We all know the struggle of balancing a hectic work life with a bustling personal life, especially when you have three kids and countless other responsibilities. So, when a friend from work noticed I was posting on LinkedIn during my vacation, they couldn’t help but ask if I was ruining my time off by staying connected. The truth is, I’ve developed a ridiculously complicated posting system that not only keeps me active online but also helps me learn and have fun in the process.
This ingenious system starts with an application I wrote that reads blog posts from my computer and enhances them using GPT, as English is my third language. Although this occasionally results in a slightly odd “voice,” it generally works wonders. To add a visual touch, I employ Stable Diffusion to generate images based on prompts I include at the top of my posts.
Once the content is polished and ready, it’s saved in Azure Cosmos DB, along with metadata about when I want it to be posted. I wrote an Azure Functions app that triggers from time to time and checks Cosmos DB for any posts that are due. If there are any, the LinkedIn API then takes over, posting the blog posts and images according to the schedule I’ve set.
Now, you might be wondering if I’ll eventually use Reinforcement Learning and Semantic Kernel to completely automate what I do in social media. While I’m considering doing some of this, I’d likely only use this approach for a small percentage of posts, just to test the waters.
In the end, unless you’re genuinely passionate about learning and experimenting, automating your social media posting like this might not be worth the effort. But for me, it’s an exciting way to dive deeper into the world of AI while staying connected with my professional network, even during my well-deserved vacations.