Lucy LoveOne of the latest interesting discussions on Slack at work has been about the value proposition of declare-ai.org, a website created (without any assistance from a Generative AI) by David Bryant Copeland that recognises the excitement around AI and wants to help people inform and be informed about its uses.
(bit off topic but, I am perplexed by the fact that David's own website doesn't have any comment about AI Content Declaration; perhaps they felt they had said all about the topic in the main website. Or perhaps they just wanted to make sure it is clear to people that their work has nothing to do with any sort of AI, maybe out of frustration for the new wave of low-quality/low-effort content blatantly created with AI and not wanting their work even remotely put on the same plate)🤷.
While I recommend checking the website (as I couldn't write any better about it than you could deduct from checking the actual site), the process it proposes is as follows:
Flow proposed by AI Content Declaration on how to determine what type of Generative AI attribution you need for your content. Diagram generated with Venngage.
Not sure if the diagram can be read in an accessible way, therefore let me repeat it here (I will use a structure that proposes first the condition that end the flow). We start asking ourselves Was Generative AI used in any way?; if the answer is no, then the answer is None (as in no contribution from generative AI). If the answer is yes, we are prompted to another question Did Generative AI produce the entire work?; answering yes will land us on Total, as in total AI contribution; if we answer no, we are asked the last question Did Generative AI contribute creatively?; yes brings us to Creative Assistance and no will bring us to Non-creative Assistance.
It is up to your own judgement to determine if you used the AI in a creative way, but the website offers you a nice way to determine what could be considered creative assistance and what not.
And once you go through this hierarchy, you can make a public declaration about it, in the best way that suits you and with the examples proposed. It allegedly also proposes to add this information in the form of meta tags... because why not?!
Now now, will I use it? Not yet! I would like to see it having more popularity; but what I like to do is to be open about the source of my work! And that's why I gladly mention at least the most important prompts I ask about to ChatGPT at the end of each article. My go-to way to use ChatGPT is to get inspiration on how to start discussing about a topic, steal good-sounding adjectives (as my English can only benefit from expanding my vocabulary), and in general to get over a writer's block that often prevents me from sharing more, and learn more about something.
More than often, I read what ChatGPT proposes to me, and I'm immediately like nope, definitely not that! as it often goes with a very bland structure, lots of repetition, and unneeded conclusion paragraphs that give no particular value to the topic asked. But more than once, when I want to make sure I understand something about a particular topic, it performs as a great teacher that never gets tired, no matter how stupider my questions become. And it's really a teacher that I'd wish for everyone to have.
Many feel ashamed to mention they'd use any sort of generative AI and they'd prefer to keep it hidden. Instead of having a sense of appreciation for this technology, a lot of people tend to judge negatively the skills of people using AI, assuming that they're lazy, not knowledgeable, not actual professionals and that they'd not be able to succeed without such tool. I personally find unfair to question people's skills' worth based on how much they like to use AI, as I don't think my critical thinking is compromised any more than it'd be if I were to acquire information from a YouTube video, an online article, a book and so on. With that premise, I use ChatGPT as one among many sources to acquire knowledge and understanding of a topic, and it never is for me the only source of information.
Rather than hide its usage, I absolutely love the idea to share with people how it helps me go past a lot of hurdles. Before ChatGPT, I also have never been particularly shy about sharing where I've learned something, or what sources I have browsed to learn and/or do something. I really love to share with people because I would not be the person that I am today if people didn't share the things I've digested to this day.
We seem to forget that not too long ago, we were all assuming developers were just copying stuff from StackOverflow and asking questions over there. So how does that exactly differ, in terms of originality and competence, from what ChatGPT offers us these days?
Have you ever wondered how many people make use of ChatGPT for writing their blog posts? What's your take on the topic? Should people be more clear to which extend they use AI or would you just be happy to have valuable content in front of your eyes, regardless of how it came to be? Let me hear your opinion in the comments!
Sources and inspiration
- Declare-AI official website
- The aforementioned discussion on Slack at work
- Cover: Title and background gradient tweaked out with the DOM Inspector from Declare-AI official website
Originally posted in oh-no.ooo (declare-ai.org - ... so, who wrote that?!), my personal website.
