I'll start by saying that today was more of a rest day, the family was home, there was a show we love to watch and then a stand-up comedy show I went to. So I only made my real-time transcription app configurable which took no time at all and also made a settings pane so that users could adjust the config from the client.
I'm not releasing yet, I've got more things I want to consider before I do so. Not in a rush :)
Funnily enough, I just did my daily reading of Atomic Habits time, and the author says the following line: "Amateurs let life get in the way, Professionals stay on schedule"
I feel like this applies to me, as I sometimes can neglect one thing for another. I feel like I need to show up more even when I don't feel like it and work through nothing-but-focus pomodoro sessions even when my brain doesn't feel like it.
That's something I'm gonna take into account when I make decisions :)
Some other questions have been rising in my mind:
LLM's are one thing in the field of AI, why is AI getting so much recognition just now? It's not like things like sheep counting or fraud detection with AI is new (to my knowledge), they're built upon concepts that were available long before the release of such LLM's using ML
How do you teach a model what to do when you have no labeled data? Unsupervised learning is really interesting me now. I did start to learn all these things at the beginning but I swittched over to practical projects for the moment with the available tools like ChatGPT.
Don't even get me started on neural networks, they tackle my brain with how they're different and that's what makes me wanna learn this.
I just keep asking myself, if I wanna make cool AI projects, doesn't it mean I need to generate data and train a model on it? This is what it feels like to me based on what I've seen during the development of this project, especially in speaker embedding extraction, which super interests me in terms of what the model helps with and how the data it's trained on helps it.
I think this arises from my lack of knowledge about how AI really works and how practical applications look.
So this is something I really want to learn as I work on projects! I think it'd be fun and interesting.
However, things like Andrew's course on Coursera go really in-depth, I don't know if that's what I need, but I'm not looking for any shortcuts :)
It's the constant recommendation by ChatGPT, probably means that detail is worth learning.
Theory, practical applications, projects.
That's my goal with this, but I still need to think this through and not act impulsively.
That's it for today,
Happy coding everyone!
Top comments (0)