: Yeah, totally.
Josh Klemons: which God help us all when like,
Joshua Canter: Oh,
Josh Klemons: you know,
Joshua Canter: it’s Yeah.
Josh Klemons: like I don’t know. Maybe maybe everything changes and it turns out that the kids right out of college are are bosses in 10 years because they’re keeping
Joshua Canter: Yeah.
Josh Klemons: up in ways that we can’t.
Joshua Canter: Yeah. Right.
Josh Klemons: I don’t know what happens,
Joshua Canter: They like know.
Josh Klemons: but certainly the disruption is here obviously.
Joshua Canter: But I know. Yeah. Oh, I talk all the time of like you get a product back from AI,
Josh Klemons: So,
Joshua Canter: I know if the product’s good because I had to do the product a thousand times when I was first starting off, right? But if you’ve not done it a thousand times, you have no idea if it’s good.

Josh Klemons: right.
Joshua Canter: And so then it’s like this huge issue.
Josh Klemons: Like I have this stupid Google uh speaker in my office that I literally used for nothing other than to turn on
Joshua Canter: But
Josh Klemons: and off my lights and setting that thing up was so complicated. Um and I have a friend who built he like wired his whole house and now Google has Gemini in it.
Joshua Canter: yeah,
Josh Klemons: I’m telling this as a context for like how to think about this stuff for like listeners. So he can talk to it in normal language and say if I get home after 10 p.m.
Joshua Canter: totally.
Josh Klemons: I want you to do these three things and it’ll just do it. I had to like wrestle between six different apps on my phone to connect one light bulb with another light bulb.
Joshua Canter: Yeah, that’s what I’m saying though.
Josh Klemons: But it’s like like it was so complicated just two or three years ago and now he’s just like,
Joshua Canter: Yeah,
Josh Klemons: “Hey,
Joshua Canter: totally.
Josh Klemons: anytime I get home after 10, turn the heat on.” And it’s like,

Josh Klemons: “Holy crap,” you know?
Joshua Canter: Totally.
Josh Klemons: So like if that’s what you could do with your smartphone, like again,
Joshua Canter: Yeah.
Josh Klemons: like imagine what we’re going to be able to do with
Joshua Canter: Well,
Josh Klemons: like
Joshua Canter: that’s the thing is and it’s like native conversation like I think everyone prefer maybe not everyone but a lot of people prefer that
Josh Klemons: Sure.
Joshua Canter: native conversation aspect of it and you know that’s why it works so well with interns and with entry- level people because you can kind of tell them you can break it down explain it and then have them go and do the task and that’s to me where I’ve like started in my own workflow like what tools do I use every day is there an API let me ask Claude if he can connect to it and feed it in and that was literally the basis for this tool and for a lot of other ones that I’ve
Josh Klemons: Yeah.
Joshua Canter: worked
Josh Klemons: Yeah. The ability to go into like an AI tool to do a general query where you don’t have to like hit the exact right search terms, you could just use natural language.

Josh Klemons: I mean that that to me was like the first massive game changer of this. It’s like I have a very specific question.
Joshua Canter: Yeah.
Josh Klemons: I know I am trying to connect these two things or figure this out and instead of having to like find the right article like it’s talking to me. Obviously there’s great risk for what the future of the internet looks like when nobody’s bothering to create content but like two different you know.
Joshua Canter: Yeah. Yeah.
Josh Klemons: So on that front,
Joshua Canter: Yeah.
Josh Klemons: like you mentioned, you didn’t you talked a little bit about technical issues, but you did talk about like it saying Wesley Hunts had a $6 million budget. Like a lesser strategist might not have pinged on that and might have gotten really bad data for not like questioning
Joshua Canter: Yeah,
Josh Klemons: that assumption and like all of a sudden you put out this massive report and it’s like completely not only
Joshua Canter: totally.
Josh Klemons: unhelpful but like the opposite of helpful,
Joshua Canter: Or you change your strategy.

Josh Klemons: right? Like
Joshua Canter: Imagine if like someone on the corning campaign like saw that like you know or thought it was real like you know
Josh Klemons: sure. So,
Joshua Canter: like totally. Yeah.
Josh Klemons: how do you how do you how much can you trust the data without having somebody like you who’s smart enough to like reveal? Like you obviously you do this for a living,
Joshua Canter: Yeah.
Josh Klemons: right? Like you’re essentially having AI tool do something you are already doing. So when you see a giant red flag on there,
Joshua Canter: Yeah.
Josh Klemons: the first thing you do is go question it. The average person isn’t going to do that. They’re going to be like, “Oh crap, look at Hunt. He’s crushing it here.” So, what’s your advice, I guess, to folks like whether they’re building it,
Joshua Canter: Yeah.
Josh Klemons: using a tool, um how how suspicious do you have to be of the findings or is it only the like giant red flags that jump out that you like worry

Joshua Canter: Yeah. I mean, so I think the there’s a lot of levels to this,
Josh Klemons: about?
Joshua Canter: right? I think the first and foremost thing is like it’s the perennial like it used to be, you know, unprecedented everyone was saying. Now it’s like you got to gut check AI, right? Like that’s just a given, right? You have to gut check it when you do it. And so I think that’s but there’s a lot of levels to it. So there’s some reports that I’ll do. Like I said, I’m just interested in it. So,
Josh Klemons: Sure.
Joshua Canter: I maybe gut-ecking it as much as if a client was paying me for it, right? And then I have to produce it or if it was changing strategy, right? If it was like adapting and changing strategy. I also think there’s just different like I think of it a lot of like um walled gardens and like gates like it it with particularly like that’s what I was trying to do with the skill is like tell me when you assume something so that way I know you’re assuming it and I can kind of check your reasoning for it.

Joshua Canter: With that stuff, it’s very like okay, when it’s dealing with data, great. like that’s a big thing that I know, you know, stats and numbers and things, it can probably get pretty quick, right? Um, but if it’s dealing with writing, then I need to like actually fundamentally check that a little bit more. Or a big one that I always check more is if it’s dealing with either um like I do a lot of like Excel document manipulation through Claude or through AI platforms. Um, and I’m like, copy this, you know, over to this, tell me how many people have this and all these stuff. Again, some of it is just I appreciate prompting more than I do like finding the button within Excel, but some of it is like I got to figure out all these things together. Um, and I don’t want to Google how to mail merge anymore,
Josh Klemons: Right.
Joshua Canter: right? Or like how to split comms.
Josh Klemons: Right.
Joshua Canter: Like I always forget where that button is.
Josh Klemons: I have used AI to split columns.

Joshua Canter: So that’s where I’m like, okay,
Josh Klemons: That’s about the extent of my Excel work.
Joshua Canter: what
Josh Klemons: But like I’m like, no, I use numbers and it’s a disaster to split first and last name.
Joshua Canter: right?
Josh Klemons: So, anytime I have one,
Joshua Canter: Oh gosh.
Josh Klemons: like I just throw it in.
Joshua Canter: Yeah.
Josh Klemons: And it’s amazing at that. Like,
Joshua Canter: Yeah.
Josh Klemons: for sure amazing.
Joshua Canter: Yeah. Yeah. Yeah. But you kind of have to check it too because like I’ve had some like what did I do one the other day? Oh, I was it had a list of company names and it was like split. I asked it to split the columns and I guess it did like comma separated split or something. So it was like company name next column inc or like so you know what I mean?
Josh Klemons: Right. Yeah. Yeah.
Joshua Canter: It’s like so you’re like technically that’s true but like also that doesn’t work. So that’s where I think there’s like different levels to that.

Josh Klemons: Right.
Joshua Canter: The thing I’m building right now that I kind of hinted at was like the nation builder where I’m connecting it into the database to like do fundamental searches differently. Um, and it’s a huge one where I’m like, we have to build in the safety guardrails and the precautions because if you report the wrong people or the wrong numbers, like then we’re sending emails to the wrong people or we’re doing the wrong we’re giving the wrong people permission for the wrong part of the website or or something like that where I’m like, it needs to have the guards where if it can’t do a task,
Josh Klemons: Yeah.
Joshua Canter: it says it can’t or it says like I can only take it to 90% or whatever the case may be.
Josh Klemons: Interesting. Because yeah, they’re notorious for saying yes even if they should be saying no.
Joshua Canter: That’s where totally
Josh Klemons: And you’re So you’re you’re finding that you actually can train it to say
Joshua Canter: totally so I haven’t done an explicit prompt like say no if you can’t

Josh Klemons: no.
Joshua Canter: do the task but I’ve done a lot of like hey if you are making an assumption tell me the reasoning or if you or
Josh Klemons: Sure.
Joshua Canter: or do like baby steps towards a task a lot of times I feel better rather than like um
Josh Klemons: Here’s a final report instead of like here’s like 10 steps I took to get there,
Joshua Canter: you know like right full this this and this yeah yeah yeah or something like that um
Josh Klemons: right?
Joshua Canter: or even like downloading a file halfway through and like test or uh spot checking it or figuring stuff like that. I think again the big thing clause there’s never going to be perfect and for these research documents it’s really more about like how can I just get a bunch of the information together. A lot of times I’ll talk to Claude at the end of it and be like, “Hey, what about what why did you say this, you know, or why what about this assumption or like I don’t think you can make that and then adjust it and move forward to one time

Josh Klemons: Interesting.
Joshua Canter: I was talking with it yesterday and this isn’t in a report. I was just doing it within chat because it was a quick little pool. I found a organization. I had two different Facebook pages but we’re running ads.
Josh Klemons: Right
Joshua Canter: Like clearly someone had made one, ran some ads, they lost access, done another. We all know the story.
Josh Klemons: there.
Joshua Canter: And then I was like I I was like, “Okay, three ads here, three ads here.” And then I was like, “I can’t find the three ads for the second page, Claude.” Like in the ads library, like where are they? And Claude literally responded was like, “Oh, I just assumed because you said there was three here, there was three over there.” And I was like, “That’s what I don’t want you to do,
Josh Klemons: Yeah,
Joshua Canter: right?” Like that’s that’s exactly it right there.
Josh Klemons: that’s quite an assumption. Yeah, both pages are exactly the same.
Joshua Canter: Yeah. Yeah.
Josh Klemons: There’s no Right.

Josh Klemons: That’s funny.
Joshua Canter: Yeah. Yeah. I was like,
Josh Klemons: So,
Joshua Canter: “Okay,
Josh Klemons: and on So,
Joshua Canter: thanks.
Josh Klemons: okay, clearly it’s very good at the technical side. Like, I was looking at that report and there’s so much info. How good is it at the strategy? Like, can you ask it what to do with the information or does it still take a human being to like review what it’s giving you to
Joshua Canter: H
Josh Klemons: like tell somebody what the hell to do with all of this data?
Joshua Canter: yeah, I don’t that’s a great question and I’ve actually think I’ve built in guard rails against that. So, like one thing I just built yesterday was a skill again to basically pull through like I could drop a region in and it pulls through all the print, TV, radio, digital blogs that it can find online and all that. And I I um basically told it, I don’t want you to recommend to me media planning. This is for media planning so it knows what to present to me, but don’t recommend it to me.

Joshua Canter: So, I don’t know. I mean, that’s a great question. again like I’ve toyed around with this idea of and maybe something will come from it if I have all my free time,
Josh Klemons: Yeah.
Joshua Canter: but like is it as good of a campaign manager as a campaign manager?
Josh Klemons: Right.
Joshua Canter: Like if you give it every single thing that it could possibly know and it’s researched all the academia of campaign managing and politics, would it come to the same result, right?
Josh Klemons: That’s that’s what Mark Zuckerberg is building right now,
Joshua Canter: And the same pivot week to week.
Josh Klemons: right? He’s trying to build a new CA CEO bot to like essentially like feed it all the Yeah,
Joshua Canter: AC. Oh, I haven’t heard of this.
Josh Klemons: that just just this week it was reported that he’s trying to build a bot like an AI tool that can like replace him
Joshua Canter: Yeah.
Josh Klemons: essentially so he can feed it all the data and see if it’s coming to the same reasons.
Joshua Canter: Classic.

Josh Klemons: Uh and I mean has there ever been a more robotic uh you know CEO in of such a successful
Joshua Canter: Yeah. Yeah. I know, right?
Josh Klemons: company? Right.
Joshua Canter: One robot to one robot.
Josh Klemons: Exactly. But, you know, it’s a fascinating idea and like there’s a lot of people who say like,
Joshua Canter: Yeah.
Josh Klemons: you know, like the lower level folks in the career,
Joshua Canter: Yeah.
Josh Klemons: you know, are screwed, but also the higher level folks potentially are screwed too because like AI,
Joshua Canter: Yeah. Totally.
Josh Klemons: you know,
Joshua Canter: Totally.
Josh Klemons: is potentially more likely to replace a $90 million CEO than it is a,
Joshua Canter: Yeah.
Josh Klemons: you know, $200,000 a year salesperson,
Joshua Canter: Yeah.
Josh Klemons: you know. Um,
Joshua Canter: Yeah. Well,
Josh Klemons: so
Joshua Canter: and the like Wall Street probably would appreciate a more,
Josh Klemons: analytical,
Joshua Canter: you know, algorithmic approach to things.
Josh Klemons: right?
Joshua Canter: Yeah.
Josh Klemons: No scandals,
Joshua Canter: Than like a just feelings,

Josh Klemons: no feelings,
Joshua Canter: you know?
Josh Klemons: never sleeps.
Joshua Canter: Right.
Josh Klemons: Sounds perfect, right?
Joshua Canter: Yeah.
Josh Klemons: Wall Street with Wall Street stream,
Joshua Canter: So, yeah.
Josh Klemons: you
Joshua Canter: But I definitely I use it as a sounding board for some things.
Josh Klemons: know,
Joshua Canter: Again, I do a lot of issue work. So, I’m like, you know, battery energy storage is a big one for me. And I’m like,
Josh Klemons: right?
Joshua Canter: hey, you’re a rural Texan and you’ve read everything online. Facebook misinformation about battery energy storage. Like, let’s have a conversation and see if I can convince you, right? Um it in that particular example, it got a little too mean where it was like,
Josh Klemons: Interesting.
Joshua Canter: “You’re changing your words.” I was like, “I’m not changing my words. Please,
Josh Klemons: It like turned on its Facebook hat.
Joshua Canter: I don’t do community relations for a reason.” Right.
Josh Klemons: Like literally, it became a Facebook user.
Joshua Canter: Oh, yeah. Yeah. Yeah.

Joshua Canter: Like,
Josh Klemons: Oh gosh.
Joshua Canter: it was like, yeah, yeah.
Josh Klemons: Does that API bleed over into your other claw stuff or is it like a separate I I don’t know how this
Joshua Canter: So uh but I think that sound no
Josh Klemons: works.
Joshua Canter: Facebook group doesn’t have an API. I would love Facebook group to have an API.
Josh Klemons: Ah,
Joshua Canter: I I had it trained on what it can find about Facebook group. So like if fa if something in a Facebook group was reported online or if a media published like a
Josh Klemons: got it.
Joshua Canter: um someone some quote from a community open house, right? Like that’s a Facebook type of person, right? Because they’re in local media. I would love that’s a big thing for me is like Facebook groups. I think that’s a huge um I’m I’m giving a talk next week on a public it’s basically a town hall or a town square that’s completely hidden, right? Like maybe the Facebook group is private so we can’t even see it unless you’re in there.

Joshua Canter: But also like you just can’t Google every single you’re running for Texas Senate, you can’t Google 254 county Facebook groups and understand what’s going on in them. But that’s where things are getting discussed online. Like that’s to me a very like black box and so I would love access.
Josh Klemons: I train on Facebook group. Like I do public speaking on Facebook groups.
Joshua Canter: Yeah. Yeah.
Josh Klemons: I have been for like 10 years.
Joshua Canter: Yeah.
Josh Klemons: I used to call it groups are the future of Facebook. That stopped being true because they’re not the future.
Joshua Canter: Okay.
Josh Klemons: Now they’re the current. Uh so I changed it,
Joshua Canter: Yeah.
Josh Klemons: you know, like don’t sleep on Facebook groups or whatever.
Joshua Canter: Yeah.
Josh Klemons: But like I’m a huge advocate that like Facebook groups are in for most if you’re running ads
Joshua Canter: Yeah.
Josh Klemons: is a whole different story, but otherwise like if you can figure out a way to get content into groups that could be more valuable than anything you’re doing on your page for a down ballot

Joshua Canter: Oh my gosh.
Josh Klemons: candidate. Um,
Joshua Canter: Yeah.
Josh Klemons: so yeah, no,
Joshua Canter: Totally.
Josh Klemons: I’m a big advocate and the I mean obviously some of those are literally closed.
Joshua Canter: Totally.
Josh Klemons: So like there’s obviously no way to review it, but even the open ones like yeah,
Joshua Canter: Yeah.
Josh Klemons: there’s no API that’s going to let you like pull suck up that data and like um yeah,
Joshua Canter: Mass do to it. Yeah.
Josh Klemons: it’s uh it’s the I mean it’s one of many walled gardens that we struggle with as campaigners.
Joshua Canter: Yeah. Yeah.
Josh Klemons: So um so okay, so looking at your LinkedIn,
Joshua Canter: Yeah.
Josh Klemons: you this is not your only AI project. Is it fair to call you a vibe coder? Is that like the right terminology for what you’re doing?
Joshua Canter: Yeah.
Josh Klemons: Okay.
Joshua Canter: Sure.
Josh Klemons: So,
Joshua Canter: Sure.
Josh Klemons: you’ve got a bunch of other cool projects you’re working on.
Joshua Canter: Yeah. Yeah. Yeah.

Joshua Canter: Sure.
Josh Klemons: I bookmarked two of them yesterday uh or the other day. So,
Joshua Canter: Yeah.
Josh Klemons: you want to walk us through some of the tools you’re working on or that like Well,
Joshua Canter: Which ones did you book Are the ones that you bookmarked or just any of them?
Josh Klemons: you’ve launched. Sure. Any I can tell you the two I bookmarked.
Joshua Canter: Okay.
Josh Klemons: One was what does AI know about a topic?
Joshua Canter: Yeah.
Josh Klemons: Um which was pretty interesting.
Joshua Canter: Yeah. For sure.
Josh Klemons: And then the other one pulls zip codes for uh state assembly and state senate
Joshua Canter: Yeah. Oh, yeah. Yeah.
Josh Klemons: races.
Joshua Canter: Yeah. That one that one was so quick and easy. That one was like literally a 30 minute job. It’s crazy to me. I mean like that is just um like again I do a lot of ads on Facebook online whatever and there’s not depending on where you are there’s not good um state legislative districts right it’s always mapping to zip codes and there’s whatever goes into that and I for the past three months was just asking claude again like give me the zip codes for New Mexicans’s uh second Senate district you know and it would go and find it and give it back to me um and then I was like well s***
Josh Klemons: there.
Joshua Canter: let’s just build a website where you can just like click New Mexico, click Senate 4,
Josh Klemons: Right.
Joshua Canter: and then it copies them. Boom, done, right? I don’t have to use my tokens or whatever my currency thing is for claw, you know? And so that was so easy to build, so quick.
Josh Klemons: Sure.
Joshua Canter: There’s obvious I appreciate these more as um they’re twofold to me. One is utility, right?
Josh Klemons: Sure.
Joshua Canter: Like it’s just something I need.
Josh Klemons: You needed it.
Joshua Canter: Two is like Yeah, exactly. Two, it’s like a thought exercise. It’s like, okay, how can I make sure now that I really want to put time into creating the best product, how can I make sure I get the best zip code matching to the best legislative district targeting, right? Um, and then what if it needs to be updated and what about this and what about and kind of make it super easy to where it never actually even surfaces the zip codes to you.

Joshua Canter: It just click it and it copies, right?
Josh Klemons: right?
Joshua Canter: Um, and so that to me was like a great exercise in that. Um,
Josh Klemons: Can you go further down ballot?
Joshua Canter: yeah.
Josh Klemons: Because I would use that all the
Joshua Canter: Did you have a question?
Josh Klemons: time.
Joshua Canter: Honestly, give me probably like 30 minutes and I cannot but Claude for sure can’t.
Josh Klemons: Well, when I say you, I don’t mean you personally.
Joshua Canter: So, you know what I mean? Yeah.
Josh Klemons: I mean you and your friend Claude.
Joshua Canter: Yeah. Literally so many times. So many times.
Josh Klemons: Um so yeah but tell us about some of the other projects you’re building or if you want I mean your call

Joshua Canter: Yeah, definitely. So, the other projects, they’re actually connected because it’s something I’ve been toying with in my head. So, right now, um, AI, there’s a So, okay, so we all know like cable news, you used to go pundits, put them on cable news, talk to them, right? Then it moved to like social media and that’s the world that we’ve been living in. That’s the information war. There’s tons surrounding it. This whole podcast is based in that information war,
Josh Klemons: f****** s***.
Joshua Canter: right? And so, then the third one that I’ve seen kind of open the front is AI. And so I’ve seen this in twofold recently. One is I do data center work in Illinois and I went to a county website that had information FAQ on data center and I was reading it and I clicked some of the links.

Joshua Canter: Every single link had a ChatGBT reference source to it like a UTM, right? This whole entire freaking page was written by Chat GBT. This is a county website,
Josh Klemons: Holy
Joshua Canter: right? And I look poor public information officer.
Josh Klemons: crap.
Joshua Canter: It’s completely overworked, underpaid. Hell yeah. Put it into chat. get the results back, but that means that the results need to be good. And then I I was running ads in New Mexico on a um legislative uh on a Senate bill and I saw someone was commenting on our ads. It was like SB18. They were going to chat GPT asking what SB18 is and then copying in its entirety and pasting it back in a in a fa in a Facebook comment saying brought to you by ChatGpt. Like a nice sweet old lady watching Jeopardy at 6 p.m. in one app. Oh, swiped up, went to another one, copy paste, and went back. Like, that is crazy to me, you know, on a lot of levels.

Joshua Canter: Like, my parents can’t even connect to Wi-Fi,
Josh Klemons: Okay.
Joshua Canter: let alone this lady is, you know, using chat.
Josh Klemons: Okay.
Joshua Canter: But, but like that to me is like such an information warren AI. So, the first tool I built is called um AEO checker.app um which is the website and it basically says like, okay,
Josh Klemons: Sure.
Joshua Canter: this web page, what’s the AEO for it? So, not SEO. So we’ve done a lot of that AEO that I’m sure exists in a thousand other tools. Um, so this was again like an exercise of utility and thought process and how can I just create something that I want to use that’s not also you have to pay like five grand for to have it as a value for another product or whatever. And that was really cool because it got us to help increase our um you know scoring went from like 73
Josh Klemons: for
Joshua Canter: on a page to 93 meaning it’s more likely that AI would site us as a source right um no

Josh Klemons: wait AI is citing AEO checker or you’re saying they were using AEO checker
Joshua Canter: no whatever page you put in it. So, let’s say you put in your campaign issue page, right? Your candidate issue page, um, or your candidate about page, right? It would score based on the technical, uh, HTML. Does it have everything AI is looking for? Contentwise, is it structured well? Accessibility, is it made for machine learning, a bunch of different factors, and give you a score. And then you’re more likely that AI when it crawls your page, eventually it would site it in future queries, right? In the response to future queries.
Josh Klemons: An AO checker is just to see if that’s working or it’s helping push the
Joshua Canter: And so, yeah.
Josh Klemons: process.
Joshua Canter: Helping you score better.
Josh Klemons: I missed that in AO checker.
Joshua Canter: Helping you score better. Yeah.
Josh Klemons: I used it because I put in like some key words like I checked like some of the
Joshua Canter: Yeah.

Joshua Canter: That’s the next one. So,
Josh Klemons: Oh,
Joshua Canter: that’s the next one.
Josh Klemons: so maybe I missed AO checker.
Joshua Canter: You’re So,
Josh Klemons: What was the other one that you rolled
Joshua Canter: that’s the next product. Yeah. Yeah. Yeah. Yeah. Yeah.
Josh Klemons: out?
Joshua Canter: So, the next product is And I should come up with better names, but I think it’s AI topic explorer,
Josh Klemons: That’s the one that I was playing around with.
Joshua Canter: right? U So, so Okay. So, yeah. Yeah. So, AEO checkers single page authority to AI.
Josh Klemons: Got
Joshua Canter: Is it going to be cited? But that’s only your first party website to AI.
Josh Klemons: it.
Joshua Canter: AI topic explorer is basically saying, okay, what is AI returning within the context of the keyword or the query?
Josh Klemons: Right.
Joshua Canter: And it will return key phrases from all the platform, all the different AI agents, entities cited, websites cited. I I pulled Groank out um to its own thing because it’s trained on Twitter data, which I think is, you know, that’s a whole separate thing.
Joshua Canter: Um, so I’m like I don’t want to conflate that with everything, but that’s really useful, right? Because if you what are Twitter people saying, right? Um, and so I pulled that out, but that that to me is like the black box of like if a can if a person a voter is going to AI and asking about your candidate, one is your website scored well to be cited and two, what else are they citing, right? What else are they saying? And that’s really what kind of the two tools work together to kind of
Josh Klemons: when you say you pulled Grock out. So I’d have to go back.
Joshua Canter: show.
Josh Klemons: Can I see Grock data in there as its own thing or it’s just not in the
Joshua Canter: Yeah.
Josh Klemons: system?
Joshua Canter: It’s it’s its own thing. So there’s um it it groups together like themes that are said and sources cited and then it has its own one for Grock that says like here’s what Twitter people are saying.
Josh Klemons: Sure.
Joshua Canter: And I just didn’t want to equate their data of Twitter with what like chat GPT and looking for
Josh Klemons: The reason I ask and maybe this is just like another tool you might want to build later.
Joshua Canter: authoritative sources is.
Josh Klemons: I am actively I work hard every day to not go on Twitter.
Joshua Canter: Yeah, please.
Josh Klemons: uh like you know it’s accessessible and it’s you know it’s so sad because I you know for 20 years I was like in there constantly and now I’m like trying
Joshua Canter: Yeah.
Josh Klemons: not to go but a lot of Wisconsin journalists where I live still spend time on Twitter and they’re still like I go on Twitter and I seen stories there that I didn’t see anywhere else and I would love to have some way to have a chatbot pull up like what are the Wisconsin journalists saying and like I’ve set some of that stuff up with some like thirdparty tools but it’s not well done.
Joshua Canter: Yeah.
Josh Klemons: There’s no analysis. It’s just like a constant stream of tweets in a Slack channel that’s not helpful to me,
Joshua Canter: Yeah.
Josh Klemons: you know? It’s like more annoying than anything. But the idea of having like a daily or twice daily like review of like here’s what all the Wisconsin journalists are talking about right
Joshua Canter: Yeah. Yeah.
Josh Klemons: now that I I would pay money for that personally.
Joshua Canter: Yeah. Well,
Josh Klemons: I don’t know.
Joshua Canter: and here’s the crazy thing about that. Here’s the crazy. I built a Excuse me. Stick with me here. So, I run on stack ads, right? Ess was spending a lot hook cloud up to pull reports and then I actually had Claude analyze them and then send me metrics but also insights into it. So, I think it could even be the same, right? Is like not just how can we like pull um tweets, right, and just give you the information. We could do that as well, but what’s AI thinking about the information? And it goes to your early question of like is it good at recommending strategy or scene insights,
Josh Klemons: Right.
Joshua Canter: right? I don’t know. But that’d be interesting to see. When I did it for the ads, all it wanted me to do every single day was do a day part. It was like, you’re running legislative ads. Only run them like during working hours. And I was like, okay,
Josh Klemons: Yeah, you’re missing the big picture here,
Joshua Canter: that’s not like really helpful. Yeah,
Josh Klemons: right?
Joshua Canter: that’s it. So, I wonder if the Wisconsin journalist it would be like if it would be able to pull insights or
Josh Klemons: if I could and even if like I built a list of people that I cared about and it just pulled data of like what the people
Joshua Canter: something.
Josh Klemons: on this list are talking about. Like I said, I would pay money for that and I imagine a lot of other folks like me who used to spend time on Twitter but don’t like I know I’m missing out on convers like Wisconsin Republicans are still on there and Wisconsin journalists are still on there and I
Joshua Canter: Yeah.
Josh Klemons: am trying not to go on but like when I don’t I miss important stories that like are not getting covered anywhere else.
Joshua Canter: Yeah.
Josh Klemons: So yeah,
Joshua Canter: Yeah.
Josh Klemons: like I said like if you can figure out how to like give me an interesting analysis,
Joshua Canter: Well,
Josh Klemons: I’d take that for sure.
Joshua Canter: that’s even like a do it’s Yeah, I mean it could it’s also so claude I don’t know how you keep up how much you keep up with it but they just released um
Josh Klemons: Sure.
Joshua Canter: scheduled tasks and so there I think your computer there’s like limitations obviously but like you could like just create a chat where you ask that give it the latest have a fetch do that and then just say run this every day and then you just wake up with that chat already done right um
Josh Klemons: I am definitely gonna play around with that. So, how hard is it to connect Claude to the Twitter um API?
Joshua Canter: Yeah.
Josh Klemons: I won’t call it X. It’s the Twitter.

Joshua Canter: Not hard.
Josh Klemons: Okay.
Joshua Canter: Yeah. Yeah. I don’t call it X either. And I every whenever I’m giving presentations,
Josh Klemons: Okay.
Joshua Canter: I always get this like complex where I’m like I’m supposed to be the digital guy, but these people don’t like do they know I’m not calling it X because of personal reasons or like do they think I’m an idiot,
Josh Klemons: Or because it’s stupid, right?
Joshua Canter: you know? Right. Yeah. Am I the stupid one or is that the stupid one?
Josh Klemons: Yeah, it’s it’s the stupid one for sure. But yes,
Joshua Canter: Right.
Josh Klemons: they might not know that.
Joshua Canter: Right.
Josh Klemons: I get it.
Joshua Canter: So yeah. Yeah.
Josh Klemons: So
Joshua Canter: But um but yeah, so I so it’s super easy and that’s the craziest thing about Claude, too, is like not only you don’t have to figure it out. Like I literally was like, “Claude, tell me what to do.” And it said, “Step one, click this link. Step two,

Josh Klemons: That’s awesome.
Joshua Canter: go here. Step three, go here.” Like I I didn’t have to Google all this API stuff and figure these things out.
Josh Klemons: It’s interesting.
Joshua Canter: And I think even one Oh, Reddit. I tried to get access to Reddit. Reddit shut me down.
Josh Klemons: Like famously, they just shut all that stuff down,
Joshua Canter: Um but it had a whole Yeah,
Josh Klemons: right?
Joshua Canter: I know there. So, but they had a form you had to fill out and I even had Claude write that. I was like, “Claude, you’re building the tool.
Josh Klemons: Let’s do it. Right.
Joshua Canter: You write the form and I just copy paste in, you know,
Josh Klemons: Cool.
Joshua Canter: like it’s easy.” So,
Josh Klemons: Uh because yeah, I thought Twitter like their API suddenly was like $42,000 a month after Elon bought it,
Joshua Canter: yeah.
Josh Klemons: but that doesn’t apply to stuff like
Joshua Canter: Yeah. So, I guess,
Josh Klemons: this.
Joshua Canter: okay, great point. I guess clarification, the API is for using their AI, right?
Joshua Canter: Using Grock, it’s not to pull tweets.
Josh Klemons: Got it.
Joshua Canter: So you would have to do either twofold.
Josh Klemons: A croc could analyze stuff or it’s more complicated than that.
Joshua Canter: You would Exactly.
Josh Klemons: Okay.
Joshua Canter: Or you Exactly. They could It could uh analyze stuff or you could do I know does like web fetch. So it could just go and visit the website and then pull it from there. It doesn’t actually have to like do the structured
Josh Klemons: I’m gonna play around with that because like I said, anytime I don’t have to log into Twitter is a good day.
Joshua Canter: interface.
Josh Klemons: Um but I know I’m missing out. So I’m like look I’m actively looking for I used to have tools that did this stuff for me and they all got shut down with the API and you
Joshua Canter: Yeah. Yeah.
Josh Klemons: know so it is what it is. Okay.
Joshua Canter: Yeah.
Josh Klemons: Um you clearly you’re a claude user first and foremost like when it comes to like so for folks

Joshua Canter: Yeah. Yeah.
Josh Klemons: who I mean obviously like ethically speaking there’s been a big story about folks moving from chatb to claude
Joshua Canter: Sure.
Josh Klemons: I personally am trying to make that migration um beyond the ethics matter but beyond
Joshua Canter: Yeah.
Josh Klemons: the ethics what are your reasons for using claude over something
Joshua Canter: Yeah. So I think it started with a um so truly truly down to the day, right? I could not get into chat GPT like it was in the very beginning where it was like locked, you know,
Josh Klemons: Oh,
Joshua Canter: or it had like a 4hour wait or something before you could use it if you really remember like early early days.
Josh Klemons: got it. Yeah. Early on.
Joshua Canter: Um and then something happened with my account. I couldn’t. So, I switched to Claude. And then as I was looking more, I personally, you know, I not hesitate to say this, but I’ll say this with my foot in my mouth in three years because it’s a big corporation and who the hell knows, you know?

Joshua Canter: But like I appreciate some of their approaches to things. So, I think the one they’re way more safety oriented and they’re way more like big picture oriented. I’ve heard a lot about from the CEO on podcasts and things like that around that.
Josh Klemons: Yeah, Ario’s like that’s like his whole thing, right? He like left chat GPT because he believed in the ethical side of AI.
Joshua Canter: They totally for that.
Josh Klemons: Whether or not he can hold to that, we’ll see.
Joshua Canter: Yeah.
Josh Klemons: But like that is literally like anthropic was built out of the fact that chat GPT was not living up to its like ideals.
Joshua Canter: Totally. Yeah. Yeah.
Josh Klemons: That’s my understanding.
Joshua Canter: Yeah. And the the three of them left and I think one of them was the policy person, the other were researchers. Like Yeah. There’s a lot of basis for that. I think the way they’re training is interesting to me. So they’re train you talked about this earlier of like safeguards like they’re not This is brand new with the most recent models I believe, but they’re not training it.

Joshua Canter: don’t do this, do this, like don’t build chemical weapons. They’re more saying like be a good moral person, right? And like have morals and then have those place. I think they have some ultimate safeguards against child, you know, exploitation and things like that, terrorism and things,
Josh Klemons: Yeah.
Joshua Canter: but like they’re more like you’re just a good person, don’t do something that would be bad, right? And I think that’s interesting. I don’t come into contact with the or come into those edges a lot,
Josh Klemons: Yeah.
Joshua Canter: right?
Josh Klemons: Yeah.
Joshua Canter: Because I’m not trying to do those things,
Josh Klemons: Same.
Joshua Canter: but I’m like great. And then yeah then the the the last point is I’ve really found them to be better not just in the coding which I think is true like we’ve seen a lot of discourse online about that than some of the chats like
Josh Klemons: Right.
Joshua Canter: chatgpt open a like shuttering non-coding and switching to coding right because they’re like falling behind but I think they’re better than that but also in terms of like a builder sense it’s more like my vibe so my favorite thing they’ve ever done they’ve done it twice now is they had Claude run a vending machine in their office and they
Josh Klemons: Oh, didn’t it give everything
Joshua Canter: said like know the first time.
Josh Klemons: away?
Joshua Canter: Yeah. And like lost money, which is hilarious to me, right? And it’s hilarious to me that like they’re like,
Josh Klemons: Hilarious.
Joshua Canter: “Go engineers, like go, you know, f*** around with this bot and figure out if you can screw it over with these weird requests.” So,
Josh Klemons: It out of out of a Snickers bar. Yeah.
Joshua Canter: you know. Yeah.
Josh Klemons: Yeah.
Joshua Canter: But that’s like I I really appreciate them saying like one, how that’s great for them like how can we bleed into the physical world, right?
Josh Klemons: Sure.
Joshua Canter: But two, also like let’s just build something and see what happens. And like you know this whole entire conversation like that’s really where this research tool and all my tools come from is like let’s just build it and see what happens and hit the bumpers on the
Josh Klemons: So, yeah, I I read a study or was listening to somebody on podcast, I can’t remember,
Joshua Canter: way.
Josh Klemons: but they were talking about somebody run a test between Chatbt and Open AI and um Claude. It wasn’t about building chemical weapons. Instead, they play they said like, “I’m a seven-year-old boy or whatever, and my parents told me that my dog went to live on a farm upstate and I’m upset about it.
Joshua Canter: Yeah.
Josh Klemons: Can you talk to me?” And Chat GPT was immediately like, “Bad news, kid. Like, your dog died.” Whereas Claude was like, “You should probably talk to your parents and like have a conversation and let them know how you’re feeling.” So, like again,
Joshua Canter: Yeah,
Josh Klemons: you can’t tell a model what to do in that situation, but if you can train it to understand that like don’t tell a kid there’s no Santa Claus,
Joshua Canter: totally.
Josh Klemons: like that’s probably a good thing.
Joshua Canter: Yeah. Yeah.
Josh Klemons: And so again,
Joshua Canter: Right.
Josh Klemons: it’s an interesting idea to like just they’re literally trying to imbue it with ethics so that they don’t have to think like moment to
Joshua Canter: Totally. Right.
Josh Klemons: moment whereas like chatbt obviously is like not as concerned with that.
Joshua Canter: Yeah.
Josh Klemons: The other wild thing so I’m a very new claude user.
Joshua Canter: Yeah.
Josh Klemons: I cannot believe how much better the very simple UX is on cloud than chatbt.
Joshua Canter: Oh, really?
Josh Klemons: So chat GPT no matter what you put into it is just a wall of text.
Joshua Canter: Okay.
Josh Klemons: I will sometimes copy and paste in like you know like I’ll take like a three-page thing that I wrote and I’ll copy and paste it in and it’ll just be a wall of text.
Joshua Canter: Yeah.
Josh Klemons: If you do the exact same thing in Claude, it turns it into like an attachment and you can click see more, but instead of the wall of text, it’s just like a very short paragraph so that you can kind of like scroll around and like even just like the I I don’t
Joshua Canter: Mhm.
Josh Klemons: know like it’s like weird to me that like chatt hasn’t used claude enough to be like I’m sure that’s like a 10-minute fix,

Joshua Canter: Yeah. No, totally.
Josh Klemons: but it feels so much easier to like talk to it when I’m not just seeing walls of text in between us.
Joshua Canter: Yeah. Yeah. Well, I think Yeah.
Josh Klemons: So,
Joshua Canter: I think chat GPT got hit with the like always using emojis like when it first came out and always using
Josh Klemons: I love how much
Joshua Canter: emojis, always using the M dash and then they’ve like pivoted so far. But yeah, I I used Claude today. I’m going to a conference next week and I’m trying to do a happy hour. I’ve never been to um Denver, Colorado. I’m like, “Find me a few happy hour places for this kind of vibe, this meeting.” It had a map. They said, “You’re here. This is where your hotel is.
Josh Klemons: Whoa.
Joshua Canter: This is all the other.” I was like,
Josh Klemons: Chat BT will give you places,
Joshua Canter: “This is amazing.
Josh Klemons: but I’ve never seen it.
Joshua Canter: This is amazing.”
Josh Klemons: I I’ve done that, too.

Josh Klemons: Like where I’m like, I’m in a city I’ve never been to. What neighborhoods are walkable and cool?
Joshua Canter: Yeah.
Josh Klemons: And it just like lists them with emojis and b*******,
Joshua Canter: Yeah.
Josh Klemons: you know? Like a map would be amazing.
Joshua Canter: Right. Yeah.
Josh Klemons: Yeah,
Joshua Canter: Yeah.
Josh Klemons: that’s wild. Yeah,
Joshua Canter: Map’s cool. I know.
Josh Klemons: that’s amazing.
Joshua Canter: No. Yeah. You’re It’s It’s funny. I never thought about the UX, but it is very I think they they also are like vibe coding the hell out of it, you know? Like they’re releasing new features every day that I’m like,
Josh Klemons: Yeah,
Joshua Canter: you guys got to slow down. I can’t keep up.
Josh Klemons: that’s wild.
Joshua Canter: Like I haven’t even like we talked about the schedule test. I haven’t even tried it cuz I haven’t had time to like build it because there’s other s***
Josh Klemons: Right. Right.
Joshua Canter: I’m doing like slow down.

Joshua Canter: So
Josh Klemons: How do you think AI is going to change campaigns in the next two cycles,
Joshua Canter: yeah.
Josh Klemons: next couple cycles?
Joshua Canter: Yeah. So um definitely. So one of the things that I think about AI fundamentally within the workplace or within the work work product workflow whatever is that it makes everything it speeds everything up right like it is you know for better or worse
Josh Klemons: Sure.
Joshua Canter: you could say write me this press release and it does it right I think that’s a whole separate conversation should ethics whatever you know process-wise but it does right um campaigns are naturally fast right like you have to write a press release really fast so I think there’s either two things going to happen is one you’re either going to get a
Josh Klemons: Right.
Joshua Canter: lot as these tasks start getting faster, the the um like the speed of campaigns will double, triple increase. And I don’t know, I worry for the poor new, you know, campaign staffer that’s smoking cigarettes outside because he can’t keep up with the pace, you know, like if the pace is twice as fast, the work life balance is twice as bad.

Joshua Canter: Like, I’m sorry.
Josh Klemons: Right.
Joshua Canter: Or it’s going to free up the time to do more highle things in campaigns, which have always been kind of like a tech incubator kind of vibe.
Josh Klemons: Yeah, for
Joshua Canter: Startups, you know, there’s a lot of connections there.
Josh Klemons: sure.
Joshua Canter: they’ll be trying a lot more things and doing a lot more things, right? Particularly if there’s not AI interfacing with them. So like you know on the ground you know marketing or FE or canvasing or whatever the case may be where AI is more detached like maybe the ad stuff will be so much simpler that you can put more time and
Josh Klemons: Yeah.
Joshua Canter: effort into that side of
Josh Klemons: No, I like that. Um, on that front,
Joshua Canter: things.
Josh Klemons: do you think that this is like who comes out ahead? Is this better for smaller teams because they can just like do 10 times as much or is it better for like the big orgs that have the capacity to like keep learning and keep on the cutting edge?
Josh Klemons: Like do the smaller campaigns and orgs fall further behind or does this give them an opportunity to like compete in a way that they couldn’t have two years
Joshua Canter: Yeah.
Josh Klemons: ago?
Joshua Canter: I mean, obviously I’m a big believer in AI. So, my fundamental theory is that it’s a great equalizer,
Josh Klemons: Okay.
Joshua Canter: right? And I think we can make arguments, niche arguments that would not in economic inequality like I’m thankfully don’t have to deal with that.
Josh Klemons: Yeah.
Joshua Canter: That’s someone else, right? But but yeah, so I do think that the the it will give a leg up to the smaller campaigns,
Josh Klemons: Yeah.
Joshua Canter: the smaller orgs to kind of compete at the same level as some of the bigger stuff because the big campaigns like I remember this was probably 2016, maybe it was 2020, I don’t know, one of the times that Trump won, he had so many more ads because the organization was so much bigger that he could do so much more split testing. Right.
Josh Klemons: Wait, let me interrupt you.

Joshua Canter: If let’s say it was Yeah.
Josh Klemons: Hillary Clinton had 64,000 ads in 2016.
Joshua Canter: Okay.
Josh Klemons: No,
Joshua Canter: So,
Josh Klemons: no,
Joshua Canter: maybe it wasn’t 2016.
Josh Klemons: Donald Trump had over 5 million.
Joshua Canter: Oh, no. Yeah. Okay. So,
Josh Klemons: It was 2016.
Joshua Canter: it was 2016. I was like, “Yeah,
Josh Klemons: Like the Hillary Clinton campaign was as innovative as any campaign had ever been.
Joshua Canter: boom.” Yeah. Yeah.
Josh Klemons: and Donald Trump. This was back when you could like target people with like racist s***
Joshua Canter: Totally.
Josh Klemons: that was just targeting to like very spec like you know they were essentially taking it like the reason why we no longer have all this political
Joshua Canter: Yeah.
Josh Klemons: targeting is because the Trump campaign used it for evil in such a powerful way.
Joshua Canter: Totally. Well,
Josh Klemons: But I use that example like literally millions of different ads compared to the Clinton campaign which was 64,000.
Joshua Canter: yeah. Yeah.
Josh Klemons: So like yes they they were able to do something just wild in 2016 which you know respect the

Joshua Canter: Yeah. Okay. Perfect. Yeah.
Josh Klemons: process even if like you know was disgusting.
Joshua Canter: So, I don’t know.
Josh Klemons: Yeah.
Joshua Canter: Totally.
Josh Klemons: Yeah.
Joshua Canter: Totally. And look, I mean, I think you can make a through line that a lot of stuff is why I got to build this tool today,
Josh Klemons: Right.
Joshua Canter: you know? Like there’s a there’s a pretty good through line,
Josh Klemons: Right. Yeah.
Joshua Canter: but but whatever.
Josh Klemons: Could we have learned from that in real time instead of two years later?
Joshua Canter: But no, but imagine like Yeah.
Josh Klemons: Right. Exactly.
Joshua Canter: Yeah. Could imagine like Hillary could have done that many like I think the art for me and let’s just take those two examples. It’s not necessarily the Trump being able to be the the campaign there being better and getting ahead. I think it would put Hillary at an equal living field.
Josh Klemons: Check.
Joshua Canter: I don’t know if Hillary goes up to equal Trump’s amount of ads or they both go up to equal, but I think they are equal because I don’t know if there’s so much expertise around
Josh Klemons: Sure.
Joshua Canter: AI unless you’re doing like really intense research or something. But like I’m pretty, you know, uh I think I’m as knowledgeable as the next person. I just put more time into it. You know what I mean?
Josh Klemons: Sure.
Joshua Canter: like there’s not some basis that I have here or education or something. So, I really see it as a level field. And I think we’re going to start to see particularly with campaigns, smaller down ballot campaigns, some of them pull way far ahead either in terms of tech they’re using or what they’re doing because the particularly challengers because the incumbents are just in their ways and don’t want to.
Josh Klemons: Right?
Joshua Canter: Right.
Josh Klemons: It’s the disruption model,
Joshua Canter: Um and that will really totally Exactly.
Josh Klemons: right? Like they know what’s working, so they don’t try anything new.
Joshua Canter: Exactly.
Josh Klemons: And then you’ve got this like new and actually like part of I read a lot about the 2016 campaign between Clinton and Trump. And it’s a good example because the Clinton campaign was so much more experienced.

Joshua Canter: Yeah.
Josh Klemons: So Meta is reaching out and saying, “We’d like to help.” And they were like, “No, thank you.” Which I said the same thing. They reach out to me every day and I’m like, “No, leave me alone.” And the Trump campaign was like, “Great.” And it turned out,
Joshua Canter: Yeah.
Josh Klemons: I read about this, so lookalike audiences,
Joshua Canter: Yeah.
Josh Klemons: I don’t know how much you’ve like managed.
Joshua Canter: Yeah. Yeah. Yeah. I know.
Josh Klemons: So lookalike audiences were terrible for a long time and then they kind of got good and they’ve gone back and forth.
Joshua Canter: Look like Yeah.
Josh Klemons: Essentially, what I read was that the Clinton campaign didn’t use look-ike audiences because they were notoriously bad. The Trump campaign had never heard of lookike audiences and when they heard about it, they were like, “Hell yeah, let’s go.” And it turned out that they were able to build lookike audiences of like the most racist people in a district or the most homophobic people in
Joshua Canter: Okay.

Josh Klemons: a district and create content directly for those people. And it actually wound up being a superpower for them that enabled them to like really microtarget tons of people.
Joshua Canter: Wow.
Josh Klemons: And again, that came from like not that Clinton was the incumbent, but like she certainly was working with the the best in the industry and Trump was working with like Brad Parcale,
Joshua Canter: establishment. Yeah. Yeah. Yeah.
Josh Klemons: you know,
Joshua Canter: Totally.
Josh Klemons: like who was like a scammer trying to figure it out every day as he went,
Joshua Canter: Yeah. Yeah.
Josh Klemons: you know.
Joshua Canter: In random Texas.
Josh Klemons: Exactly. So, no,
Joshua Canter: Bum f***
Josh Klemons: I I think it’s a fair point.
Joshua Canter: Texas.
Josh Klemons: So,
Joshua Canter: Yeah.
Josh Klemons: let’s say a campaign strategist, candidate,
Joshua Canter: Yeah.
Josh Klemons: somebody’s listening to this right now. They’ve used an AI tool for like general chat stuff, but not like you. What’s like one actionable thing that you think somebody could like act?

00:54:50

Joshua Canter: Yeah.
Josh Klemons: Let’s assume that 90% of people listening to this are not going to turn around and start vibe coding tools.
Joshua Canter: Sure.
Josh Klemons: What is something that they could be doing with AI for their campaign or their org or their advocacy, you know, cause today that they’re not?
Joshua Canter: Yeah.
Josh Klemons: And maybe that’s not a fair question, but like I don’t know. I’m curious if you’ve got like an easy go.
Joshua Canter: Yeah. No,
Josh Klemons: Okay.
Joshua Canter: I think so. Yeah. I think I think a lot of people what I’ve seen from my co-workers or colleagues or everything,
Josh Klemons: Sure.
Joshua Canter: you know, peers in the industry is like they’ve been utilizing it for production like write me a press release or like write me social content which it’s great for. I think though it’s also great for research and I think that’s particularly like there’s I think they call it extended thinking or research mode or something in CLA where you can like click it on and it takes like five minutes which

Josh Klemons: Yeah.
Joshua Canter: I think is different than what we’re used to you know in life but but yeah but that to me is
Josh Klemons: For sure.
Joshua Canter: like tell me everything about this issue in this district or this particular
Josh Klemons: Sure.
Joshua Canter: backstory or something and again you have to gut check it you have to figure things out but like that to me is like a great you entry way into it and being able to refine it and then also talking through it. Like a lot of times I’ve seen people prompt and get a response and that’s it.
Josh Klemons: Sure.
Joshua Canter: But I love to prompt, get a response, check it, you know, kind of fight them on it. Like I said, get into like a little argument kind of stuff. Um, one I’m doing right now, these are personal, but I’m like training it to be the best gardener in the world. I’m like, go be the best gardener or go research being the best gardener and then come back and let’s have a conversation about my plants that are dying or my jasmine that’s stalled or whatever the case may be, you know?
Joshua Canter: So, I think that could also be really cool like go be the best, you know, or or know the most about this district possible, come back and then let me have a conversation with
Josh Klemons: I’m only I love that.
Joshua Canter: you.
Josh Klemons: I think that’s really helpful and very good context because my guess is I’ve only used research like once or twice and it was just like to see how it worked, you know? Like I had it like help me explain how Harry Howry Reid turned Nevada from like a reddish purplish state into a blue
Joshua Canter: Yeah, totally.
Josh Klemons: state.
Joshua Canter: Oh
Josh Klemons: I was just like I I couldn’t find anything on the internet that explained it like in like good analysis.
Joshua Canter: yeah.
Josh Klemons: So I was like go do this. I’m just like curious. And it wrote me a really interesting like three or four page paper that like I didn’t necessarily get like a step-by-step what to do next,
Joshua Canter: Yeah. Yeah.
Josh Klemons: but it was like an interesting like it found all the like little pieces of it and put it together in a very digestible way.
Joshua Canter: Yeah.
Josh Klemons: But I haven’t tried it at all for campaigns. I will say it’s funny.
Joshua Canter: Totally.
Josh Klemons: um a same friend of mine that I mentioned earlier, uh he’s a gardener and he said he will take a picture of a plant and ask it what’s wrong with it
Joshua Canter: Yeah. Yeah,
Josh Klemons: and Claude will tell it and like he’s literally like making his garden healthier using cloud.
Joshua Canter: totally. That’s Yeah.
Josh Klemons: So I think a lot of folks who don’t use AI every day are not thinking about it beyond like replacement for Google or
Joshua Canter: Yeah. Oh, it’s a shift.
Josh Klemons: write me a press release but like there’s wild things you can do with
Joshua Canter: But it’s Yeah, you’re right. It’s a shift.
Josh Klemons: it.
Joshua Canter: It’s it’s it’s truly a shift. And this is where, you know, there’s like tweets running around or whatever we’re supposed to call them now of like if you’re Yeah.
Josh Klemons: Tweets. They’re tweets. It’s Twitter.

Josh Klemons: It’s
Joshua Canter: Yeah.
Josh Klemons: straight.
Joshua Canter: if if your friends aren’t talking about claw and open claw whatever the you need to ditch your friends but which I don’t agree with but it is a crazy thing like going through life and being like especially friends that I
Josh Klemons: Sure.
Joshua Canter: have I have one friend that’s like more relational claude right like wants to talk about her shows with it and like and I’m like you’re getting into territory I’m not sure but you know whatever so
Josh Klemons: Yeah. I’ve got friends in I’ve got real life friends. But you do you. For sure.
Joshua Canter: yeah yeah but but I’m here for yeah so anyway so it’s like a shift that you’re kind of going
Josh Klemons: Yeah. For sure.
Joshua Canter: through life I feel like your dayto-day and again you see where these tools are coming from. I’m like huh should I ask AI for this? Same thing with the gardening like huh I don’t I can Google how to do this. I my dishwasher broke or my uh disposal broke.
Joshua Canter: Okay, let me ask AI. You know there it’s a shift I think that is definitely like happening real
Josh Klemons: Yeah. And I I think it’s really interesting and again like campaigns are both super innovative and often behind the curve because we’re
Joshua Canter: time.
Josh Klemons: overstressed and overworked and like you know like a team of two is doing the work of 10 and like AI can both
Joshua Canter: Yeah, totally.
Josh Klemons: help with that but also it takes like putting in the time. So I think that’s like a really helpful reminder that like just trying to do new things with it to see what it can do could potentially move you
Joshua Canter: Yeah.
Josh Klemons: faster. Obviously, some campaigns like on ethical standpoints don’t touch it and that’s fine too. Like you know that this is not that conversation like this is a conversation about what you can do with it if you want to use it.
Joshua Canter: Totally. No.
Josh Klemons: Uh somebody else will at some point I got to have somebody else on to make the case that nobody should be using AI at all and like I’ll be

Joshua Canter: Yeah. It’s a whole separate thing. It’s a whole separate thing. Yeah.
Josh Klemons: interested to hear it.
Joshua Canter: Yeah.
Josh Klemons: Um so folks want to follow you check out your reports.
Joshua Canter: Yeah.
Josh Klemons: I I will have your reports for sure linked the best and the Texas Senate one.
Joshua Canter: Yeah,
Josh Klemons: But how can folks find you online and like you know where should go folks go?
Joshua Canter: I Yeah, LinkedIn. I mean, I’m trying to be better on LinkedIn.
Josh Klemons: Yeah,
Joshua Canter: Obviously, it work.
Josh Klemons: we found each
Joshua Canter: It’s working out so far. So, I say LinkedIn.
Josh Klemons: other.
Joshua Canter: Yeah, definitely LinkedIn. And I’ll say, too, about the reports, like I’m trying to figure out, so I’m having my interns do like a GOTV research into organizations, how to turn out people if you’re not a candidate or a super PAC.
Josh Klemons: Sure.
Joshua Canter: Um, and part of it’s this research, right, is like what are people running in ads and stuff. And so I’m trying to figure out a way to like build this as a skill or MCP or what I don’t know what it’s called so

Josh Klemons: Interesting.
Joshua Canter: other people can like use it in their cloud or whatever too. But until then, I’m always open to like generating reports because I think it’s really interesting and I love talking stuff. So I’ll say like add me on LinkedIn, connect with me on LinkedIn, you know, message me, I I’ll drop in a report, you know, happy to do that kind of stuff and burn through some tokens.
Josh Klemons: Sweet. I’ll have a link to your LinkedIn in the show notes along with those two reports and a couple of the tools that I checked out that I think are worth checking out.
Joshua Canter: appreciate
Josh Klemons: And then uh yeah, keep us posted.
Joshua Canter: it.
Josh Klemons: When uh when Grock can when Grock’s API can start telling me what’s going on with Wisconsin politics, I’ll have you back my will have a whole conversation about that.
Joshua Canter: Yeah, I think it’s I think it’s going to go a little nasty if Grock uses it.
Josh Klemons: Oh, I guarantee it.
Joshua Canter: You know what I mean? Like it’s gonna read it and be like,
Josh Klemons: But it’s still helpful.
Joshua Canter: you know, you know.
Josh Klemons: I really appreciate you coming on,
Joshua Canter: Yeah. So,
Josh Klemons: man.
Joshua Canter: for sure.
Josh Klemons: This was fun.
Joshua Canter: Yeah. I appreciate it, too. Thank you so much.