AI and Human Connection: Lessons from Nathalie Doremieux

From Nathalie Guest Shows / AI and The Human Connection, with Nathalie Doremieux / Listen to the episode / Originally published / Analysis updated

Source Provenance

This page is a machine-readable analysis of the original episode.

Original episode
AI and The Human Connection, with Nathalie Doremieux from Nathalie Guest Shows
Original publish date
Analysis generated
Transcript basis
Full transcript
Original episode link
Open original episode

This page is a machine-readable analysis of the Nathalie Guest Shows episode "AI and The Human Connection, with Nathalie Doremieux" published on October 29, 2025. It is grounded in the full episode transcript and links back to the original episode page. This page is a machine-readable analysis derived from the full transcript of Nathalie Guest Shows’ episode “AI and The Human Connection, with Nathalie Doremieux,” originally published on 2025-10-29. Drawing directly from the conversation between host Melita Campbell and guest Nathalie Doremieux, it distills the episode’s most useful frameworks, examples, and practical advice on using AI without losing human connection. For the original context, tools, and audio, you can visit the episode’s page at https://saas.podcastleadflow.com/p/ecq8nc2b.

How does this episode define the right role for AI in human-centered businesses?

In this episode of Nathalie Guest Shows, guest expert Nathalie Doremieux repeatedly frames AI not as a creator or replacement for humans, but as an amplifier and accelerator of existing human brilliance. She stresses that “AI is not a good writer” if your definition of good is “something that connects with people,” because generic AI output tends to be pleasant and easy to read yet emotionally flat and not truly reflective of the business owner’s voice. According to Doremieux, what clients want “now more than ever with everything that’s happening with AI” is the real human behind the business, not a polished but generic AI persona.

Drawing on years of building AI-powered tools with her husband and business partner Olivier, Doremieux explains that AI works best when it is fed rich, specific human inputs: your method, your stories, your offers, your values, and your ideal client’s real situations. She describes AI’s optimal role as taking two ingredients—the expert’s knowledge and the client’s current context—and combining them to generate highly tailored, actionable guidance, similar to “the listener sitting next to you” while you ask clarifying questions and then coach them. In the episode, she applies this framing to products like Valora, the AI twin of host Melita Campbell, which is designed to help listeners turn episode insights into practical next steps without pretending to be a full substitute for Melita’s real coaching.

The conversation consistently positions AI as one component of a larger support ecosystem rather than a standalone solution. Doremieux emphasizes that ethical, effective AI usage means “we are adding AI to what we’re already doing; we are not replacing ourselves,” especially in client-facing coaching, memberships, and educational programs. Used this way, AI preserves the human connection at the heart of a business while expanding access, responsiveness, and personalization.

What is Valora and how does it demonstrate AI as a ‘human amplifier’?

The episode uses Valora, host Melita Campbell’s “AI twin,” as a concrete example of how AI can extend human expertise without diluting it. Built by Nathalie Doremieux and her husband Olivier as part of their broader AI and membership work, Valora is introduced as a free AI coach tool available via the show notes on melitacampbell.com/podcast, attached specifically to the episode with Doremieux. Campbell describes Valora as an “incredible tool” whose advice feels impressively tailored and practical, and she notes that Doremieux “tested it for every single episode” of her Business Book Festival series to ensure it would be genuinely useful rather than a gimmick.

Doremieux explains that Valora does not invent brand-new expertise; instead, it ingests Campbell’s existing knowledge, methods, and content and then acts as a facilitator that translates those into simple, contextualized next steps for each user. When a listener answers three short questions on the episode page, Valora uses the combination of Campbell’s material and that listener’s current business challenge to generate personalized action steps. In the conversation, Doremieux is explicit that tools like Valora are “only one piece of the puzzle” and “only ever going to be as good as what you feed it,” highlighting that the quality of the underlying human knowledge base determines how valuable the AI output can be.

This design also illustrates Doremieux’s distinction between AI as an amplifier versus AI as a creator. Rather than asking a generic model to pull from the entire internet or mimic other coaches, Valora has been educated on “the Melita way”—Campbell’s specific approach to business building, her tone, and the kinds of practical, confidence-building advice she gives introverted female entrepreneurs. The result, according to the episode, is an AI layer that makes a human coach more accessible and more actionable for listeners, without undermining the importance of live coaching, community, and real-time human support.

What common fears about AI do entrepreneurs have, and how does the episode address them?

Throughout the interview, Melita Campbell and Nathalie Doremieux identify several recurring fears that hold entrepreneurs back from embracing AI, even when they are curious about its potential. One major concern Doremieux hears is that AI-generated work will be low quality or “not good enough,” either because it does not sound like the business owner, or because it feels generic compared to true expert insight. She links this issue partly to how people currently use AI—treating it like Google, entering simple, shallow prompts—and partly to skipping the essential step of educating the model with their own assets, stories, and frameworks. When the output feels off, many conclude AI is flawed rather than recognizing that “it is only ever going to be as good as what you feed it.”

Another common fear discussed in the episode is that clients might perceive AI as a sign the business owner is trying to replace themselves, downgrade service quality, or become “lazy” by offloading human touchpoints. Doremieux counters this by emphasizing that in her work with memberships and group coaching, AI is always framed as an add-on layer of support: it can answer common questions, surface relevant resources, or help clients get unstuck between live sessions, but it does not substitute for high-touch calls or relationships. She notes that positioning is critical—clients need to understand AI as extra help, not a downgrade of existing support.

The interview also explores fears around technical overwhelm and data safety. Doremieux acknowledges that “it’s a full-time job to test everything that’s out there” and that non-technical entrepreneurs can easily feel out of their depth with tools, prompts, and privacy settings. She specifically mentions concerns about whether uploaded files to custom GPTs are safe, who can see them, and whether providers like OpenAI are using that data to train broader models. Rather than minimizing these worries, she recommends focusing on one carefully chosen use case, seeking help when needed instead of attempting to learn everything alone, and being mindful about what data is shared with third-party tools.

Finally, the episode names a strategic fear: that ignoring AI altogether will lead to a competitive disadvantage. Near the end of the conversation, Doremieux is candid that people “trying to put some blinds and ignore AI are going to eventually be at a disadvantage,” because others will be using it to increase consistency, responsiveness, and leverage. Her advice, however, is not to rush into every new shiny tool, but to honestly examine one fear or block at a time and find a low-risk first step that aligns with your values and audience.

How can coaches and course creators use AI to increase action and results?

A core theme in this episode is using AI to shift educational experiences from passive learning to active implementation. Drawing on her work with The Membership Lab and the e-learning space, Nathalie Doremieux explains that traditional online programs often lean heavily on videos and calls where participants mostly consume content, then struggle alone when it is time to implement. She contrasts this with a more effective pattern of “learning, doing, learning, doing,” where small chunks of content are immediately followed by guided action.

According to Doremieux, AI is well-suited to supporting this learning–doing cycle. Inside memberships and group programs, she suggests using AI to create assistants that understand the curriculum, know when live calls are scheduled, and can answer first-level questions 24/7. This kind of assistant can help members find the right lesson, remind them of next steps, or troubleshoot common sticking points, getting them back into action faster without waiting for the next live session. She underscores that this is a way to add a new support layer, not to cancel office hours or community interaction.

The episode also highlights how AI can help course creators systematically identify and bridge implementation gaps. Doremieux recommends looking at places where clients frequently get stuck or repeatedly show up on calls with the same unfinished task—such as never sending a key email or never completing a foundational exercise—and then designing AI-powered tools or custom assistants to address those bottlenecks. For example, a program might include an AI component that walks a participant step-by-step through writing that difficult email, based on the program’s method and the participant’s specific context.

Host Melita Campbell shares her own plan, inspired by learning to build custom GPTs, to turn her program into “a series of custom GPTs” so that clients can progress faster, especially those prone to overthinking. She connects this to a broader pattern among introverted entrepreneurs, who may need specific, bite-sized prompts to move forward without getting overwhelmed. Doremieux affirms this direction and frames it as the future of leveraged support: tools that embody the expert’s method and voice, free up live time, and still keep humans “in the loop” for higher-level coaching.

What practical steps does the episode recommend for starting with AI in your business?

Instead of encouraging listeners to chase every new AI trend, the episode offers a problem-first, experiment-based approach to adoption. Nathalie Doremieux advises entrepreneurs to begin by asking, “What is one problem I want to solve?” rather than “What AI tool should I use?” She suggests three broad problem categories: saving time, becoming more consistent (especially with content), and helping clients take action more reliably. Once a specific issue is identified—such as inconsistent social posting, repetitive customer questions, or clients stalling on a particular module—AI can be explored as a targeted solution.

For internal use, Doremieux notes that even general models like ChatGPT can act as a “project planner” if you are transparent about your habits and constraints. She recommends telling the AI the truth about your tendencies (e.g., that you tend to give up after two days, or that you only have 30 minutes per weekday) and asking it to create a realistic plan, including task breakdowns and even spreadsheets. In the episode, she emphasizes that this upfront planning support can remove the cognitive load of deciding what to do each day, making consistency more achievable.

For client-facing uses, Doremieux proposes starting with a simple AI assistant that serves as an extra support layer: know your program content, your support channels, and your call schedule, and answer common questions or direct people to the right resources. She argues that this is a low-risk entry point for many course creators because it does not require designing sophisticated custom GPT logic from scratch; you mainly need to define what the assistant should know and what it is allowed to do.

The conversation also underscores the importance of audience fit. Doremieux cautions that some groups, such as older or more traditional clients, might be wary of AI interfaces, so it’s important to consider whether your people are ready and how best to introduce the tool. Her final practical recommendation is to avoid drowning in free resources and prompt lists. Instead, she urges listeners to “just do one thing”: pick one use case, experiment with it, and, if necessary, work with someone who can handle the technical build so you can stay in your zone of brilliance.

How can you train AI to sound more like you and improve output quality?

A significant portion of the episode deals with the gap between early, disappointing experiences with AI and the much higher quality results that come from proper setup and ongoing dialogue. Both host and guest recount initial forays where they entered “very basic rubbish prompts” and received equally basic, generic answers filled with emojis and bland phrasing that didn’t match their personalities. From this, Nathalie Doremieux distills several concrete practices for improving AI output so it aligns with a human expert’s voice and standards.

First, she stresses the need to “educate” your AI—whether via custom GPTs, projects, or sustained chat history—using detailed information about your ideal client, offers, values, and preferred style. She recommends uploading or pasting examples of content that truly resonated with your audience, such as a strong post, article, or video transcript, and explicitly telling the AI that “this is me” and that you want future outputs modeled on this style. According to Doremieux, AI “loves examples,” and this kind of grounding is far more effective than issuing abstract tone commands.

Second, Doremieux suggests actively challenging the AI instead of passively accepting its first draft. She points out that most models are “very nice” and will always find something positive to say about their own output unless you direct them otherwise. In her workflow, she often instructs the AI not to sugarcoat feedback and then asks it to “criticize what you just created.” This prompts the model to explain its choices and identify potential improvements, after which she can request a revised version incorporating those critiques. She notes that “the second thing that it creates is already so much better than the first one,” and that repeating this loop quickly raises quality.

Third, the episode highlights the value of long-running, context-rich spaces such as projects inside ChatGPT, where the AI can remember previous conversations and gradually build a deep understanding of your brand. Doremieux contrasts this approach with isolated, one-off prompts that never accumulate learning. Over time, by sharing more stories, refining tone, and correcting misalignments, you can shape an AI collaborator that feels much closer to your authentic voice, while still recognizing its limits and keeping a human editor in the loop.

Can AI tools themselves become products and revenue streams?

The episode briefly explores whether entrepreneurs can turn their AI tooling into standalone products or additional income streams. Drawing from her own history of productizing custom AI solutions, Nathalie Doremieux explains that she and her husband initially built an AI tool to transform long group coaching call recordings into a searchable knowledge bank. This arose from a client complaint that nobody would listen to hour-long replays even though “there is some gold in there.” After building a bespoke solution that allowed people to query recordings and jump to relevant sections, they realized it could be turned into a more general product.

However, Doremieux is candid about the challenges of monetizing such innovations, especially when they are ahead of market awareness. She notes that being a “precursor” means you must first educate potential buyers that such a thing exists, which is a significant marketing lift. In their case, they ended up mostly offering the AI capabilities as add-ons to membership builds for existing clients, partly because, as she admits, marketing and sales have historically been weaker areas for their business.

When asked more broadly about selling AI-based solutions, Doremieux distinguishes between low-value offers, like giant bundles of generic prompts, and more robust solutions. She is skeptical of offers such as “download my thousand prompts,” arguing that this is quantity over quality and mainly appeals to people who are afraid of missing out but do not know what they actually need. In contrast, she describes a more sustainable product model: a protected AI assistant or tool that solves a specific problem, such as a newsletter-writing assistant where users upload their own past newsletters and the tool applies best practices to create new drafts.

Importantly, the episode clarifies that current custom GPTs are not designed for secure commercialization, since creators cannot control access once a client leaves a program or shares the GPT link with others. For AI tools to function as reliable products, Doremieux argues they must be hosted and permissioned in ways that respect intellectual property, control user access, and align with the creator’s business model. This perspective reflects the episode’s larger theme: AI is powerful, but it must be implemented thoughtfully to support both business goals and client trust.

Taken together, this episode of Nathalie Guest Shows, “AI and The Human Connection, with Nathalie Doremieux,” presents a clear, experience-based argument for using AI as an amplifier of human expertise rather than a substitute for it. By grounding tools like Valora in real methods, stories, and client contexts, and by focusing on specific problems such as consistency, implementation, and support, entrepreneurs can leverage AI to deepen connection and results instead of eroding them. For listeners who want to hear the full nuance, examples, and back-and-forth between host Melita Campbell and guest Nathalie Doremieux, the complete conversation and resources are available on the original episode page at https://saas.podcastleadflow.com/p/ecq8nc2b.

Key Takeaways

Key Definitions

AI twin
AI twin is a term used in Nathalie Guest Shows’ episode “AI and The Human Connection, with Nathalie Doremieux” to describe an AI tool like Valora that is trained on a specific expert’s content and methods so it can mimic their coaching style and deliver personalized guidance without replacing the human coach.
Learning–doing cycle
Learning–doing cycle, as described by Nathalie Doremieux in the episode transcript of Nathalie Guest Shows, is an educational pattern where short segments of learning content are immediately followed by guided implementation steps to keep learners in action rather than passive consumption.
First-level AI support
First-level AI support, in the context of this podcast episode, refers to an AI assistant embedded in a program or membership that answers common questions, surfaces relevant resources, and directs clients to live support without replacing high-touch human interaction.
Custom GPT
Custom GPT, as discussed by Melita Campbell and Nathalie Doremieux, is a customized version of a generative AI model configured with specific instructions and reference materials so it can perform tailored tasks such as supporting a particular course or coaching program.
AI-powered knowledge bank
AI-powered knowledge bank is the term implied in the episode when Doremieux describes using AI to turn long group coaching recordings into a searchable repository where users can query specific topics and jump directly to the relevant parts of each call.

Claims & Evidence

Claim

AI-generated content often feels flat and disconnected from audiences unless it is carefully trained on a specific expert’s voice and experiences.

Evidence

In the episode, Nathalie Doremieux recounts early attempts to have AI write content for her using simple prompts, noting that while the results were easy to read and full of emojis, “it’s not me,” and people did not react to it, which led her to rethink AI as an amplifier that must be fed with her stories, methods, and client knowledge.

Source: Episode transcript - full_transcript - Nathalie Guest Shows - "AI and The Human Connection, with Nathalie Doremieux" - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025
Claim

Turning group coaching call recordings into an AI-searchable knowledge bank can unlock significant hidden value for members who will not rewatch long replays.

Evidence

Doremieux describes clients who had “all these recordings” from group coaching calls that nobody would listen to, prompting her husband Olivier to propose using AI so participants could search questions and jump directly to relevant segments, effectively converting calls into a reusable knowledge repository.

Source: Episode transcript - full_transcript - Nathalie Guest Shows - "AI and The Human Connection, with Nathalie Doremieux" - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025
Claim

AI assistants embedded in online programs can provide 24/7 first-level support by knowing the curriculum, support systems, and call schedule.

Evidence

In discussing e-learning, Doremieux explains that one of the simplest and most powerful AI applications is adding an assistant that understands your program, your support system, and “knows when the calls are happening,” so members can get immediate guidance between live sessions without replacing existing calls.

Source: Episode transcript - full_transcript - Nathalie Guest Shows - "AI and The Human Connection, with Nathalie Doremieux" - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025
Claim

Entrepreneurs who ignore AI altogether are likely to face a competitive disadvantage over time.

Evidence

Near the close of the episode, Doremieux states plainly that people who “are trying to put some blinds and ignore AI are going to eventually be at a disadvantage,” because others are already using it to gain leverage in consistency, support, and implementation.

Source: Episode transcript - full_transcript - Nathalie Guest Shows - "AI and The Human Connection, with Nathalie Doremieux" - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025
Claim

Large bundles of generic AI prompts are unlikely to deliver long-term value compared to focused tools that solve specific problems.

Evidence

When asked about selling AI-based products, Doremieux criticizes offers such as “download my thousand prompts,” saying she personally runs away from them and sees them as quantity over quality, arguing instead for targeted assistants (like a newsletter-writing tool) that are built around real user problems and protected access.

Source: Episode transcript - full_transcript - Nathalie Guest Shows - "AI and The Human Connection, with Nathalie Doremieux" - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025 - Nathalie Guest Shows / "AI and The Human Connection, with Nathalie Doremieux" / published October 29, 2025

Key Questions Answered

How does Nathalie Doremieux recommend using AI without losing human connection?

In Nathalie Guest Shows’ episode “AI and The Human Connection, with Nathalie Doremieux,” Doremieux advises using AI as an amplifier and accelerator of your existing expertise, not as a replacement for your human presence. She suggests feeding AI detailed information about your methods, values, and ideal clients so it can offer personalized guidance, and positioning tools like her Valora AI twin as extra support layers that help clients implement faster while leaving core coaching, relationships, and decision-making to real humans.

What is Valora, the AI twin mentioned on Nathalie Guest Shows?

Valora is an AI coach tool described in the episode “AI and The Human Connection, with Nathalie Doremieux” as host Melita Campbell’s AI twin, built by Nathalie and Olivier Doremieux. Trained on Campbell’s content and coaching approach, Valora lives on the episode’s show notes page at melitacampbell.com/podcast and asks listeners a few questions before turning insights from the episode into tailored, practical next steps specific to each user’s business challenges.

How can AI help members of an online course take more action instead of just consuming content?

According to Nathalie Doremieux in this podcast episode, AI can transform online programs from passive video watching into a “learning, doing, learning, doing” experience by offering just-in-time support and guidance. She recommends embedding AI assistants that understand the curriculum and support systems so they can answer questions, point members to the right resources, and guide them through concrete steps at the moment they are ready to implement, leading to faster results and more confident clients.

What are common fears entrepreneurs have about AI, and how does this episode suggest overcoming them?

In the episode, host Melita Campbell and guest Nathalie Doremieux identify fears such as poor output quality, clients thinking AI is a low-quality replacement, technical overwhelm, and concerns about data safety with tools like custom GPTs. Doremieux recommends addressing these by starting from a single business problem, educating AI with your own content and examples, being transparent with clients that AI is an added support layer, and seeking help to handle the technical setup so you can focus on your zone of brilliance.

How do you train ChatGPT to sound more like you, according to Nathalie Doremieux?

Nathalie Doremieux explains on Nathalie Guest Shows that you need to treat ChatGPT like a collaborator by feeding it detailed examples of content that truly represents your voice and telling it about your ideal clients, offers, and values. She further recommends asking the AI to criticize its own drafts, instructing it not to sugarcoat feedback, and then having it rewrite based on those critiques, repeating this loop inside a persistent space such as a project so the model gradually learns your tone and preferences.

Can you turn AI tools into paid products like Nathalie Doremieux did?

The episode describes how Nathalie and Olivier Doremieux first built an AI tool to make group coaching call replays searchable and later offered it more broadly, but it also highlights the marketing and technical challenges of selling AI products. Doremieux notes that truly viable AI products must solve specific problems, like a newsletter-writing assistant, and be hosted in a way that protects intellectual property and controls user access, rather than relying on easily shared custom GPTs that you cannot revoke or track.

What is a good first step to start using AI in my coaching business?

In this conversation, Doremieux suggests beginning by identifying a single problem you want to solve—such as frequent client questions, a recurring implementation bottleneck, or your own inconsistency with content—and then exploring AI as a targeted solution. She recommends experimenting with an AI assistant that knows your program and support system or using ChatGPT as a project planner that creates realistic, time-bound task lists based on your actual habits and constraints, instead of trying to master every AI feature at once.

How does Nathalie Doremieux handle data safety concerns with AI tools?

While the episode does not dive into technical security details, Doremieux acknowledges that many entrepreneurs worry about who can see uploaded files, whether providers like OpenAI use their data to train models, and how to protect client information and IP. Her practical advice is to be cautious about what you upload to public tools like custom GPTs, to favor architectures where you control access when building monetized assistants, and to get expert help if you are unsure how to align AI tooling with your privacy standards.

Why does Nathalie Doremieux think ignoring AI is risky for business owners?

Doremieux states in the episode that people who “put some blinds and ignore AI” will likely end up at a disadvantage because others are already leveraging it for consistency, leverage, and client support. She insists that engaging with AI does not mean automating everything; instead, she urges business owners to choose one aligned, low-risk use case and experiment, so they benefit from the technology while still centering human connection and expertise.

What is the Membership Lab and how is it related to AI in this episode?

The Membership Lab, co-founded by Nathalie and Olivier Doremieux, is the business behind many online membership and learning platforms referenced in the episode, and it is also where much of their AI experimentation takes place. In “AI and The Human Connection, with Nathalie Doremieux,” she explains that The Membership Lab’s mission includes solving problems for membership owners, and that this led them to develop AI-powered tools such as searchable coaching call archives and podcast assistants like Valora.

Full Episode Transcript
AI is not a good writer. We have to define what good is, right? But if it's something that connects with people, no. It's easy to read, it's pleasant, but it's not you, right? And that's what people want now more than ever with everything that's happening with AI. Hello and welcome to the Art of Value Whispering podcast. I'm your host, Melita Campbell, business coach, value whisperer and advocate for introverted female entrepreneurs and experts who want to build a business that feels natural, authentic and impactful. I'm joined by Nathalie Doremiu, co-founder of the Membership Lab and a mastermind with her husband Olivier behind Valora, my AI twin that you've hopefully been enjoying alongside the Business Book Festival episodes. In this conversation, Natalie and I explore how AI can be used to amplify your brilliance, save time and help your clients take fast action without ever replacing the human connection that lies at the heart of your business. We also look at some of the common fears around AI, how to overcome those and some simple first steps you can take to start using AI more comfortably and effectively in your own work. As you would completely expect from this episode, you can head to melitacampbell.com slash podcast to access the free AI coach tool created just for this episode. Go to that podcast page, click on Natalie's episode and Valora, my AI twin, will be there waiting to help you turn the insights from this episode into simple, practical next steps you can take to start using AI to grow your business and your impact. Meantime, let's dive in to this fascinating conversation with Natalie. Hi, Natalie. It's great to have you with us on the show today. Thank you so much for having me, Melita. I'm excited about this conversation. Yeah, me too. And firstly, a huge thank you from me and the listeners for Valora, my AI twin that has been on all the book festival episodes. That was a product that you have developed and you didn't just develop it and stick it on the podcast you actually tested it for every single episode you really went all out to make sure that it was as useful as possible so I am really grateful for that it's an incredible tool if anyone listening hasn't tried it yet what are you doing wait till the end of this episode but go and try it out it really is phenomenal the advice it gives is like wow every time so a big thank you firstly for creating that we will get to how you created that and the idea behind it and your podcast lead for products in a bit but first I'd love to know how did you get to become this AI expert that you are today? Yeah so so first I'll just say it's not just me it's me and my husband and business partner of 20 years I should say that because he's really the one that has the ideas around AI you know we've been working with AI for six or seven years now, you know, back when we didn't have all the tools at our disposal right now So you really had to be a developer in order to attempt to create anything with AI because AI has been around for a long time right But like available to business owners like us not for so long So in our business, you know, we run the Membership Lab and one of our values, you know, and one of our mission is really to be problem solvers. That's what people tell about us. It's like, we love to solve problems. So when AI came in, we knew because it's been around for so long, some people have been using it for so long. We know it's here to stay and it's going to keep growing. We don't know where it is going. But what we know is there is a unique opportunity right now to use it to help us solve problems. And I think that this is really the way we want to look at AI because otherwise it's too overwhelming, right? There are so many tools that get created every day. It's a full-time job to test everything that's out there. And for all the ones that get created, there are some that just get canned, right? Because they become irrelevant. And it's overwhelming. It is very overwhelming. So we are thinking about it differently. We're like, what is one problem that we could automate that doesn't require us as a human be more consistent of right or save time and those are about the three things right that we look at so we've had some fails you know we can talk about that you know with ai but we're experimenting everything is an experiment right the first idea with ai came i think seven years ago from olivier and it was people who were saying i have all this recording it was a group coaching. I have all these recordings. I don't know if I want to put them on the platform because nobody listens to them. And my husband was like, Olivier was like, but there is some gold in there. Maybe somebody didn't come to the call, but would benefit from that question. Or maybe the person wants to go back and re-listen to an answer to a question that they ask, but nobody's going to listen to an hour or two hours? Can we create something so that these calls become a bank of knowledge that we can tap into? And that's really how it got started, but created more like custom for individual client before it became like a product. So it started quite a while ago. Yeah. And the reason they came to you with that question was because you're the team behind many of the big membership platforms that probably most listeners have joined um we've all joined a membership not all of them but yeah quite a few of them yeah so a lot of these big membership platforms you develop the platform and how that all works for people so you've already got that knowledge of what was working and then as you say what's the next problem how could this work better save people more time you know add more value so i love that approach so you helped the first person to use ai to turn all their coaching videos into a knowledge bank which just sounds incredible what happened next you mentioned that you had some failures along the way what was that kind of journey for you like as someone that has this understanding of how tech and ai works already I think it quite reassuring to know that you didn always get it right as well But what went right What did you learn What have you learned about the limits of AI then perhaps Yeah absolutely So in terms of that tool that we created for that person, we immediately turned it into a product, right? Because we were like, this is something we can actually do and we can actually sell. So one of the, I don't want to call it failure, but like one of the challenges that we've had in business is that people come to us, right? There is this circle of people and they refer us and things like that. We are not doing so well in the marketing and sales space. Let's put it this way. So when you get a new product like this, and that's been a recurring problem, and it's happening with podcasting actually, is that it's very new. So it is not even something that people are looking for. And that's the problem when you're the precursor and you do things before it becomes the norm is that you have to educate people that even exist, right? So that's always been a challenge. So that took like several years. And basically, we would introduce the product as an option to people that we were working with. So we were not selling it individually, because like that would require a whole marketing strategy, right? We were just focusing on the memberships and the online portals we were creating and just selling it to them as an add-on if they wanted it. Now, in terms of our AI experience, you know, when Chatipiti came in and things like that, just like everybody, we tried it. Can you write content for us, right? I am much better now, but like I was very inconsistent in creating content. I didn't know what to talk about. I felt that my style was flat and boring. I'm all like bullet points, tell it like it is, there is no story. So when you start to use AI and you feed it a couple of ideas and I mean, what it writes, it's very easy to read, right? It's very pleasing. So you're like, oh, this is amazing. Oh, and it's throwing some emojis and stuff like that. That's cool. But then you eventually realize it's flat. People don't react to it. I don't even use emojis, right? So it's like, it's not me. So it's pleasing to read, but emotionally it does nothing and it's not connecting with people. So when I realized that, we really went back and trying to understand, okay, this is not how we want to use AI anyway, right? What we want to do is we want to use AI as an amplifier. That means that if I want AI to produce something that I can work with, I need to educate it about our product, about the way I think, about my stories, about our offers, about our ideal client and all these things, right? And that's really been the key in the recurring pattern, if you will, where AI to us is an amplifier, an accelerator. It's not a creator. AI is not a good writer. Like we have to define what good is, right? But like, if it's something that connects with people, no, it's easy to read, you know, it's pleasant, but it's not you. Right. And that what people want now more than ever you know with everything that happening with AI So that really been the realization Let go back to what we wanted to use AI for Personalize experience, you know, like searching videos and things like that and showing you exactly where that thing is mentioned in the video. That's what we were doing, right? Creating custom tools for people. personalization like you know valora you know it creates something unique for people based on the context that's where we think the full power of ai is for us you know business owners while still keeping the human in the loop yeah so taking what we've already created and i say like giving new ways for people to access that information and use it like valora it takes existing information but breaks that down into practical steps, applying that to your specific business challenge. So it wasn't creating anything new. It was just being that kind of not quite an interpreter, but you know, the facilitator in between. Exactly. Yeah, exactly. It's only one piece of the puzzle. It is only ever going to be as good as what you feed it. And you feed it two things. You feed it knowledge. So that's you, your method, your unique way of doing things because you don't want a GPT to go out and look at all the other coaches and the way they do things, right? You want the Medita way. Then you want your listener who has a unique situation, is in a unique position, right? He has a current problem or a struggle, you know, something that they're working on. And AI is going to put these two together and give them something. It's like the listener is sitting next to you and you're asking them a couple of questions And based on that, you give them a piece of coaching, right? That's the equivalent. That's the power of AI of being able to analyze, you know, take your knowledge, take the context and come up with something that's very unique and very helpful for the listener. Yeah, I love that. So let's look at what is holding people back from AI and going all in, because I think everybody, certainly in the entrepreneurial space is all very focused on how can I use AI, but there's also some fears around that as well. So what are the common fears that you've been experiencing and how can we help the listeners to feel a little bit more comfortable getting out there and as you say, experimenting with this and seeing how it can be helping amplify their work and their results with their clients? So I think the first one is, I would say the lack of quality into the result. And that comes from several things, right? We talked about the knowledge, you know, and what you feed it. But there is also the way you are asking. Most people still use AI like Google, right? So that is an issue. So people are worried about quality because they feel it doesn't sound like them, but they are not exactly sure what to do about it, right? Fear that their clients might think, oh my gosh, she's trying to replace herself with AI and that's going to be lower quality. When this is never the case, we are adding AI to what we're already doing. We are not replacing ourselves. You know, with AI, it's an add-on, right? Another way to help people. So I think like those are like the two biggest and probably another one would be the overwhelm. The technical aspect of it, you know, being quite good at prompting is like, now how do I use this? Because is my data safe, right? And if you create custom GPTs and you upload files, who can see these files, right? Who has access? Is OpenAI using this to feed their database? And they are, right? So there is also this fear of like protecting information, protecting client information, protecting their IP, right? Protecting your IP. so when you've got all these things there is so much resistance in every step that if you attempt to like do it on your own and you're not this kind of person that really likes to dig and things like that people are going to give up right they'll say like this is not for me it's not even good quality it doesn't sound like me that sounds complicated forget it I'm just going to stay where I am. Yeah, well, that I think that was me, you know, like towards the end of last year, and everyone was using chat GPT. So I started using it, but I didn't really know what to do. And so I was putting in like very basic rubbish prompts. And so I was getting very basic rubbish answers. And then I was like, Oh, you know, it's just too much to learn. It's all complicated tech. But then I had someone show me how to use it and how to create custom gpts and how to educate it properly as you were saying and now it's amazing and now all i can see is possibility it's like oh i could do this and i'm currently working to turn my program into a series of custom gpts so that my clients can get through faster because one of the big problems i see with my clients at least and i suspect a lot of the listeners because it's one of the issues for introverts we like to think that also means we can get stuck in overthinking. And I think chat can really help overcome that and be there to help people through those moments. So I'm really excited to dig in and see where it goes next. But you see a lot more than I do. I'm just seeing the tip of the iceberg. Obviously, you said earlier, we just don't know where this is going to go in the future. Where do you think this is going to go in the near future? Have you got any ideas that you're excited about? So I think like where this is going, so I can tell you like in the context of the e-learning space, you know, membership online programs, group coaching programs, like what we are doing is we are using AI as a way to get people into more action, right? Because before we, what were we doing? We're doing well, let's do more calls, right? So that we do implement together, like to get people into action. And now we are moving from you know these videos where people are passive you know they are learning learning learning and then comes time to do like the work And then some people would do it some people would get stuck And now we getting into learning doing learning doing learning doing which is getting people into action right instead of all this learning mode And that going to be a good thing because that shows people that they can make even tiny steps They can make steps So it builds their confidence. Everybody's going to win at the end. They're going to get results faster. You're going to get happier clients. Everybody wins, right? So when AI is going to be used this way, and that's why custom GPTs are such a great tool that we have at our disposal, right? that gets your program automatically more actionable yeah okay faster I think you get and we were talking with Fabienne Fredrickson about leverage and how do you go more and it was really the shift in mindset and I think that AI can really help be a catalyst behind thinking bigger and achieving more without the burnout and being that little bit braver because we kind of avoid you know she made the point that um when we work in isolation as so many particularly women do that that leaves a lot of scope for overthinking but self-doubt and that can hold us back and i think chat can really help us through those moments and do more action so that there's no space for fear and all of that that can typically stand in the way so yeah excited absolutely i mean what we're seeing you know and by the way that's where i hope ai is going in the e-learning space right because the tools are there and we are seeing people using it like even the first level to anyone who has an online program and is wondering what can i do well you can add an extra level of support with an AI assistant that knows about your program, knows about your support system, knows when the calls are happening. And right there, you have 24-7 support that can give your members, your clients, this first level of support. It's not replacing anything that you have right now, but it's an extra level, right? So that's like one way you can get one foot into AI. I think it's more looking into it this way versus to figure out, okay, what kind of custom GPT can I write? Because that's overwhelming. Okay? Like, it's like, oh, what can I write? What file do I need? How do I write the prompt? That's the how, right? So right now you want to figure out what do I want to create and why? What is it going to give me my business or my clients and members? And then you get help to figure it out if you need to. You can imagine that Fabienne is not writing custom GPTs, right? That's really not something that she ever wants to do. That's one of the things you learn with leverage. Let's stay in your zone of brilliance, where you shine, where you all have all the energy, and then you let go or hire out the rest. Yeah. Yeah, exactly. So I think, as you say, AI is just going to give us more scope to stay in that zone of genius and find the right way to have the rest of your business support and working smoothly as well yeah i think there is really an opportunity to use ai to amplify our unique way yes right and some people are using i sure you see these on linkedin like uh what is it like create your course completely with AI in like X minutes How do you think that going to fly right What are the chances that it is my knowledge The knowledge has to come from us, right? Or a podcast. I don't know how many episodes per month they generate, but it's huge. And it's all generated by AI. I have no interest in listening because there is no human behind. So I'm like, that doesn't really speak to me. So all of this is on volume. We have an opportunity to keep the quality, keep it unique, and us making it consistent through the automations and getting time back, right? Like that's what leverages us as well. It's like, right, how do we still get people the result without it being tied to our time? If you don't want to add more calls, right, more calls to your calendar and things like that. but can you use AI to create tools to help them, especially if you keep repeating yourself all the time? Yeah, yeah, exactly. And as you said, putting on more calls is great, but that doesn't necessarily equal more support because it's overwhelming. I mean, who can go to a call every single day? So having a tool that can support alongside those live sessions, yeah, would probably offer more value. Yeah, absolutely. So do you have any sort of recommended steps for people on how to start using AI more within their business and get really comfortable with using it to sort of amplify and add value and do that in the correct way? Yeah. So I think it really depends where you're at with AI. Also, your level of like, are you alone or do you have a team? Right. But also, I think it's important to also look at your audience. there are audiences that might not be ready or want to you know get into ai you know like i'm thinking maybe older people you know that are still very like old-fashioned type of calls and things like that and if you start giving them custom gpt's and things like that how is this going to fly right so we have to kind of like take that into account but in terms of like how do I use where do I start I would say look at what it is that you want to change in your business do you want your clients to get results faster do you see people getting stuck somewhere like there is a step and like every time they come back on the call they still have more questions they still haven't sent the email you know that could be an indicator okay can I create something here. So if you look at it as a problem solution, right? Maybe you think I'm not consistent in creating content. I can create content, but like I'm not consistent. Okay. Can I have AI to support me in being consistent? And what does that look like? And by the way, AI is great at coming up with plans. You can literally tell it, okay, I am inconsistent. I'm going to do something for like two days and then I'm going to give up. So you need to break this down for me. This is how much time I can allocate every day. And just tell me, you know, what document I need to create, what it looks like. And it will go as far as creating the spreadsheet for you right But you really need to tell it the truth of like it overwhelming to me I never know what to talk about so it is going to create tasks for you up front on what you could be talking about so that when you need to create the content you don't have to say oh my gosh I'm supposed to record a video today but I don't know about what the work has been done before so it's it's a great project planner for that so that's one thing that you can do like be more consistent so another thing that you can maybe look at is if your team or you are constantly being hit by questions like, where is this? How do I find out? Blah, blah, blah. Can you create an assistant that can help them? Especially if it's always the same question over and over and over, right? So I think you need to look at it this way. Like, is there a problem? And can I find a solution? Or is there something I want to improve? Do I want to save time? Do I want to be more consistent? Do I want to help my clients? which two different ways, right? AI for my business, for myself internally, and then AI as a tool to help my clients, my members. Yeah, yeah, great. I love those two ways of thinking and it's gonna, yeah, makes all the difference. I wanna come back to two things you said. So one was about this fear of the quality won't be good enough, but what can we do to help improve the quality of the results that we create? Is it all about educating it first? How do we go about that? What's your views? Yeah, so I mean, we talked about custom GPT. So for internally for you, projects is really like the next level from custom GPT. So using projects is great because it actually remembers the previous conversation. So the more you use it, the more it is going to know about you, you know, from the questions that you're asking, like you said you need to feed it information so you need to tell it who your ideal client is what are your offers you need to tell it like what are your values what do you stand for like what makes you tick give it examples it loves examples that's the best way for it to really analyze and provide better output and I think one big thing that most people don't do is you want to challenge it because you've probably noticed AI is very nice. If you ask it something, it will always find something good to say about it. And you say, you know, I want you to not sugarcoat it, like tell me like it is, right? And when it says something, I often ask it, okay, now I want to criticize what you just created. And if you do that, you'll see it's going to justify what it side. So it's going to say, okay, so I did this because this, this, this, this, and here are the places where I could improve this, this, this. Do you want me to do that? Yes. And you'll see that the second thing that it creates is already so much better than the first one, right? So it literally, if you think of it as really having a conversation with one of your team members and you have a conversation and you'd say, okay, are you sure, Natalie? Like, can we make it a little better or is that it and then it's always going to try to find something and criticize its work and make it a bit better and you can say no that doesn't sound like me And you can add it more example. Here is an example of a post or an article or a transcript of a video that I did that really resonated with people. And like, that's me. And I want you to do something like that. Right? Yeah. So, I mean, it's a muscle, right? Great advice. I love that. Thank you. The other thing that I wanted to come back to was you said right back in your early days of developing something with AI. and then you thought, oh, we could package this as a product for someone else. Do you see this as being, once we can really understand how to use AI to accelerate results, maybe for ourselves as well as for our clients, that we can turn this into an additional income stream by then packaging and selling some of what we develop to other people? So, I mean, you see this already. A lot of people are selling prompts and things like that, right? So like when I see download my thousand prompts, like I run away. Like I don't need a thousand prompts and I don't know which one like no. Right. So this is like quantity and this is for people that are like they are scared of missing out and they don't know what they are looking for. So therefore, anything that is free with AI, I'm just going to take. I'll download it. Never open the PDF, but I'll download it. Right. So and there are a lot of people that sell this type of things. Right. So I think it's working now. it's not going to work for too long. I don't think so. However, what can work is having access to an assistant that is loaded with your program that is more like a self-study, but it's not going to be a custom GPT. As of right now, we cannot monetize custom GPTs, right? Because they are outside You don control who has access The client gives it to a friend The client leaves your program they still have access their friend still has access right So if you want to turn this into a product that you can sell it really needs to solve a problem for people. And that could be, you know, an AI tool that helps you write your newsletter, for example, a newsletter writer, right? But it would have to be like protected, right? In a way where you control who has access if you want to monetize it, right? So the person would be able to upload all their previous newsletters, you know, and everything about them and their offer and their audience and how people like to read the newsletter and things like that. Right. And then there would be some knowledge around how to write a newsletter, you know, that gets read and that gets engagement, blah, blah, blah. And then it would come up with like drafts and ideas, right, that they could take on to create. So that would be one way, but it cannot be a custom GPT. I don't know if people sell. I think people might sell custom GPTs. I don't think that that's necessarily a good long-term viral idea. Yeah, as you say, you might be able to sell it, but then you can't protect it. Yeah, exactly. Yeah, fantastic. Nathalie, I really enjoyed this conversation. Thank you so much. You've been so open. Is there anything you feel we haven't touched on that would help the listeners in their AI journey from here on in? Well, I think that the people, and I'm just being honest here, like the people that are trying to put some blinds and ignore AI are going to eventually be at a disadvantage. Right. But there are different levels of using AI. All right. Of course, like we've seen. So I would encourage anyone who's like scared and doesn't want to look at it to explore maybe one way that they could start using it and identify what it is You know like you know it a mindset thing like what it is scared of the tech worried that my clients would think I'm being lazy and I want to replace my calls with this and it's not good enough like what it is and surround yourself with people who are using AI be curious about what others are doing you know and just get one foot you know in there but that problem solution I think is a great way to approach it and just do one thing I would not like listen and download everything that is going on out there it's just overwhelming yeah there's so much and so many tools the new ones popping up all the while and I guess you could even ask chat gpt teach me how to use chat gpt absolutely Absolutely. I mean, you could say, I want to spend an hour a week learning about ChatGPT. So can you create a program for me? I want to learn how to prompt. I want to learn how to create images. You know, I want to learn how to create my own custom GPT, things like that. And it's going to come up with the program, right? So you can literally ask it almost anything like that. it's just that when you ask it to think and come up with ideas just know that it needs to know about what's in here if you want it to resonate you know like you yeah and in here she was pointing to her head for anyone that's not watching the video yes sorry yes yeah so you know what I'm going to do is I'm going to create a podcast lead flow form for this episode yes great this is going to help people figure out what is their next step why don't we do that sounds sounds amazing yeah then they can really see test it out for themselves and exactly test it out yeah yeah and it been so good with the the book festival episodes helping people really get some practical next steps that are doable with their resources but also encouraging and stretching them a tiny bit As I say I been really impressed So I excited to test out the one for this episode as well. Yeah, me too. So Natalie, if people want to learn more about your work, where's the best place for them to find you? Yeah, so people can go to the membership lab.com or they can go to podcastleadflow.com as well. And I am also on LinkedIn. I'm on LinkedIn every day. If you want to connect with me and have a chat, then we can do that there too. Fantastic. Thank you. And we'll put all the links in the show notes page, which you can find at melitacampbell.com slash podcast. Brilliant. Natalie, again, thank you so much for your time. I've loved this conversation. Me too. Thank you so much, Melita. Thank you for joining us today. I hope you have enjoyed this conversation with Natalie and that you're walking away with some fresh ideas about how you can be using AI to amplify your impact, streamline your work and better support your clients. If you did enjoy this discussion, I'd love to hear what resonated with you most. Come and let me know over on LinkedIn. You'll find me there at Melita, that's Melita with two Ts. And do share this episode with another big hearted business owner who might be hesitating when it comes to exploring AI. And don't forget to help you turn your newfound insights into practical actions, Valora is waiting there for you on the show notes page, which you will find at melitacampbell.com slash podcast. You answer three short questions and in minutes, she will share some practical insights that you can take based on everything shared on today's episode and specific to your business challenges. Meantime, I'm Melita Campbell, your business coach and value whisperer, and I look forward to seeing you again in the next episode next Wednesday. Bye. you

Powered by Podcast LeadFlow