Cookie Consent by Free Privacy Policy Generator

The Best Of

Go to the Best Of the SEO Community.

Noah
Noah
Jan 19, 2025, 10:00 AM
Forwarded from another channel:
Forwarded thread from another channel:
Emina Demiri-Watson
Emina Demiri-Watson
Dec 18, 2024, 8:13 AM
Hiya,
I'm doing a talk at WTS Fest in Feb and wanted to gather some thoughts from the hive to include in my ppt. Would you be so kind to answer the following questions - nothing too long, just quick bits
1. How would you define ethical AI?
2. Are you actively thinking about ethics when using LLMs?
3. Are you doing anything on that front in your company?
thanks a ton!
MIchael Buckbee
MIchael Buckbee
Dec 18, 2024, 8:20 AM
Are you asking about ethical AI use or how AI's perceive brands/companies acting ethically?
Emina Demiri-Watson
Emina Demiri-Watson
Dec 18, 2024, 8:58 AM
It's ethical use of AI - ie. dealing with some of the consequences such as environmental impact
MIchael Buckbee
MIchael Buckbee
Dec 18, 2024, 9:10 AM
I think that the direct impacts we've seen so far (like environmental impact) are somewhat overstated and the potential societal impact is understated.
Emina Demiri-Watson
Emina Demiri-Watson
Dec 18, 2024, 9:13 AM
I tend to agree with that actually. It's not the only thing I've started looking at but it's the one that is getting researched the most at the moment and easiest to action from a user perceptive.
MIchael Buckbee
MIchael Buckbee
Dec 18, 2024, 9:13 AM
There's sort of a blend as well with "appropriate" vs "inappropriate" uses. For example, because they reflect data sourced by flawed humans, using them falsely neutral judges of things feels _very_ unethical. I'm thinking of things like using AIs to filter job applicants, etc.
MIchael Buckbee
MIchael Buckbee
Dec 18, 2024, 9:14 AM
Is the action "don't use AIs so much?"
Emina Demiri-Watson
Emina Demiri-Watson
Dec 18, 2024, 9:18 AM
There are a lot of actions - one of them is making sure that you are aware of the inherent bias. I mean its all use of LLMs for SEO specifically. I have a few things we are already looking into but in general the ppt is about the journey - asking questions and not just going for it without any considerations. With some examples where we are in this journey - ie. some simple workflows we built in Make and the type of things we had in mind while doing it, some google colab bits we have been using and again some questions Ive been thinking about....
MIchael Buckbee
MIchael Buckbee
Dec 18, 2024, 9:20 AM
So, I've been coming at this from the other side - where for SEO I've been trying to get companies to think about how their ethical actions impact their search results in AI answers as to date the LLMs have a much broader definition of things like "best".
MIchael Buckbee
MIchael Buckbee
Dec 18, 2024, 9:21 AM
Here's the list of questions that we're encouraging folks to ask the LLMs about their brand.
MIchael Buckbee
MIchael Buckbee
Dec 18, 2024, 9:21 AM
I don't know if this then crosses back over to "Is this an ethical way to use AIs?" or not.
MIchael Buckbee
MIchael Buckbee
Dec 18, 2024, 9:22 AM
Like I don't think it's quite right to ask `Is Emina Demiri-Watson ethical?`
MIchael Buckbee
MIchael Buckbee
Dec 18, 2024, 9:26 AM
Maybe marrying these two ideas up: how long is it ethical for the LLMs to hold onto viewpoints or opinions about a company? Here's the example I always point to:
MIchael Buckbee
MIchael Buckbee
Dec 18, 2024, 9:26 AM
But that was _years_ ago now - should ChatGPT still be letting those actions influence how it today?
Emina Demiri-Watson
Emina Demiri-Watson
Dec 18, 2024, 9:28 AM
Nice one. Thanks. I’ll see if and how it fits!
MIchael Buckbee
MIchael Buckbee
Dec 18, 2024, 9:28 AM
Good luck on your talk! Super fascinating subject.
Franzi Dietrich
Franzi Dietrich
Dec 18, 2024, 9:50 AM
another take - first thing that popped into my mind was protecting intelectual property and copyrights especially on the content creation/video/image side cause yes, I always have that in mind and therefore would never take something ready made from an AI (even if it would be googd quality already^^)
RP
RP
Dec 18, 2024, 1:28 PM
Ethical as : rewrite this (someone else's content) and make it unique.
Creating fakes (images/video) and not declaring them as such.
Copying and pasting onto a client's website with human touch to save time and increase profit.
We do not have a company policy regarding AI/LLM, maybe that is something companies should start adopting/implementing and publishing to build trust.
Mikael do Nascimento Araújo
Mikael do Nascimento Araújo
Dec 18, 2024, 3:36 PM
*How would you define ethical AI?*
> It's about being transparent about the use of AI-related resources (data, text or media processing) in internal processes and their impact on the organization's audience or customers
*Are you actively thinking about ethics when using LLMs?*
> Yes. There are several biases caused by the way LLM models have been created. As much of the source material for the creation of the models came from languages in so-called developed countries, a number of cultural, gender and racial biases are built into the models.
>
> We've observed this a lot in various tests we've carried out to translate articles into other languages (language and cultural biases reflected in the articles' arguments), as well as gender and racial biases in the creation of images.
*Are you doing anything on that front in your company?*
> Unfortunately not. Managers still don't have a critical eye for the question. In fact, what I see is that there is still widespread unfamiliarity with these impacts. Only the professionals who are at the forefront (carrying out tests, creating hypotheses, etc.) are beginning to realize the fundamental impacts (both positive and negative) of AI models.
Ryan Mendenhall
Ryan Mendenhall
Dec 19, 2024, 9:33 AM
1. How would you define ethical AI?
> Generally, ethical AI would be designed to help and not hurt. In the LLM space, it wouldn't try to tell you something that isn't true, either to hide something, obfuscate it, or outright make something up. On the user's side using it ethically would look like using it for magnanimous means, i.e. trying to help people do good things.
1. Are you actively thinking about ethics when using LLMs?
> As a user I'm trying to create with it so the end consumer of the content goes away with more knowledge, unique insights and just all around being a better person.
1. Are you doing anything on that front in your company?
> We are actively involved with creating systems that use LLM as part of a larger content automation process, so we're including safeguards to make sure any output is a value add.
Mika Lepistö
Mika Lepistö
Dec 19, 2024, 9:39 AM
1. It depends on the AI application/use case. If you can add some more detail I could respond better. For example, wearables have a whole different impact on my view of AI ethics than using an LLM to help with content.
2. As a consumer, no. As someone who's involved in an industry creating content it may have trained from, yes.
Emina Demiri-Watson
Emina Demiri-Watson
Jan 2, 2025, 7:02 AM
@Mika Lepistö thanks. Yes it's more using LLMs in SEO and not things like wearables.
Mika Lepistö
Mika Lepistö
Jan 2, 2025, 8:13 AM
This is a hard topic to discuss at surface level depth, for me at least.
From the context of an LLM, I would define ethical AI being an LLM that was trained on data owned by or licensed by the entity training it.
The debate happening legally outside that scope is interesting, and I'm personally conflicted, which adds complexity to the previous statement.
On one side it theoretically is "learning" like humans can. On the other side it is basically predictive text that learned the predictive process by consuming published works.
If the published works were put out on the internet for anyone to consume, then is that not like putting your trash at the street? It's intentionally publishing for anyone to dig through, including bots. We just didn't see all the bots and what they were doing. But, just because it was published freely doesn't negate IP/copyright law.
If it's paid content, and someone paid to access that content within the license terms, and learns from it (programmatically or as a human) is that not legal?
Circling back, it feels like the easy place for me to land is if the output of the LLM doesn't violate the license, is that's legal.
The ethics of business are gray, so if it's legal is it ethical? Maybe not by the standards of humans who feel screwed over, but are those people OK with using it? If so, then like myself with a conflicted/double standard position I think we may need to accept it's ethical if we choose to participate in using it en masse, because we are accepting it.
Ethics are also subjective moving targets, and separate from legality although there can be overlap. A thing that was accepted as ethical in history may not be now due to changes in societal norms. On a different thing the opposite may be true.
Summarize that beyond "it's complicated"! ????
Maddy Osman
Maddy Osman
Jan 2, 2025, 10:20 AM
Hey @Emina Demiri-Watson — we tackled the topic from the generative AI perspective in this article:
Happy to dig in more if you’d like ????
Some quick responses to your questions:
*How would you define ethical AI?*
Essentially, being thoughtful about how you’re going to use the outputs. If there’s a commercial use case, like generating an image for a business website, did you provide the inputs? Because if you didn’t, the end result was trained on other’s work and you’re probably not even aware of who that creator is.
Another big ethical dilemma revolves around how carefully you must validate details with any outputs shared, since AI is confidently incorrect and hallucinates.
*Are you actively thinking about ethics when using LLMs?*
Yes, by nature of my work with creating content and wanting to lead by example for good karma around my body of content work!
*Are you doing anything on that front in your company?*
We’ve detailed our AI policy in terms of how we use AI internally and what clients can expect here:
Is using AI plagiarism? To understand the ethics of AI-generated content, we must consider how it’s created and the different forms of plagiarism.
The Blogsmith: Is It Plagiarism To Use AI-Generated Content? The Ethics of Content Creation in an AI World
Our generative AI policy has zero tolerance for AI-generated content, and we take several steps to ensure all content we create is accurate and original.
The Blogsmith: Our Generative AI Policy
Emina Demiri-Watson
Emina Demiri-Watson
Jan 3, 2025, 2:13 AM
Love this! Thanks @Mika Lepistö and @Maddy Osman! I agree it’s a minefield.Maddy can I use your policy as an example? We are now working on ours as well and looking into the new ISO for it. Part of me feels like maybe its to early for this talk. I don't think I’ll have any actual answers for attendees but at least we can show where we are with this and get people thinking I guess…
Maddy Osman
Maddy Osman
Jan 3, 2025, 7:36 AM
@Emina Demiri-Watson yes, please feel free to use as an example :) I think it will be an awesome talk!
Emina Demiri-Watson
Emina Demiri-Watson
Jan 3, 2025, 11:24 AM
Awww thanks @Maddy Osman! Fingers crossed!
MIchael Buckbee
MIchael Buckbee
Jan 4, 2025, 9:53 AM
@Maddy Osman This isn't like a gotcha, but the line on generative AI seems really blurry. I've had some content marketing + SEO's give me pushback for both Grammarly ("that's the job of a copy editor and Grammarly sometimes makes mistakes.") and Clearscope ("that's the job of a search strategist").
Maddy Osman
Maddy Osman
Jan 6, 2025, 10:53 AM
@MIchael Buckbee I'm not sure I totally agree (in terms of these specifically mentioned tools) but I I think I follow what you're saying (please correct me if I'm not).
In terms of this policy, it would probably help for me to go a step deeper to share more about how precisely we're using these tools. For example, with Grammarly, we use it as a basic spelling and grammar checker but we *mostly* use it for the custom style guide feature that allows me to basically dump all the rules I've outlined in my ridiculously long style guide () for easy reference as a quality check. We do not use any of the generative AI features of Grammarly/Frase. I have those turned off as an organization in Grammarly.
In terms of Clearscope, in our process, it's just another tool for data, like Semrush or Ahrefs. I'm kind of generalizing by adding it to the AI policy because it's more machine learning than (generative) AI. At any rate, we do keyword research on Semrush, Clearscope, Frase, and MarketMuse for every piece of content that has a search purpose. So that's part of our SEO team's job to curate all this info together and then pass the baton to the writer.
The Blogsmith Writing Style Guide is free for all to use and we hope it helps you to effectively create content for humans and robots.
The Blogsmith: The Blogsmith's Writing Style Guide

Our Values

What we believe in

Building friendships

Kindness

Giving

Elevating others

Creating Signal

Discussing ideas respectfully

What has no home here

Diminishing others

Gatekeeping

Taking without giving back

Spamming others

Arguing

Selling links and guest posts


Sign up for our Newsletter

Join our mailing list for updates

By signing up, you agree to our Privacy Policy and Terms of Service. We may send you occasional newsletters and promotional emails about our products and services. You can opt-out at any time.

Apply now to join our amazing community.

Powered by MODXModx Logo
the blazing fast + secure open source CMS.