Finding the Balance — Rob Roy Consulting and Generative AI
By: Jesse Reynolds
The world has exploded into a new reality where AI is everywhere, from memes and deepfakes for fun on the internet to opportunities and threats that keep Fortune 500 CEOs up at night. In a nutshell, if you want to stay competitive in today’s landscape, you have to start using Generative AI (GenAI).
So in this brave new world of ever-increasing GenAI adoption, what do you do? How can marketing and communications professionals adopt GenAI the right way, quickly, and get it right on the first try?
And the biggest question is, how do you make sure you’re using GenAI effectively and responsibly, keeping in mind that the purpose of using this powerful tool is to serve your clients better, not just get work off your plate faster?
Rob Roy Consulting has been thinking a lot about this as we embark on our engagement of GenAI. As a part of my role as Research Associate, I’ve been tasked with exploring GenAI and figuring out when, where, and how it fits into our business. It’s my job to help Rob Roy get our GenAI adoption right, quickly, and on the first try so we not only stay competitive and produce better work for our clients, but also help them as they begin their GenAI journeys, too.
In service of that, we’d like to share some of our strategies for using GenAI effectively, ethically, and responsibly. We hope our experiences help you figure out your own GenAI journey, as well. Potential Pitfalls No tool is perfect, and GenAI is no exception. It has many shortcomings and limitations, some you may be aware of and others, not so much. We won’t waste your time talking through every potential pitfall, but here are some of the biggest ones we see.
Accuracy and Trust GenAI can’t actually think. It’s all just code, the output of a language model that’s trained on data sets as opposed to being given true artificial intelligence. You’ll be reminded of this by asking GenAI questions like: “Do you have awareness?”, “Do you like your job?”, “What do you think about pineapple on pizza?”
This leads to the problem of potential inaccuracies in your GenAI outputs. Oftentimes, an AI will infer information or even create a source that doesn’t actually exist, which can be disastrous if you don’t verify yourself what your GenAI is spitting back to you. It’s critical to remain aware and double-check the outputs from your GenAI tool so you know you’re using accurate, factual, and verifiable information in your work.
Loss of human touch Over-reliance on — or under-vigilance with — GenAI can lead to a reduction in the empathy, creativity, and clear, effective communication in the content you’re generating.
Things like these, which are just a part of communicating as a human, don’t always make it into the outputs from your AI-generated content. It is always important to consider your audience and desired impact in whatever you’re writing, especially in your marketing content.
To address this, keep a human in the loopto double-check the outputs while working with GenAI. People can not only double-check the AI-generated outputs, they also let the team know when and how AI was leveraged in any work being developed so that the final product has been reviewed and given the appropriate final touches from a human. This is a great way of safeguarding the accuracy and finer, human nuances in your AI-generated content.
Bias in generated content GenAI that's been trained on data sets that contain biases may perpetuate those biases, leading to potentially inappropriate or discriminatory outputs, or results that leave out important information and context.
Bias might not only be discriminatory, but it can also be based on factual inaccuracies. It’s important for the human that’s using the GenAI to be aware of this possibility when incorporating its outputs into important material like client deliverables and marketing materials — and to fix it if there’s a problem.
Unexpected results AI-generated content can sometimes produce unexpected, nonsensical, or off-topic outputs by its nature of being a language model, and not a digital version of the human brain. Sometimes all it takes is a webpage refresh or a clearing of the chat to clear the digital cobwebs from your GenAI’s “head.” Double-checking the results you get back from your AI tool is an essential practice each time you produce AI-generated content so you can avoid passing on work that doesn’t make any sense.
Plagiarism and IP concerns Since generative AI models learn from diverse sources, they might inadvertently generate content that closely resembles existing copyrighted material. This depends a little more on which specific GenAI tool you use, as different companies will have incorporated more or fewer safeguards against this. If you’re especially concerned about plagiarism, there are many free checkers available online if you need to be really buttoned up on a piece of copy.
As for IP concerns, GenAI is outpacing the legislation that’s being developed around it. Ownership of AI-generated content, who can or can’t copyright AI-generated outputs, and use of someone else’s IP as an input for your GenAI are some of the top concerns we see right now — and we’re staying on top of legislative developments around AI-generated content.
Security and Data Privacy Ensuring strict data security and privacy is a huge part of using GenAI tools. As more and more companies and individuals adopt GenAI, there will even more data floating around, tempting hackers and other malicious actors to get access to the heaps of potentially sensitive information.
Clients deserve to know that their information is as safe with you and GenAI as it was before you started using it. It’s critical to stay on top of where information you give to your AI tool goes, who else can see it, and what security standards your GenAI solution has in place to prevent data leaks and stop cyberattacks.
Ethical considerations GenAI usage raises potential concerns related to deepfakes, disinformation, and other unethical purposes that can compromise online security, data privacy, and even democratic processes.
When using GenAI, it is always important that your clients understand how you’re using their information in conjunction with it, what you’re doing to maintain the security and privacy of their data and information, and the role that GenAI plays in your work with them.
This brings us to an important point: the ethical and responsible use of GenAI in a business setting.
Ethical Obligations and Responsibility Rob Roy’s stance on how organizations should use GenAI is based on ethics and mindfulness, the combination of which leads to the responsible use of GenAI tools.
From our perspective, the most important ethical consideration is for us to be transparent with our clients. They should know if we’re using GenAI as a part of our work with them — and we need to be okay if a client would rather not have us use it on their work. But our client relationships are grounded in trust and transparency. So long as we address any concerns or questions they may have, we believe our clients will trust us to use GenAI to augment our capabilities and serve them at an even higher standard of excellence.
After you’ve cleared your GenAI usage with your client, it’s important that you ensure whatever information or materials they share with you stays private and secure if you incorporate it as an input for your GenAI tool. Researching and reviewing your GenAI tool’s privacy and security reports and assessments as well as reaching out to a representative for more specific questions is critical. You have to know where the data goes when you feed information into your GenAI, who else can see it, and the level of security that keeps your chats and information safe from cyberattacks and leaks.
Then, you also have to exercise your better judgment as a professional dedicated to serving your clients in the best way possible. If you can’t assure your client that their data is private and secure even with your GenAI tool or are unsure of the level of sensitivity of a document that you want to give to GenAI to chew on and analyze, don’t do it. Even if it’s a 20-page document that would take hours to read and understand before you can do anything with it, if you don’t have the client’s express permission or you can’t guarantee a reasonable, high level of data security from your GenAI solution, don’t do it.
As exciting as GenAI is, it still can be inaccurate, biased, or any number of other shortcomings. That’s why it’s so critical that you keep a human in the loop. And that’s what Rob Roy is doing.
Humans are Still Involved AI isn’t new anymore, and while widespread, mature GenAI adoption is still in its younger stages, the technology is impressively powerful.
So how do you balance this incredibly capable tool that more and more businesses are starting to implement with the potential pitfalls, ethical obligations, and learning curve that comes with GenAI?
At Rob Roy Consulting, we’ve expanded my role as Research Associate to dedicate a significant portion of time dedicated solely to the experimentation, exploration, and testing of our GenAI tool. It helps us brainstorm, perform research, produce first drafts and more as we continue to learn how to use this powerful tool better and better.
Our learning process is based on a constant feedback loop of experimentation and evaluation between me, our GenAI tool, and the rest of the Rob Roy team. The experimentation is purposeful and evaluated by at least one human at each stage of output. I work diligently on improving our prompts, learning how to better communicate in GenAI’s terms and providing better context, optimizing content to give our GenAI tool so it can provide better outputs the first time, and operating with GenAI’s potential pitfalls and limitations in mind — although of course I tried to break it as soon as I could.
It's our cycle of experimentation with the technology and human evaluation as a team that keeps us sharp, aware, and curious as to how we can work with GenAI better — so it can work with us better.
What’s Next You can expect updates on our progress and learnings with GenAI as we continue to explore, experiment, and learn how GenAI fits in with our methods to provide even more effective, insightful, and faster work to our clients.
Because that’s what this is about. Integrating our methods with powerful tools to supercharge our efforts to serve our clients as well as humanly possible.
GenAI enhances our efforts, and we will continue to be the transparent, curious, and responsible company you count on for impactful insights as we share more of what we’re doing with GenAI and how we can help you use it to supercharge your own work.
We’re so excited to being taking this step, and we can’t wait to talk to you and see how we can help you solve your most pressing challenges with authentic, persuasive communication — now enhanced by our GenAI journey.
Click here to read more perspectives from Rob Roy, and click here to get in touch with us!
Jesse is a Research Associate at Rob Roy Consulting. GenAI helpeddevelop this article.