In short
Podcast Episode Summary: Ask Us ANYTHING: 3 NEW ChatGPT Tricks! + 🫨Should Jay Move to Japan?🫨 | Ep. 415
Podcast Overview Podcast Title: Do This, NOT That: Marketing Tips with Jay Schwedelson Description: The podcast offers marketing insights and practical advice aimed at enhancing marketing strategies. It explores decision-making, audience understanding, and marketing psychology.
Episode Highlights
Introduction
- Jay Schwedelson welcomes listeners to a new format of the "Ask Us Anything" episode.
- The focus is on answering audience questions, both serious and absurd.
Main Discussion Points
- AI Insights and Misconceptions
- Stats from AI:
- Up to 50% of statistics generated by AI can be misleading or fabricated.
- Jay shares personal experiences where he was misled by AI-generated stats.
- Best Practices for AI Prompting:
- Verified Responses: Always request "verified answers that are 100% not inferred" to avoid inaccuracies.
- Prompting Techniques: Instead of crafting prompts, ask AI to create the best prompt for your desired outcome.
- Timeliness of Information: Specify time periods in prompts to ensure receiving up-to-date information. Without a time frame, data may be outdated by over a year.
- Audience Questions
- Work-related Question: Jessica from Denver asks about Jay's second favorite thing about AI.
- Ridiculous Question: Shay from Guam questions whether Jay would prefer living in Guam or San Francisco.
- Jay humorously expresses a lack of knowledge about Guam and shares his preferences for living in Japan due to cultural aspects such as privacy, advanced vending machines, and manageable portion sizes in meals.
Fun Moments
- Jay humorously critiques living in places with annoying associations (e.g., Scranton, PA due to "The Office").
- Positive portrayal of Japan includes:
- Vending machines that offer diverse products (not just snacks).
- Advanced toilet technology enhancing comfort.
- Cultural preference for moderate portion sizes.
Key Takeaways
- AI Limitations: Users must remain critical of AI outputs and verify statistics to avoid spreading misinformation.
- Effective Prompting: Learning to prompt AI effectively can significantly improve the quality of responses received.
- Cultural Preferences: Personal anecdotes offer insights into lifestyle preferences, highlighting the charm of different locales based on their unique offerings.
Conclusion
- Jay encourages listeners to engage further by asking questions and subscribing to the podcast.
- Reminder about the upcoming Guru Conference, featuring notable speakers and free virtual seats.
Additional Resources
- Follow Jay on:
- [YouTube](https://www.youtube.com/@schwedelson)
- [TikTok](https://www.tiktok.com/@schwedelson)
- [Instagram](https://www.instagram.com/jayschwedelson/)
- Register for the [Guru Conference](http://guruconference.com).
---
This episode combines practical marketing advice with lighthearted discussions, showcasing Jay's engaging style and expertise while demystifying the use of AI in marketing.
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Transcript
Automatic transcript. May contain errors.0:00Welcome to Do This Not That, the podcast for marketers. We share quick tips, things you can do right now, and then we add a little bit of chaos at the end of every episode. We also keep it short, like this intro. Let's check it out. We are back for Ask Us Anything from the Do This Not That podcast presented by Marigold. You know, normally we release this Ask Us Anything on Wednesdays, but we're trying it out on Fridays. We're mixing it up. It's always about testing to see what people like and when. And so thanks for sticking with us. And if you've never heard this episode, what we do here is we get questions in all week long.
0:35We get in work questions. We get in absolutely ridiculous questions. And we try to tackle one of each. And if you want to submit a question, you can go to jschwedelson.com. There's a button that says podcast. Another one that says ask us anything. We love the questions. And we got some audio questions in this week. So we're going to try to tackle those. So let's go to the work question before the ridiculous question. Here we go. Jessica from Denver, what's your second favorite thing about AI? All right, Jessica from Denver. That's an interesting question. What is my second favorite thing about AI?
1:11Well, I'm going to tell you right now what are three things, so that way the second one fits in there somewhere. What are three things that you can do on AI that you should be doing on AI that maybe you're not, like some quick tip trick things that are actually kind of a game changer? The first one is a little bit frustrating. You may or may not realize this. Everybody knows that when you go to ChatGPT or whatever, they'll give you hallucinations. They'll give you answers that aren't necessarily real. Even if they give you a stat, people don't realize when you get a stat back from, let's say, ChatGPT, it says 34 % of people, whatever industry you're in, it sounds very official.
1:50Sometimes it even says a source. You know, 34 % of people from Acme company, according to whatever study, blah, blah, blah. and you're like, wow, that sounds real. And then you put it in the slide deck, you present it, and guess what? It wasn't real. And how do I know that? Because I've gotten burned in presentations before. And so here's the wild stat, and I'll tell you what you got to do. This is real. And this is actually ironically verified, according to ChatGPT itself, that up to 50%, on average, over 30%, so between 30 % and 50 % of stats, Plus, newest items and niche details that are provided back on Chachapiti are inferred and they're not necessarily accurate or real.
2:33So what does that mean? That means basically, on average, about half the stuff that you get back from Chachapiti is just made up. It's just completely made up. So that's annoying. So what do you do about that? When you're looking for any real information on Chachapiti, this is all you got to do and this is what you should do every single time. You say, oh, what's the latest stat on whatever, or what's the average whatever? And then you say in that same prompt, only give me verified answers that are 100 % not inferred and provide exact source. The key thing there are verified answers and that it's not inferred.
3:13Because if you don't tell it to stop inferring, it does infer. It infers by default. So you want to be doing this on everything that you ask ChatGPT that you're asking it to not infer. Okay. Now, the other quick tip is, the second quick tip is the fact that I have no idea how to prompt, right? I'm not the greatest prompt writer. You know, you go to the bar and you write whatever you want to write and you keep messing around with, is this one good? How do I really get it to be so good? Who knows, right? There's a simple way around that. Okay. All you have to do, instead of trying to come up with the right prompt, what you want to do is you want to ask ChatGPT what the prompt is.
3:55What does that mean? So whatever it is that you want to do, I want to make muffins. I want to get my email open rate higher. I want to make a business plan for my plumbing business. Whatever you want to do, you go to ChatGPT and you say, I want to do, and then you put it in there. I want to do this. Write me the perfect ChatGPT prompt for this with exact wording, format, and style. And then it will give you the prompt for you then to use to get the result, information, instructions, thing that you want. Stop trying to create the best prompt and ask ChatGPT, what is the best prompt? That's it. All right, the last one, and this one's kind of like not used a lot, and I think it's an epic fail.
4:45One of the big benefits, and everything's on the free version of ChatGPT or any of the other free versions, any of the tools, but I like ChatGPT, right? One of the benefits of ChatGPT now, ChatGPT 5, is that it can go out and get real-time information, okay? And the reason that you need to make sure that you're asking for kind of this real-time information is, and this is per ChatGPT, this is confirmed. If you don't put in a time period with whatever it is you're asking ChatGPT, okay, if you say, I want the latest metrics in the healthcare industry about social media posts on LinkedIn after 4 p.m.
5:21and make sure it's all happening within the last 12 months or whatever. If you don't put a time period in your prompt, what happens is it's giving you old data. So ChatGPT is, this is based on themselves. They say, on average, if you don't include a time period in your prompt, the data you get back, on average, especially in the marketing category, is at least 12 months old. So anytime you're getting the stats or information or tactics or whatever back, if you don't ask ChatGPT for a time period in your prompt, this stuff is out of date. It is old. So when you do go and you ask ChatGPT something, like, so, for example, I'm such a loser.
6:01I'm not kidding. I asked ChatGPT about pop culture stuff. So I go on ChatGPT and I say this, show me the top five pop culture trends in the last seven days that people are talking about. And that last seven days is key. You could say, show me the direct-to-consumer marketing tactics that are working the best in the last 60 days. Show me in the nonprofit sector, what are the call-to-action buttons that are performing the best in the last six months, you must give it a time period now. If you don't, you're just getting back old information. It's a simple trick that enough people are doing. All right, let's get to the ridiculous portion of the podcast.
6:47We have another audio question coming in. Let's hear what it is. Hi, Jay. This is Shay from Guam. Would you rather leave in Guam or would you rather stay here in San Francisco. All right. Shay from Guam. Would I rather live in Guam or San Francisco? Okay. Well, I've done no research on this, so I've never been to Guam. This is so rude, but if you forced me and said, where is Guam on a map and you took off all the labels, I would give it a 0 % chance I could find Guam. I miss you though. You're a wonderful person. I want to come visit you in Guam. I just don't know where it is. San Francisco, I don't want to live in San Francisco.
7:28No, I'm nothing against San Francisco. I'll tell you some places I would and would not want to live. So not that you asked. So I'll tell you where I don't want to live. I don't want to live any place that is known for something annoying. So for example, like I don't want to live in Scranton, Pennsylvania at all. Why? Because it's known as the setting for the show, The Office. That to me is annoying because then anytime people visit or they ask you where you're from and you say I'm from Scranton, Pennsylvania. They're like, oh, are you friends with Jim or Pam from The Office? Ha, ha, ha, ha, ha.
8:01It's like, oh, do you have a Dunder Mifflin shirt? It's annoying. So I don't want to do that because my entire life I don't want to answer questions about like or if you're from like Roswell, New Mexico, because that's like where all the alien stuff is, like the alien sightings or whatever. How annoying must it be to live in Roswell, New Mexico? If you go to like the diner to go eat, I guarantee you on the menu of every restaurant is like alien shaped pancakes. I can't. I can't handle like an entire where I live, everything to be about the thing that I don't even care about. Like who cares? If there are aliens, great.
8:37Come hang out with me. I need new friends. Where would I want to live? I would want to live in Japan. This is not a flex. I've been to Japan a couple times. It is the greatest place on earth. Why? I'll give you a couple of reasons why. Number one, nobody talks to each other. Amazing. So like when you go on the train, you're not going to deal with somebody trying to talk to you. They don't do it. They will not talk to you. Amazing. I channel that wherever I go. I love that. First of all, they also have vending machines. Like in the United States, we have vending machines like, oh, you want to get a soda?
9:10Okay, great. They have vending machines there where you can get like ramen soup or a tie if you're missing a tie or underwear. I mean, their vending machines are wild. And the toilets in Japan are incredible. I mean, I don't know if I've ever broke down toilets on this show, but they have all these buttons. You could warm your toilet seat. It sprays out all this water stuff. It plays music if you want it to. Oh, there's so many wonderful things there. And I wouldn't be a slob because when you go and you order food at a restaurant there, the portion sizes are like for a human. They're not for a cow.
9:48Okay. So you get the amount of food. If you want to eat it all, great. You don't wind up looking like Jabba the Hutt, right? It's amazing because they don't give you so much food. It's awesome. What am I talking about? I don't know. Listen, you're awesome. Let's all move to Japan. A few things. Number one, if you're not registered for Guru Conference, you stink. And I don't like you because we are working so hard on this thing. All right, we have Nicole Kidman. We have Lance Bass. We have a major announcement. We even told you another celebrity is coming. We are almost out of these virtual seats.
10:23I keep saying that, but nobody believes me, but it's true because we only buy a certain amount. And they're free. This thing is November 6th and 7th. You go to guruconference.com. What are you doing? Register. And if you leave this thing a review or you leave a comment on Spotify, I'll answer it. We need the reviews and whatever because it circulates the show and we can share more nonsense. And the world needs more nonsense. You're very awesome. I appreciate you and keep it real. You did it. You made it to the end, but wait, the party is not over. Listen, I want to keep hanging out. Subscribe to this podcast.
10:55And if it wasn't the worst podcast you've ever listened to, give it a five-star review. Why not? But you know what? I want to do even more with you. Go to gurumediahub.com and we can partner there. You could find out about all of our free events, all of our stuff. And if you're epically bored, go to jschwedelson.com and we could stay connected. You could find my newsletter and everything else I got going on. Thanks for being here. and hope you subscribe.
From the publisher
Sometimes the best insights come from the strangest questions, and this Ask Us Anything proves it. Jay Schwedelson riffs on AI tricks you probably aren’t using but should, why half your “official-sounding” stats might be totally fake, and the one detail that makes your prompts 10x better. And just when it seems like things can’t get weirder, a question about Guam leads to toilets, vending machines, and why Japan might secretly be the best place on Earth.
ㅤ
Best Moments:
(01:04) Why half the stats you get from AI might be made up
(03:00) The simple phrase that forces AI to stop inferring fake numbers
(03:45) How to get ChatGPT to write the perfect prompt for you
(05:13) The one trick that keeps AI responses fresh and not a year out of date
(07:01) Guam vs San Francisco and the real problem with Scranton, PA
(08:45) Why Japanese vending machines, toilets, and portion sizes make life better
ㅤ
Check out Jay’s YOUTUBE Channel: https://www.youtube.com/@schwedelson
Check out Jay’s TIKTOK: https://www.tiktok.com/@schwedelson
Check Out Jay's INSTAGRAM: https://www.instagram.com/jayschwedelson/




