Compiling a prompting guide for #HustleGPT community. ✍️
Small changes in prompts can lead to big differences in output, especially if your using HustleGPT to write your venture's blogs.
Sneak preview below 🧵
Setting the tone is important in business negotiations and in prompting LLMs to give you juicy marketing content 🧃
1. Academic
If your building something backed by science, make sure it sounds like it
2. Informal
Trying to connect with your customers like their your buddy? Dial the tone the other way
3. Persuasive
Or maybe your product doesn't quiet sell itself and you need to turn up the charm. Simple words like "convincing" or "assertive" can make the output feel more powerful.
4. Humorous
Everything doesn't always have to be serious and #ChatGPT can say some things that'll make you chuckle.
ChatGPT: Why did the tomato turn red? Because it saw the salad dressing!
5. Serious
No segue. From Humorous to Serious. This is real life.
Stay tuned for the initial release of the guide!
These screenshots were from a single tip and by the end of the day there should be around 100 tips to get the most out of #HustleGPT 🚀
- Google Ad Campaign was suspended for god knows why
- Continuing to hit the response limit on GPT-4. I'm making it WORK 😂
- A lot of sites that distribute #NotionTemplates want giant affiliate fees
One thing you need to consider when designing your #Firestore database is how much data you'll need to build each document reference.
This will be important if you ever add notifications to your #flutter app since the notification should bring the user to a relevant screen.
For example, say you have the following collections:
Users -> Posts -> Comments
And you want to notify a user about a comment on their own post. In this case, the notification will only need the post ID and the comment ID since we can infer the user ID.
But what if you wanted to notify a user about a reply to their comment on someone else's post?
You'd need to send all 3 IDs so your app can resolve the destination.
Alternatively, you could flatten the database so that users, posts, and comments are top-level.