Every day on social media, in videos & podcasts, and within direct messages & emails, I see and learn about new ways generative AI can be used to supplement or even replace work that’s being done.
☑️ Write blog posts and email newsletters.
☑️ Conduct pre-call research and generate sales letters.
☑️ Build prototypes of videos or entire marketing campaigns
The more businesses and marketers are willing to invest time in training, guiding, and customizing available AI tools, the better results they see, and the deeper AI is able to integrate into and improve their work.
All of which I’m incredibly excited and optimistic about.
But there’s a ‘but’ there, isn’t there?
It’s great that we’re able to bring increasingly advanced and capable artificial intelligence to the fore, but where does AI stop and Humanity remain firmly in place?
At what point do we say, no, that particular task should never have the human replaced by the machine?
It’s a potent question because we’re rapidly approaching the point where virtually any task you can imagine could be assumed by AI. It’s no longer a question of “could” but rather one of “should.”
So that’s the topic we’re going to explore today.
To AI or Not To AI, That Is The Question
As you may know, I’m the proud owner of 6 different podcasts, and counting. I host 5+ podcasts for Agorapulse, and have my AI in Marketing: Unpacked podcast as part of my side hustle. Since I’m a podcaster now, I thought it would be a great idea to do a deep dive into the industry and try to glean best practices when it comes to production & promotion, so I attended Podcast Movement 2024 in Washington D.C., which was an eye-opening experience.
My biggest fear (aside from my normal travel and peopling anxieties) was that I’d attend a bunch of sessions on podcast marketing and learn nothing I didn’t already know about marketing.
Fortunately, while there was certainly a great deal of superficial marketing advice (start a newsletter! 🤦🏼♂️), there were more than enough tips, insights, and nuggets of wisdom for me to come home feeling extremely optimistic. If you have a podcast and would like to know more about what I learned, just ask, I’m happy to share.
As you might imagine, there were dozens of sessions that included or were entirely about AI, from tools to tactics. However, it was during a session on interviewing that a topic came up which sparked my interest – and today’s topic.
The panelists were discussion interview prep and interviewee prep, and the question of whether or not to share questions with the guest came up. The panel unanimously agreed never to share questions with guests in advance – which I don’t agree with. I understood though that these panelists and their podcasts were more consumer-facing, news/celebrity guest formatted, and not the B2B business format all my shows enjoy.
But then one panelist said something that I found startling – “Never use AI to draft the questions.”
And again, all of the other panelists chimed in their agreement, and went on to bemoan how generic and simple such AI-generated questions would be, and why would you ever do that to your show or guest.
I remained quiet because I felt that wasn’t the forum for a heated debate on the topic, but dang, I couldn’t disagree more! It’s only thanks to AI’s help that I’m able to pull off all the interviews I have to do every week to host 6+ shows. And yes, the questions are predominantly created by AI.
And my guests respond all the time, “that’s a great question!” So clearly these aren’t generic, uninteresting topics of discussion. The difference, I believe, lies in the quality of my pre-work, which we’ve talked about before. Because I’m using a custom GPT to craft the entire podcast interview, including deciding in advance what outcome I want from the show and ensuring the line of questioning leads us to that point, the output is excellent.
I suspect these panelists, if they’ve even tested AI at all, simply opened a new chat and asked for interview questions, and therefore received the most generic – the most probable – questions possible.
The fact is, as Chris Penn demonstrated in our interview, one could use AI to actually analyze every interview someone has ever given and generate questions they’ve never been asked before.
I share that story because it illustrates part of the question of whether or not to use AI. The part that has to do with whether or not AI can perform at a level sufficient to make it worthwhile to use. And we’re increasingly getting to points where that’s not a question if one has the expertise to implement AI appropriately.
Whether it’s a matter of engineering the right advanced prompts, using the best tools, and investing the right amount of time and resources, there will soon be few, if any, tasks that couldn’t be accomplished using AI.
So that boils the matter down to the core question: when is it inappropriate to use AI? When would you not only be better off using a human, but perhaps unquestioningly so?
Determining When Humans Must Be Involved
Since we’re talking about AI, I thought it would be fair to allow AI to participate in the conversation. When asked, here’s what ChatGPT 4o had to say:
When AI Should Be Used:
- Data-Driven Tasks: AI excels in processing large volumes of data, identifying patterns, and making predictions. Tasks like analyzing customer data, automating reporting, or optimizing ad campaigns are perfect for AI. It can crunch numbers faster and more accurately than humans, leading to better-informed decisions.
- Personalization at Scale: AI can personalize experiences for thousands of customers simultaneously, something a human team simply can’t achieve. For example, AI-driven email marketing can tailor content to individual preferences based on past behavior.
- Routine and Repetitive Tasks: Anything that’s mundane or repetitive, like scheduling posts, sorting customer inquiries, or updating databases, is prime for automation with AI. This frees up human time for more strategic work.
- 24/7 Customer Support: Chatbots and AI-driven customer service tools can handle basic inquiries around the clock, providing instant support and freeing up human agents for more complex issues.
When Human Involvement is Essential:
- Strategic Decision-Making: While AI can provide data and insights, the strategic direction and big-picture decisions should still be guided by humans. Understanding context, interpreting nuances, and aligning with broader business goals requires human judgment.
- Creative Work: Creativity involves more than just combining elements in a new way; it’s about storytelling, empathy, and connecting with an audience on a deeper level. While AI can assist by providing suggestions or generating content, the core creative ideas and final outputs should be driven by human creativity.
- Ethics and Values: Decisions that impact company values, ethics, or social responsibility should never be left to AI alone. Humans need to ensure that the business aligns with its moral compass, something AI can’t adequately understand or uphold.
- Crisis Management: In situations where a brand’s reputation is on the line, human intuition, experience, and empathy are critical. AI might misinterpret the situation or fail to grasp the emotional undercurrents that need to be addressed.
- Building Relationships: Marketing, at its core, is about relationships. Whether it’s engaging with customers on social media, networking with partners, or negotiating deals, the human touch is irreplaceable. AI can assist, but it can’t replicate the genuine connection that humans can build.
I would say that’s spot-on.
In my mind, before asking ChatGPT, my thoughts were focused on strategy, empathy, and relationships.
My podcasts for Agorapulse are all about building relationships – both with the interviewee and the larger audience overall – and therefore my role as interviewer is essential. While AI can now carry on conversations and have pre-determined steps and approaches, it would fail utterly at the goal of building a relationship with the interviewee. When discussing challenges they may have faced in their marketing journey, an AI would not be able to authentically demonstrate empathy.
But I also think it’s important to highlight ChatGPT’s final comment, that “AI can assist, but it can’t replicate.” I believe we’re at a stage where AI is essential in order to improve everything we’re doing beyond our capabilities. While AI can handle a lot, it’s crucial to recognize when the human touch is not just better but essential.
If you’ll forgive a sports metaphor for a moment, it’s akin to Michael Jordan being one of the greatest basketball players of all time. Someone that, if you were to play them 1 on 1, he’d wipe the floor with you. And yet, to actually achieve championships, Jordan was always surrounded by an excellent team and support staff. The greatest player in the world cannot win a championship alone.
Similarly, in the world of business and marketing today, professionals must look for new and innovative ways to bring AI onto their teams and give them that additional advantage. If you’re Michael Jordan, AI is Kobe Bryant. An incredible resource and arguably as good as – or even better than – Jordan, individually. Together, you’re a championship-calibre team.
In every area above, which ChatGPT and I agree that human involvement is essential, the assistance of AI can help strengthen at activity and ensure success.
- Strategic Decision-Making: AI can ensure decision-makers have the best possible information, as well as help identify gaps in knowledge or awareness.
- Creative Work: AI can provide critical brainstorming and research which can often suck creative energy from marketers.
- Ethics and Values: AI can analyze assets and identify topics, language, or direction which might not be inclusive, or troublesome in other ways.
- Crisis Management: AI can offer alternative suggestions and advice which might actually prove to be industry best practice, rather than emotional response.
- Building Relationships: AI can help identify opportunities for smart marketers to pursue, expediting relationship-building activities.
AI is a powerful tool that can amplify business and marketing efforts, making them more efficient and data-driven. However, it’s essential to recognize its limits and know when the human touch is not just valuable but necessary. Strategic thinking, creativity, ethical considerations, and relationship-building are areas where humans must take the lead, ensuring that AI remains a tool that enhances rather than replaces the human element in business.
If this is a topic that you’re struggling with within your business, or perhaps more basically, you’re looking for guidance on how to approach and integrate AI, let’s jump on a call and talk it through. I can help you consider your ultimate goal or objectives, what strategy you should pursue to get there, and which AI tools or tactics can be applied to help you along the way.

Related
Discover more from The AI Hat
Subscribe to get the latest posts sent to your email.