AGI models (aka artificial general intelligence) have swiftly taken over the world in the last few weeks and have become a subject of love and fear, and most of the time, a combination of both.
It has also shaken mobile marketers. ♥️ Undeniable love : “Will I ever have to write long and tedious articles again?” and 😱 Fear: “Has ChatGPT replaced me all together at last?”.
Mobile marketers, fear not. You have not found your new fountain of youth nor will you now have to change your career path - but here is how AGI can be a great benefit to you 🚀
AGI (Artificial General Intelligence) models are a class of AI models that are designed to exhibit human-like intelligence and perform a wide range of tasks across different domains. Basically, they’re trying to copy our brains 🧠 - hard.
They use deep learning techniques to analyse large amounts of data and learn complex patterns and relationships in the data. And that’s when it gets interesting 🧐
What your brain can’t do (as well as)
AGI (Artificial General Intelligence) models can provide several benefits by analyzing vast amounts of data from user interactions with mobile apps, websites, and other touchpoints. That could include:
- Personalized marketing: AGI models can analyze user behavior and preferences to provide personalized recommendations and offers, which can increase engagement and conversion rates.
- Customer segmentation: AGI models can segment users based on their behavior, demographics, and other attributes, allowing marketers to tailor their messaging and promotions to specific customer groups.
- Predictive analytics: AGI models can predict future user behavior and purchase intent
- Real-time optimization: AGI models can analyze user behavior in real-time, allowing marketers to make quick and data-driven decisions
- Chatbots and virtual assistants powered by AI
You can be a great marketer but ingesting such large sum of data is close to impossible to a human brain 🤯 - AGI models will do the dirty work for you.
But be sure, that’s probably all it will do correctly.
A few ways you can still really benefit from it:
- Know your prompts
The best way to make use of AGI models is to make sure your prompts are on point. ☝🏾
A prompt is the text or message that a user inputs into the chatbot, which the bot then uses to generate a response. It is essential for using ChatGPT effectively - it can differ with Dall-E.
Here is a an exemple of 2 prompts for a simple mobile marketing enquiry:
Prompt: You are George Clooney promoting a gaming app on Instagram
Hey everyone, it's George Clooney here and I've got something fun to share with you all! If you're looking for a way to unwind and have some fun, check out this amazing gaming app! With tons of games to choose from, you're sure to find something you love. Trust me, this app is a game-changer! Download it now and let the fun begin! #GamingApp #FunTimes #GameOn #GeorgeClooney
Prompt: You are George Clooney promoting a gaming app on Instagram with elegance and wit
Greetings, my fellow Instagrammers. As you know, I'm not just an actor, director, and producer, but also a bit of a gamer. And I've recently discovered this incredible gaming app that I simply have to share with you all! It's elegant, it's witty, and it's downright addictive. So if you're in the mood for some serious fun, download this app now and let the games begin! Trust me, you won't regret it. #GamingApp #EleganceAndWit #GameOn #GeorgeClooney
Framing your prompt will make it easier for ChatGPT to give you a precise answer. Don’t be afraid to add those extra details that can make all the difference.
Finding really good prompts can also be time consuming. Which defies the purpose of using AGI. So example of easy tool you can use to find the best prompts, a Google Chrome extension called AIPRM. It will feed you with great prompts covering many different topics.
- Let it run your A/B testing
A/B testing is obviously crucial to any marketer. And AI models can help mobile marketers conduct A/B testing by analyzing data collected from the two different versions of the campaign and providing insights into which version performed better. Here is how:
- Predictive Analytics: it can use historical data to predict which version of the campaign will perform better.
- Real-time data analysis: it can monitor the performance of each version of the campaign in real-time, allowing you to quickly identify which version is performing better and adjust your strategy accordingly.
- Personalization: AI models can analyze data on individual user preferences and behaviors to deliver personalized content to each user. This can help you tailor your campaigns to the needs and interests of your target audience and increase engagement.
- Optimization: AI models can automatically optimize campaigns based on performance data, adjusting various elements of the campaign such as copy, images, and targeting to improve results.
- Use it for coding
It was a revolution in the world of coding - and probably feels like it doesn’t apply to you but hang on for a minute - it might get interesting.
You’re looking to change an ad you have previously used thanks to code?
Given the right prompt, you can modify or create code to mimic gameplay, or modify playable elements.
Chat GPT can help you generate code for specific tasks by understanding your requirements and providing code snippets. You can input your desired functionality, and Chat GPT will generate code that implements it.
Example: Your ad in gif format displays pink shoes and you would like to change it to red. You feed ChatGPT with the code of your gif and ask the color to be changed to red in the code.
Great asset for developers, can also be highly valuable for marketers 🏆
Who’s doing it?
- Amazon: Amazon uses AGI models to personalize its mobile app experience for each user. By analyzing the user's browsing and purchase history, the app can make personalized recommendations and offer customized deals and promotions, improving the likelihood of a purchase.
- Spotify: Spotify uses AGI models to create personalized playlists and recommendations for each user.
- Starbucks: Starbucks uses AGI models in its mobile app to create personalized offers and promotions for its users.
- Netflix: Netflix uses AGI models to personalize its mobile app experience for each user. By analyzing the user's viewing history and preferences, the app can make personalized recommendations and offer customized content.
The limit of the model: the example of Dall-E2
Here is our script for Dall-E, a new AI system that can create realistic images and art from a description in natural language:
“Two people having coffee in a garden”.
Yes, I know what you’re thinking. Why are the faces and the coffee cups melting? 🤷♀️
Great metaphor for AGI models - it will do most of the work, but we still need our human beauty cover touch on things 💄
In conclusion, what’s the future of AGI and mobile marketing?
It’s not a replacement. It’s a tool. Your beautiful human brain still has it going strong.
There will always be a need for designers and writers for content but AGI tools will cut down on time and allow for more frequent and/or higher quality production.
Find the sweet spot and you’ve got it all figured out 😉
PS: This article was written at 30% by ChatGPT ♥️ (ok, maybe 50% - who will ever know?)