The Ledger
Every statement that passed quotation and attribution checks. Mix any filter with any other: certainty 1/5, debate potential 5/5, or both at once.
why aren't all 99 resolved? a statement only gets an assessment when the public record can support or contradict it. opinions and what-ifs never can, and 0 checkable ones are still open, waiting for their date. predictions held up or didn't; assertions are supported or contradicted. on every card: ▮▮▮▮▮ certainty · ▮▮▮▮▮ debate potential. speakers are clickable
Farroki: Contextual prompt chips directly boost AI app engagement and proper usage
“These details, like, it does actually, not even for the feeling, it has a direct impact on the usage. Like, people will probably click these things, and there's just a higher chance that they'll, they'll use it more properly if you have this.”
Farroki: ChatGPT beats Google Gemini for creating initial app mascots
“It might surprise people, but I've found also that for coming up with novel mascots, chat CPT is actually better than Gemini, better than nano banana from Google. I think nano banana is very good if you want to make iteration. So when you have the base mascot,…”
Farroki: Hire human artists for initial mascot concept before generating AI iterations
“My pro tip for anyone who's doing this is actually to bring on or hire an actual artist to generate the first version or some sort of concept, and then use that and feed it into ChatGPT or whatever you're using to generate as a reference and say, hey, can you …”
Nixon: ChatGPT Image Generation Is Absurdly Slow Compared to Nano Banana
“And so this is why I love Nano Banana Man. It's so quick. Like I used to use chat for image generation. It's just so absurdly slow.”
Schneider: Tally's ChatGPT referral traffic plummeted after a model update
“A great example of this is the Tally founder has this graph where it's like, absolute liftoff, right? It's going crazy. GPT-V drops, they, like, it changes the entire search algorithm within ChatGPT, and their referral from that, their, like, basically referra…”
Schneider: Site owners charge $500 per link on AI-referenced URLs
“They're like paying basically like 500 dollars per link that's being included on these, right?”
James: Local Google search ranking is far more valuable than ChatGPT inclusion
“Like, for this company here, like, ranking high on Google is way more valuable than going and showing up in ChatGPT.”
Isenberg: GenSpark is not a ChatGPT killer, but a valuable tool addition
“My gut take on GenSpark is, you know, is it the madness killer? Is it the ChatDBT killer? No. But is it interesting and something that you want to put as a part of, like, your cocktail of different tools that you're using to get a diverse opinion on a particul…”
Greg Isenberg: AI-assisted Fiverr gigs are software apps waiting to be built
“So these are manual workflows that are apps waiting to be built.”
Greg Isenberg: ChatGPT will soon unbundle into vertical SaaS and specialized agents
“So that's what you're going to kind of see with ChatGPT is it's going to be like unbundled. There's going to be like very specific vertical agents, vertical SAS that are specific
To that”
Fricks: Prompt AI with lowercase Gen Z slang for authentic copy
“I think the hardest part is getting the copy right. Like you, I'd say direct your, direct AI to sound like Gen Z. It's pretty good at it. Use lowercase, like those things kind of make it feel not AI.”
Smith: Custom GPTs require specialized proprietary data to monetize
“In order for a custom GPT to, I think, be able to monetize, you do need some sort of, ah, specialized data that other people don't have. Because if that same thing can be done through ChatGPT or Claude, right, you don't have a business.”
Courtney: Reverse-engineer images with ChatGPT for photorealistic Facebook ads
“If I just ask ChatGPT to reverse engineer the image and tell me exactly the prompt it would take to get that exact look and feel, and then use that when I'm making Facebook ads, it's completely different thing. It goes from like, you know, a person holding a p…”
Courtney: AJ&Smart requires staff to keep AI tools open full-time
“Well, you know, eight weeks from now, we will launch our first, like, we will have launched our first funnel, which is every team member at AJ and Smart having the ChatGPT tab, Claude tab, whatever else they're using, open the entire time as like an internal r…”
Isenberg: Use Claude or ChatGPT to write v0.dev prompts for you
“See, you don't even need to know how to prompt really well here, right? Like, there's so many guides around how do you get the most out of v-zero, for example, but it's like, just ask Claude or ask ChatGPT how to do the best prompting. Like, it will tell you h…”
Howell: Never Let ChatGPT Specify Tech Stacks for Replit Agent
“I think it just will cross too many wires if ChatGPT is like, use like Next.js and Node and this and that, because that's what Replit's good at, is kind of picking the stack that it knows how to do, so.”
Howell: Including OAuth in Specs Confuses AI Coding Agents
“If it, if ChatGPT tells you OAuth, like, that's probably going to confuse your AI agent. It's just more steps.”
API access will let marketers generate 100 ad variants in five minutes
“And when the API becomes generally available, this is going to be trivial. It's literally going to be like, This is your inspiration or like reference ad. These are all the different audiences, just like boom, boom, boom, smash them out and you'll have a hundr…”
Isenberg: Claude is better for writing, while ChatGPT excels at research
“Like Claude for me, writing is just, it's just a better at writing. Researching, you know, chat GPT, you know so there's these nuances that make these products better.”
Isenberg: ChatGPT Deep Research enables one-shot prompting better than Perplexity
“In this you're going to have to prompt a little bit more. This you're going to one shot it more often. ChatGPT you're going to one shot it more often.”
Isenberg: Perplexity's deep research completes queries faster than ChatGPT's
“Okay, so first things first is that perplexity is faster than ChatGPT. You can see that ChatGPT is about halfway done. Perplexity is done.”
Isenberg: ChatGPT Deep Research goes deeper than Perplexity
“It just goes a little deeper. It goes a little deeper than perplexity.”
Chu: Filtering raw scraped data directly with ChatGPT removes good listings
“I've noticed that it can get rid of a lot of the good quality listings if you
Just go straight into ChatGBT and have it remove what it doesn't think a dog park is, if that makes sense.”
Greg Isenberg: Creators should use Leonardo AI instead of ChatGPT for images
“Like, they do one prompt on ChatGPT, they get this, they're like, I can't do that, and they don't realize that their prompt is wrong, and that they're using the wrong product. They should be using Leonardo, not, not ChatGPT, Foro.”
Fernando: GPT-4o natively understands audio and tone, unlike DeepSeek R1
“So one thing I just learned very recently about GPT-IV and ChatGPT's Omni models is the fact that this model's breakthrough, a little bit different than R-I, is the fact that it can actually understand audio and tone and all these extra implications that we do…”
Fernando: DeepSeek via Fireworks costs $8/M tokens versus o1's $75
“I think it's about eight dollars per million tokens where normally I think ChatGPT was like 15 input and 60 dollars for output for O-one Pro.”
Cole: Writers should ideate first before refining with AI
“I think that you should always try yourself first and then weigh it against what ChatGPT comes up with and then improve the quality together. That is why I don't defer the writing to AI. I write with AI”
Cole: AI elevates writers to taste-driven editors-in-chief
“Instead of me passing this along to my junior columnist, I'm going to feed this to ChatGPT, which means we just got to elevate ourselves up to editor in chief, where our primary value is taste.”
Moura: Role framing in system prompts significantly changes LLM agent output
“Because these agents, they're kind of like impersonating roles. You do get different behaviors depending on, you do get different behaviors depending on how you give them like a different role. So you can actually replicate this on ChatGPT. Like if you ask Cha…”
Yadegari: AI apps with high latency should use background loading
“What a lot of apps do is they bring you to an analyzing screen where it loads, and you just wait on the screen. I honestly hate this, and especially for something that can take 30 to 60 seconds, depending on how complex your prompt is, and when you're waiting …”
ChatGPT cites AI-generated site Guilty Chef as top source for recipes
“And so now when you search like Burger King's spicy bean burger recipe, for example, you know, it'll give you the recipe, and it'll tell you that the number one source is, is Guilty Chef, you know.”
Wilkinson: I have completely replaced Google with Perplexity and ChatGPT
“I don't use Google anymore. Like for everything I go on perplexity or chat GPT now.”
Kim: Voice interaction with ChatGPT is a 10x improvement over text
“Like, you experience ChatGPT for the first time, and you're like, holy shit, this is cool. But then you talk to ChatGPT, and it's almost like this 10 X improvement in experience.”
Josephs: OpenAI sent Autopilot a cease and desist over ChatGPT portfolio
“OpenAI sent us a cease and desist, so we have to close that one down.”
Postma argues mainstream AI adoption requires skeuomorphic interfaces mimicking real-world workflows
“If you want to get products to be adopted, you need to make it feel like the real world, right? Basically how iPhone started with their interface notes and looked like notes. You need to make it feel right. So. These people, they're not going to use ChatGPT to…”
Isenberg: Specialized niche AI interfaces will thrive against generalists like ChatGPT
“Like there's certain things that chat GPT will not be able to be amazing at. So the interface for chatting with the PDF is, is specific to this use case. So I think more niche specific use cases, interfaces are really interesting.”
PDF AI founders made hundreds of thousands monthly before OpenAI's native update
“A few of the PDFAI guys who are building in that space. We're pretty pissed to put it lightly about what happened with the recent updates to chat GPT. Cause they were making tens, if not hundreds of thousands of dollars a month.”
Greathouse Fall: ChatGPT generates unique business ideas from identical HustleGPT prompts
“When you prompt the bot you know, with pretty much my original prompt verbatim, it's giving different ideas for everyone who approaches it, which I think is really interesting. It's not just spitting back the same thing over and over.”
Isenberg: I recovered $109,500 from a client using ChatGPT
“For the record there, you know, you're referring to how I recovered 109,000, 500 dollars from a client who wasn't responding to our emails. And I used ChatGPT to do it.”
Koe: Reverse-engineers high-performing content into LLM meta-prompts
“Anything that I like and want to incorporate into my work, I will ask Claude or ChatGPT to break it down, teach me how to do it. I will take that as if it were a, an SOP or just a set of instructions. Then I will take a prompt that helps me create prompts, sen…”
Perplexity Deep Research beats ChatGPT and Grok for audience research
“You don't necessarily need to use perplexity. You can use ChatGPT deep research. You can use Grok deep research. I just find that you know, this one perplexity deep research is a little better for this specific use case.”
Tossell: Codex Ask Feature Lacks Live Internet Access
“And it can do it and it, but it's more connected to, it's not connected to the internet. So it's like, I'm just looking at your repo and answering based on what it has, what knowledge it has to give you that response.”
Integrating marketing tools directly inside Claude eliminates context switching
“Like, you know, the challenge here for a lot of marketers using these tools is that they're switching between like, you know, 10 different things. Like their tabs open on their browser. Like, you know, they've got ChatGPT, they've got Claude, they've got maybe…”
Conley: Use Claude or ChatGPT to Audit Prompts Before AI Coding
“Something that I've been finding super helpful recently is going to a ChatGBT or Claude, whatever your favorite one is, and explaining your idea to it first, and then saying, does this sound right? Is there anything that I'm missing? And like talking to it alm…”
Pozel: ChatGPT Can Generate Ad-Account-Ready Creatives in Two Minutes
“We're going to teach people how to create ads ready for the ad account using ChatGPT. In about two minutes, it's that simple.”
GuiltyChef.com generates all recipe pages and images entirely via AI
“Every single page you're seeing here is completely generated automatically using AI down to the image. You know, and so some of them you can obviously tell it's an AI-generated image, some of them not so much but description, the prep time, cook time, ingredie…”