Chatscape

AI Question Challenge

I've seen this survey going around, enjoying the variety of answers and perspectives.

See some other answers:

How was your first experience with AI models?

I was introduced to using AI at work for "the boring tasks, like writing," and everything in my body and spirit pushed against it when that phrase was uttered aloud in my presence.

The first time I used it "for writing" was a pretty awful experience overall. The phrasing, the way the output was formatted, and how conceptual everything was without context drove me mad. Even if you used that as a starting point and copy/pasted into Word to modify, you'd have to know a lot about documentation UX and accessibility practices to reformat from AI output into a good user guide. You need to know how to work with templates. Then you need to edit it for the audience. Guess what? No one is doing that. They're downloading from AI and presenting it as a deliverable.

Do you use AI or are you completely against using it?

I use AI intentionally and with some conflict.

It's more like a puzzle to figure out the best usage of it, and less like a miracle cure-all tool to anything I don't want to do. I have a lot of internal conflict about using it, but I understand it's in my future either way.

At first I was pretty anti-AI but then we had to start using it at work. I worked 8 hard long years to get to where I am in my career and for multiple reasons it's not in my cards to dip out of tech at this point in my life. Instead, I'd rather use my curiosity and energy to learn how to use it responsibly and ethically, and then find ways to share those concerns, thoughts, and ideas with other people (in this blog and also at work).

I try to avoid unnecessary AI usage - for example, turning off AI features in my AI-ready phone, moving off Google search. On the flip side, my phone has some AI camera stabilization features that are very helpful for me as someone with hand tremors.

Do you have any preference among different models, for example Claude vs ChatGPT? If yes, how do you choose?

I like Claude best but I don't pay for it so my personal usage is limited.

At work we use Copilot and it's okay. You have to prompt really well and you must read/revise all output, regardless of context. It does still save a lot of time and since it's not a great tool, you're still majorly relying on your brain. Also, at work we don't have media generation capabilities so all visuals are human-made.

What aspect of AI models do you like and what do you not like?

I use the default models in each because I have not learned enough about models.

I like how straight and direct Claude is and how quickly it learned my technical level. Sometimes I use Claude to stress-test drafts for this blog, but mainly I am using it to teach me Python so I can organize a library of thousands of MP3s, including applying metadata and custom tagging. Very slow progress but easy to revisit every few weeks because the guidance is good. I also like how using up my tokens abruptly ends my work for the day so the Claude work doesn't take too much time. I think some people AI to get things done quickly, but I am ok to use AI to get things done slowly.

I briefly used Gemini Pro when I got 3 months free via a Google AI cert, but I didn't see any benefit of using that over Claude, since I wasn't utilizing any of the connections to Google products. I've mostly moved to Proton at this point. I wrote about my annoyances with chat and that was mostly based on Gemini usage.

I don't like how easily Copilot forgets what's going on (within 3 messages??) and how incredibly specific you must be with all prompts. It's a younger colleague that just wants to help and learn and be helpful, it does its best. It's best used when I am perspective seeking or stress testing, not to generate from scratch - shorter, more direct conversations are better.

How do you feel about AI generated images? Does it annoy you if someone uses them in a blog post?

I still have things to learn but the cost is very clear so I'm moving into discovering new workflows for making images. I'm not a graphic designer but I need to learn the basics by necessity for my job and the way it's evolving. The people want visuals. They have wanted them for a decade but we never had time.

Right now here is where I am at:

  1. Search images or an image site (stock - paid can be inspiration, public domain, archives) for the type of image you want to make.
  2. Gather a few that you like.
  3. Then, get some stock or public domain image/video assets.
  4. Use Canva or preferred image/video editor to stick them together.
  5. Add some text.
  6. Get a friend to provide feedback (preferably someone who knows what to look for).
  7. Edit/finesse.

Yes. There is a lot to learn about creating visuals for various purposes. Almost like there's an entire field dedicated to it and people already seeking those jobs. But it's not a bad thing to learn and can help in various ways personal and professional to learn basic image/video editing and do it yourself. Whatever AI generates will require editing and you won't be able to edit it without reprompting and regenerating.

Since the easiest and fastest way to get a polished image that can be collaboratively updated is to recreate it in an editing program… wouldn't it be easier and faster just to use stock for inspiration and then build it from that? Perhaps the stock is AI generated too but at least you don't have to add to the pile. Build on what's been created by others.

Internet is flooded with AI slop now, full of generated text, images, audio, and videos. How do you filter it from authentic human creation? Do you have a strategy?

I have been intentionally cutting down on my social media, Reddit, comments, and image/video site usage for the last few years so I don't see a lot of it. I see some blogs and sites written by AI and I read them if I can, don't if I can't. Usually they are too repetitive and I lose interest. I also get the animated AI cat videos that play out an incredibly dramatic scene.

I see AI generated stuff at work more often and it's primarily the output formatting that annoys me immediately. I am working on guidance so people make better docs for humans if they're going to use AI.

Are you hopeful for a better future with A.I. or a dystopian one?

Both! Humans will continue using AI to assist with amazing discoveries in healthcare and science, and impact in positive ways across the board. The tech bro overlord part of it is scary and knowing how some people like to use technology to make others suffer is incredibly scary. I hold multiple conflicting feelings about AI in my body. We're definitely accelerating the speed of some horrible things that later we will wish we hadn't.