How NOT to use AI in your business

AI is powerful, but it can also get a lot of businesses and people into a lot of trouble. ChatGPT is now acing exams, it’s able to change its personality, it now accepts visual inputs. And there are now calls for the development of AI projects to be slowed right down, from the likes of Elon Musk and Steve Wozniak. So how could you use AI badly? And what are some of the repercussions to you as business owners and marketing teams? Well, today we’re going to look over some of the examples of how you shouldn’t be using AI in your business, and ways to ensure that you’re able to use it correctly.

Hi, my name is Jordan McFadyen, and welcome to the Marketing Mindset podcast. This podcast is really to challenge both you and I to always be reassessing our marketing strategies, with new thoughts and new ideas on how you can move away from random acts of marketing, and start helping you to build a more long-term marketing strategy.

And as I said at the start of this podcast, AI technology is moving fast, like really fast, but too fast for some. There are some tech leaders, including Elon Musk and Steve Wozniak and a few others. They released an open letter recently. They’ve called for a temporary halt to artificial intelligence development. What they’ve done is they’ve cited potential risks to humanity as part of this open letter. They’ve expressed concerns about the rapid advancements in AI technology, and really trying to highlight its potential misuse.

And what they’re doing is they’re trying to urge global cooperation, essentially, in a bit of a pause in its development, really just to allow a more comprehensive evaluation of the technology’s impact on society, and also on the environment as well. So it’s been a really interesting week for AI development. What’s going to happen, we’re not entirely sure just yet, but with big tech leaders like that leading a movement like this, it does really show us how fast this technology is moving.

And look, there are some big concerns that I’m reading, and also feeling myself as well. There is a lot to think about and to deal with in terms of how fast AI is moving. And I literally went to ChatGPT and I literally put in there to find out some of the limitations and some of the things that we should be aware of. And this is literally out of ChatGPT’s own mouth, saying that we need to be concerned around data privacy and security. The tool was able to tell me that AI systems are often relying on large amounts of data and information. And there are concerns, obviously, about the protection of some sensitive information as it goes through these different AI learning machines.

There’s obviously the big ethical concerns. AI can sometimes perpetrate existing biases in data. And what this is going to do is it’s going to lead to unfair or discriminatory outcomes as well. The other big one that people are concerned about, and ChatGPT literally told me, was obviously job displacement, that fear of AI actually being able to replace human jobs.

And then there’s the dependence on data quality. Because these systems, they’re only as good as the data that they’re actually being trained on. So if they’re being trained on poor quality data, what this is going to do is it’s actually going to lead to inaccurate or unreliable results, which is obviously going to negatively affect businesses and business decisions, if you were using the tool to guide you on some of those decisions, which we’ll talk about a little bit later as well.

The other thing that ChatGPT literally told me was that one of the big concerns is legal and regulatory compliance. So the adoption may bring new legal challenges in both those spaces. And then there’s also the big lack of understanding: business leaders not fully understanding their capabilities, but also, the big one, the limitations of AI, which is going to lead to unrealistic expectations, or even worse, inappropriate applications of AI technology as we move through.

And that’s just a few examples of what AI literally told me were the risks of using AI. It went on to talking about deepfakes and misinformation, cybersecurity threats. Also got right in depth around economic inequality as well. It’s really crazy stuff to read. And even to hear it straight out of the AI’s mouth, you could say, was even more crazy to research it.

And in a real-world example as well, I even today just came across a post on LinkedIn. And what it was, it was an image-based post. And this image, it was a photo, and it showed Boris Johnson. Boris Johnson was being arrested. And it was an image that was completely made up, completely made up from AI and a tool called DALL-E. And the author of the post said he literally created that image in minutes, which is terrifying, because it did look really real.

If you’re just scrolling Facebook or just scrolling LinkedIn, and you saw this shot of Boris Johnson being arrested with four or five police officers, you would immediately think that he’s done something wrong. So I really think that it is crucial that in 2023, we need to remind ourselves that not everything online is going to be as it appears. So we really need to start putting that filter on. And I know that you and I, we’re going to be able to decipher. We might get tripped up a little bit initially, but once we do that research, we’re going to realise it’s all made up, or a deep fake. But for a lot of other people, they’re not going to have any clue. They’re not going to have any clue that this technology exists, and that these level of deep fake can actually be created as well, which is scary.

And then this filters down to businesses and business owners and marketing teams, and how they are going to use this tool. Especially, this is a big one, especially when it comes to content creation. And what I think we need to do is we need to remind ourselves that AI should really be used as a tool for enhancing productivity and enhancing your decision making, rather than actually having that focus on replacing human workers, especially when it comes to content creation. It’s not a tool for replacing someone. It’s more around how we can enhance productivity, make us work faster, make brainstorming easier, decision making around what title you should use. All that stuff is where it should be really powerful.

And I need to stress the importance of data quality. And it goes back to what I was saying before: data quality, transparency, and just understanding the limitations of these AI systems. Because we need to ensure that we’re using them in responsible and effective implementations as well. And there are a few examples that have been floating about on the internet that I wanted to share with you today, and how this could actually impact you and your business if you go ahead using these tools without really looking at it.

It was widely reported that CNET, now, CNET is like that technology publisher. So CNET had reported that they had written around 77 AI finance stories. They quietly published this on its websites. And what they then had to do is they then had to go back and retract and issue corrections to 41 of those 77 articles. And these were on basic finance articles that they’d put across their website. Basic explainer articles like, what is compound interest? And supposedly the articles, they say, were assisted by an AI engine, and then reviewed and fact-checked and edited by editorial staff. But clearly that wasn’t the case. And clearly what had happened is that team had relied too much on the tool, and pushed publish far too soon, instead of actually fact-checking what the AI tool had actually pushed out.

The main takeaway here is we don’t want to use these AI-powered tools unsupervised. It’s just like, this weekend I’m not going to suddenly get my chainsaw out and use it completely unattended. It’s a tool. We’ve got to treat AI like it is a tool.

The other example I wanted to share is that AI-powered tools are in no way a replacement for our existing research tools. At least not yet. They’re not there just yet. They can be really useful. And this is where I find them really helpful, really useful in my business. It can be really useful of generating ideas, but they are going to likely lead you astray if you’re not careful. So I wouldn’t say you’re going to use this purely as your research tool, because I honestly think that you can use it to help with your brainstorming, getting you in the right direction, but don’t take what the tool is giving you as gospel.

And honestly, I also think that when we’re talking about written content, this is where I think video is going to become so much more powerful. So to business teams and marketing teams and business owners, sorry, this is where I think video needs to become more a part of your marketing strategy, as a video requires human intervention. Even with this podcast here, obviously we can write ideas, we can write bullet points, headlines, things to make sure the conversation of this podcast and this video is going in the right direction. But still, it requires human intervention to actually publish it, to actually present it, to record it. So I do think that video on a side is actually going to be a really powerful tool as we continue with AI technology and AI tools.

Now, a third thing I want you to consider as well is, like a lot of people working in the digital marketing and the tech space, I’m really excited personally around AI-powered tools and what they’re potentially going to be capable of. I’m scared about how fast it’s moving, so I can understand what they’re saying in terms of slowing down developments. But I am also really intrigued to see what’s going to be capable over the next few years. I’m keen to see where I can fit that into my personal workflow, how we can fit it into the agency’s workflow here, and how we can make our processes in the background work much faster, and be able to offer a much better service to all of our clients.

But it’s not okay to go all in on AI without letting clients and customers, or your readers or your viewers, actually know. Otherwise, if you’re going to use AI secretly to deliver a service, or using AI in an unethical way like that, and you do make a mistake, or the AI tool makes a mistake, you’re likely going to be really embarrassed, just like those CNET examples that I gave you earlier.

So I would honestly say that if you are going to use AI tools, especially when it comes to delivery of your services, you need to let the clients know that you’re embracing this technology, and that’s where you’re embracing the technology and how you’re embracing it, especially if it is directly related to an output that you’re going to provide that client. Not necessarily the processes behind your business, and how your business works to be able to get tasks through to certain team members at certain times. But if it is something that you’re actually going to be then handing off to their client, I do think that you obviously need to make sure that that client fully understands, to save any embarrassment if there’s any type of mistakes.

And with that all being said as well, I also think that there are some great ways to use AI. I have done a podcast on this previously, so you can go back maybe just a few weeks, and we actually share a bit of a process on how you could write a blog article using AI to help with that brainstorming. But also, the things that you can ethically use AI for from right now is sparking those content ideas. That’s the big one that I use it for: coming up with content ideas, and how you can even lay out your blog post or your article.

I also find it really helpful, and I know a lot of people find it helpful, with writing drafts to business emails. Even for those mundane emails of, if someone missed a Zoom call or anything like that, just adding that into ChatGPT can be really quick. “So-and-so missed a Zoom call, how can I essentially rebook them for another call?” It’ll come out with a really well-written email, which can work really well. Obviously, you’re going to have to put a few tweaks in to make it more personal. It can help inspire your marketing copy.

Meeting summaries. This is a big one. This could save us so much time in writing summaries about different meetings, taking notes from meetings. There are tools that you can actually plug in, like Otter.ai, that will attend your Zoom meeting and then be able to actually give you a transcript. The transcript, you can then put that into a tool like ChatGPT, and it will write out the summary and the action points of what was the outcomes of that meeting. So that can be also really powerful.

There’s all those other tools that I mentioned before, like Otter.ai. That one’s great. If you invite them to a Zoom meeting, it can help transcribe the notes and help you be able to send meeting summaries to the client straight after a meeting. Grammarly is great. It’s a tool that has been around for ages, and it’s using AI technology to help you write better, write better with your emails. Then there’s tools to help you with social media videos, cropping them to different layouts, and speed up that process.

And like I said before, we’ve actually got a full podcast on the topic of AI, and how I’ve been able to use AI, and an example of how you could use AI to actually research and create a blog post, and create summaries to a blog post, creating the headings on top keywords, brainstorming titles and meta tags and all those different things. It can really help you speed up that process, and give you essentially, again, a tool to help you unblock that writer’s block.

So look, AI, it is moving fast. And we said that at the start of this podcast as well. It’s too fast for some. They’re calling for it to slow down over the next six months. But if you use it for truth, that’s where you’re going to run into problems. If you use AI as a tool, then I honestly think that you and your business and your content creation could go much further. What’s going to happen in the future, and whether development is going to slow down or not, that’s still going to be questioned. And I’m personally going to be watching with a lot of interest. But how you ethically use the tool to support your business processes is where I honestly think that you could see some really good positive gains.

And look, hopefully, if you can use some of these quick tips, and these tips of what not to do with AI, I hope that it steers you in the right direction of how powerful these tools can be, but also some of those limitations on where I don’t think you should be using these type of tools.

And if you’ve enjoyed today’s podcast, I really would ask you to consider hitting that follow button so you can get more of our content. Or please give us a star rating or a review. It is the best way to support this little podcast. And hopefully, it gives you some good ideas to help your marketing strategy grow and evolve over time, so that you can move away from those random acts of marketing.

And look, if you want to get all the access to our other podcast episodes, or any of our other blogs or articles that myself or the team create, please go to our website, donebynine.com. You’ll also find our contact section on there as well, where you can either just send us a message and have a chat to me or a member of the team. Or if you actually want to sit down and work through your marketing strategy, you can also fill out our onboarding form there as well, and we can get in touch and see about how we can actually work together in creating your marketing strategy as well. But again, thank you so much for joining us on another episode of the Marketing Mindset podcast. And we’ll look forward to seeing you on the next episode this Thursday.