AI Everything Everywhere All at Once #WeblogPoMo2025
Photo by Mitchell Luo on Unsplash
This week I've swapped over to Bearblog.dev for hosting my personal site and blog. While poking around the discover page I stumbled on Ben Daubney's post How dare you (not) use AI. I really enjoyed reading Ben's perspective and since this is something I've been thinking about a lot lately I thought I'd go ahead and hop in for the month of May and blog some of my thoughts on AI/LLMs/Generative tools. This first post is just a springboard and in no way comprehensive on all my practical or ethical conclusions! I also reserve the right to completely change my mind. So let's dive in.Â
1. The Ultra Super Jumbo Elephant in the Room... Training Data
So this has been and will continue to be my moral/ethical sticking point. For most popular models, in one way or another, their training data is largely from whatever is publicly available on the internet. Copyright and intellectual property be damned. As someone who creates for a living this is obviously a pretty clear cut case of mass art theft. Especially for generative image tools like newest one in ChatGPT. There's not really much of an argument to be made to get around this in my opinion. So this dramatically effects how I use and will use image and video generation tools. Right now I am not comfortable using the images GPT puts out for anything other than playing around and brainstorming. Even platforms that claim to have more ethical training practices like RunwayML have been found to be scraping boatloads of copyrighted content from youtube and other sites.
2. Some of us don't have the luxury to opt out completely
Now that I've mentioned RunwayML I'd like to talk about those of us in the creative fields. This is where the waters get a bit muddier for me personally. I have used and continue to use tools like ChatGPT and RunwayML for work. Adobe Premiere's new generative extend feature is a lifesaver for us video editors and actually works really well in my testing so far. Often getting the best results out of these models requires you to feed them a "real" starting point. Whether that's an image or reference video. For those of us using creative software we've pretty much been using AI tools for years now they just weren't under the grotesque AI EVERYTHING EVERYWHERE push that software companies are shoving down our throats right now. I do think some of these tools are useful and potentially more ethically neutral when removed from all the current spin. Unfortunately finding that middle ground in the current climate is pretty difficult. Much of it comes down to responsible use.Â
3. It's actually pretty fun...
As someone who's always been into emerging tech (even though I often have a more reserved approach to adoption) some of the stuff LLMs and generative tools can pull off when used appropriately and with a trained hand are pretty fun and frankly pretty useful. Whether it's proof-reading, transcription, brain storming, organizing, or even just offloading some menial data-management tasks, AI can be fun and useful. It is not, however, the god tier tool to rule them all that AI hype bros are shoving down our throats. Like most tech, in my opinion, it definitely has its place. It's just going to take some time for us to figure out.Â
These are just a few points to get things rolling. I honestly find my thoughts and opinions on AI and the surrounding discussion changing daily, but I'll use this as a spring board to have this dialogue. I'd love to hear other people's thoughts and opinions so feel free to reach out via email at howdy@sethvirtually.com. I'll be grouping these posts under the tag #WeblogPoMo2025 to stay on theme with Ben's original post that got me going. Thanks for reading and I'll see y'all again here soon.Â