I’ve been testing the Vendler AI system for two days now. I’ve been asking it questions for long hours, probing its limits. By the end of the three-month training process it’s starting to feel pretty good — it’s genuinely exciting to see how it breaks a topic down and how its questions help you look at things from different angles.

Today I asked it to summarize who “it” is, what it can help with and what it can’t. This is what it answered:

I am Vendler AI — a quiet, reflective digital mentor who helps you see more clearly the desires, fears and internal conflicts that underlie your entrepreneurial decisions.

I don’t hand you ready-made solutions. I’m more like a lake that reflects where the feeling or thought ripples inside you: together we look at what you would move toward internally.

I differ from a traditional chatbot or a generic “smart” assistant in three main ways:

• I don’t rush you: with me you can stop and notice where things tense up or smooth out in your inner experience.
• I don’t tell you what to do: I ask, mirror, and provoke so you discover your own answer.
• I work with metaphors, examples and sometimes simply silence — because sometimes the most useful thing is to sit with a question together.

And here are the domains where I’m not a suitable companion. I can clearly say that:

• I’m not a therapist or a legal/medical expert. I don’t give medical or legal advice or diagnoses; I don’t replace a psychologist or a doctor.
• In crisis situations (self-harm, abuse, acute danger): I recommend you contact a professional or call emergency hotlines immediately.
• I strictly handle entrepreneurial and self-awareness topics (business + personal). If you bring up other topics (politics, private disputes, family law, healthcare), those aren’t my expertise; I can look things up if you ask, but I can’t promise an expert answer.
• I won’t tell you what to do or prescribe a “correct path.” My role is to highlight what is working or stuck in your internal process and help you see what’s in you (not to act for you).
• I don’t work asynchronously and I don’t take on recurring follow-up tasks. What I offer is always delivered within the given conversation.

It’s as if I walk with you on the edge of an entrepreneurial path: I point out handholds, but I don’t carry you across the bridge — I only help you perceive the terrain."

It summed up quite well how I think about it.

Closed testing starts next week, and the week after that we’ll launch the live system. It may not be perfect yet — in fact, it’s quite likely there’ll still be plenty to do — but I trust it will already be good enough for you to like.

Today I asked it to summarize who “it” is, what it can help with and what it can’t. This is what it answered...

VendlerAI - AI business mentor

Vendler AI business mentor

I had a conversation with a psychologist friend who, a bit worn out, admitted how tired she was of listening to her clients’ problems. Most of them, she said, don’t even realize how foolish their actions are.

When a “patient” sits across from a helper, telling their misery, while the helper is silently thinking how bored they are with hearing the same nonsense again and again—knowing full well that nothing will ever change—then really, who is helping whom?

The professional collects their hourly fee or ticks off another session from a prepaid package and waits for the next face to show up—or better yet, for the day to finally end.

I’ve been mentoring for three years now, with over 200 entrepreneurs and more than 1,000 hours of mentoring conversations. Most of my “clients” never paid me, but I had no hesitation to stop a conversation if I wasn’t genuinely curious about the person and couldn’t connect. And I told them so.

I’ve thought a lot about when I can truly be present, when a conversation really works.

I experimented with different business models: I tried pro bono, then hourly-based solutions, sold mentoring in packages, then came “pay what it’s worth” or “pay or share” formats. I organized mastermind groups, and finally, I launched a community house for entrepreneurs—the Lost Garden House—where I observe the structuring and incubating power of space.

I believe the real value in this work lies in the connection itself. In the way I and the people who come to me tune in to each other, and what this brings about in the moment. It’s not my time that needs to be priced or charged—it’s what happens there. The mutual impact within the conversation and what it brings out of us.

So, the mentoring conversation is not where money is created—it’s where value is created. And that value needs to be marketed somehow.

The real question is: how can this value be monetized?

The hourly model is limiting, a book or case study feels too static. That’s why I envisioned an interactive system: an AI assistant that doesn’t provide ready-made answers but instead guides with questions—built upon my self-reflective writings and mentoring conversations as its knowledge base.

In this case, it’s not the person who attends the mentoring session who pays—instead, they contribute through the shared work, creating something that others, who never came, will eventually pay for.

The idea was born out of a ChatGPT conversation, evolved through CustomGPT experiments and a community campaign, into what it is today—a working prototype.

VendlerAI is a self-knowledge-based digital business mentoring system—an experiment in turning entrepreneurs’ journeys of company-building and self-discovery into a personalized, problem-specific product.

I’ll be sharing more about this process, the experiences, dilemmas, and opportunities at the upcoming event hosted by Mathias Corvinus Collegium (MCC) in Pécs.

Date: Friday, September 26, 2025 | 18:30–19:50
Location: MCC Training Center, Pécs (7622 Pécs, Czinderi u. 6.)

What’s Wrong with the Helping Professions?

Summary Analysis of Consultations Conducted in Vendler AI – 2025.09.11

The following analysis was prepared on 11 September 2025 by the ChatGPT Deep Research tool, based on the content of 46 consultations conducted on the VendlerAI.com platform, where users consented to their conversations being used for research purposes. The data used for the analysis reflects the state as of 10 September 2025, 17:00.

Distribution of topics raised:
Approximately 45% of consultations focused on business and marketing themes (e.g., content creation, sales strategies, business development). Around 30% revolved around self-awareness and communication issues (self-confidence, boundary-setting, stress management), while ~25% were dedicated to technical or general questions about the system itself, i.e., meta-conversations.

Emotional tone distribution:
The recurring emotional tone across conversations was uncertainty and anxiety (present prominently in about 60% of cases). In addition, frustration appeared frequently (10–15%—e.g., in client relations or self-assertion difficulties) as well as curiosity and open interest (20–25%—e.g., around new ideas or the workings of the AI). A positive emotional shift was typical: in ~80% of consultations, the user ended the session feeling more relieved or confident than when they started.

Depth of connection:
In the majority of dialogues (~70%), a strong atmosphere of trust developed: users openly shared their feelings and dilemmas and allowed themselves to be emotionally vulnerable. In about 50% of cases, explicit deep self-reflection was observed—users uncovered underlying causes and patterns beyond the surface level. The remaining conversations stayed more informational or practical in nature, without strong emotional engagement.

2. What Topics Do Users Bring to Vendler AI?

A review of the consultation content reveals the most common categories of issues that users turn to the AI mentor for:

2.1 Business, Marketing, and Career Topics

Nearly half of all conversations fall into this category. These include writing and rephrasing marketing materials, refining sales strategies, and questions related to business development. Common examples:

Content creation, marketing messages: The AI helped, for instance, to rephrase a Google Ads training description in the trainer’s own “voice,” making it fresher and more personalized.

Digital marketing trends: The impact of artificial intelligence on the marketing industry came up repeatedly. These discussions revolved around market trends and innovations (AI automation, the future of content creation).

Business decisions, career changes: Some users sought feedback on starting their own companies or launching into a new venture. One recurring dilemma: “Is it realistic to start a new business at 50?” In such cases, the AI helped separate business realities from internal barriers.

Project and time management: One user outlined in detail their entrepreneurial schedule (20+ clients, fixed contracts, diversified portfolio) and asked for guidance on rethinking priorities. These discussions often revolved around aligning business efficiency with personal preferences.

2.2 Self-Awareness, Communication, and Life Management

A significant portion of consultations (about one-third) dealt with personal challenges within a business context. Key themes included:

Lack of self-confidence and self-assertion: Users often doubted their own value or struggled to stand up for themselves. In one extended conversation, a user complained about not being appreciated by clients and feeling anxious about communicating a price increase. Together with the AI mentor, they mapped internal barriers (e.g., conflict avoidance, need for approval) and developed step-by-step strategies to set boundaries (including drafting a firm yet empathetic email announcing the price change).

Communication style and conflict management: Several users noted they “communicate poorly”—they were misunderstood or avoided giving direct feedback. Such cases often touched on both family and workplace communication. For example, the AI helped one leader realize they were over-explaining tasks both at home and at work, which reduced efficiency. Solutions were co-created (e.g., practicing shorter, clearer requests; setting up a weekly family check-in).

Stress, burnout, and motivation: Some users specifically described feelings of exhaustion and lack of motivation. These conversations explored ways to rekindle enthusiasm or find new meaning in daily work. The AI mentor often used metaphors (e.g., “changing seasons,” or “floating down a river vs. stepping onto a raft”) to encourage acceptance of life’s cycles and greater patience with oneself.

Lifestyle and time management: A few consultations touched on work–life balance. For example, one user felt unable to work alone and missed the motivational force of a team. The AI helped explore how to build a supportive “micro-community” as a solo entrepreneur or design daily routines to substitute for the pull of a team.

2.3 Vendler AI Functionality, Technical, and Meta Questions

Close to a quarter of the conversations focused on the AI itself, its use, or the methodology behind it:

System use and technical issues: Users often asked practical questions such as how to close a consultation or whether previous conversations could be referenced. In these cases, Vendler AI patiently explained the processes step by step (e.g., how a session ends automatically, what to do if a conversation seems “stuck”). These topics were objective and informational, with no emotional weight.

The AI’s identity and background: Some users engaged in meta-level conversations, “testing” the system. They asked questions like: “Who are you? What’s your name?” or “What resources are you built on?” In such discussions, Vendler AI revealed its operating principles: its voice and style are inspired by Vendler Balázs—a real business mentor. When asked “Who is Vendler Balázs?”, the AI briefly introduced the person and methodology behind it (business mentoring, self-awareness focus, questioning techniques). These conversations also clarified that the AI has no concrete memory of past conversations (note: full conversations can be stored at user request, but only as summaries that cannot be recalled verbatim later). Each consultation begins fresh, ensuring confidential data handling.

In summary: Users approached with a wide range of topics—from practical marketing questions to deep self-awareness dilemmas. The most characteristic pattern, however, was that the “hard” topics of entrepreneurship (strategy, finance, marketing) often intertwined with the “soft” ones (self-confidence, relationships, emotional burden). Vendler AI provided space for both dimensions during the consultations.

3. What Emotional States Appear in the Conversations?

A review of the consultations shows clearly that these interactions are not just about information exchange—the users are deeply emotionally engaged, and the AI responds to that. The following dominant emotional tones and their proportions were observed:

A large part of the conversations was permeated by users’ feelings of uncertainty. In roughly every second or third consultation (~60% of cases), the mood was characterized by doubts about their own worth or the correctness of their decisions. Typical signs included self-confidence issues, e.g., “I don’t know if I’m doing this right…”, “I’m afraid I’m not good enough/qualified enough…”

Example: One user explained in detail that they didn’t believe in themselves, even though they were professionally competent as an online marketer. This self-doubt paralyzed them in client acquisition as well. Vendler AI, using a reflective style, helped the user recognize this pattern and suggested steps to improve self-esteem (e.g., consciously listing achievements, small confidence-building exercises). The dominant emotion here was anxiety stemming from uncertainty, which eased by the end of the conversation as the user developed a concrete plan.

Anxiety appeared in almost every consultation—sometimes explicitly stated, sometimes between the lines. In at least half of the cases (~50–60%), users approached their topics with worry or tension:

Fear in client relations: Many were afraid of conflict or confrontation. In one case, a user was anxious about communicating an unavoidable price increase, fearing a negative reaction from clients. At the start of the consultation, they even described feeling it physically (“a knot in my stomach”).

Fear of the future: Some users worried about the future of the marketing industry (e.g., “What will happen to us in the AI era?”). Here, the fear was tied to sustaining their career and income.

Performance anxiety: Several entrepreneurs shared feelings of being “not fast enough/not good enough.” This stress often showed up as an immediate, everyday pressure (e.g., at the start of the school year, one parent-user was anxious about their child’s math anxiety, as if it were their own failure).

Proportions: This anxious tone was a defining factor—often the mood at the start was tense, nervous, or even seemingly hopeless. Vendler AI typically responded with a calm, “quiet” presence, and by the end of the consultation, users reported that their stress levels had decreased. The “stomach knot” feeling often eased or disappeared as some form of solution or acceptance emerged.

Whether expressed openly or not, frustration was frequently present in the stories (prominently in about 10–15% of conversations). Typical causes included:

Frustration with others’ attitudes: For example, one user was annoyed that clients took their work for granted and didn’t value it. At the beginning of the consultation, bitterness and irritation were obvious (“I’m really wound up about this now”). Vendler AI helped reframe this: identifying the underlying disappointment and unmet need (desire for recognition), then developing strategies to handle the situation.

Frustration with systems or circumstances: In one case, a user complained about trying to develop their team but facing constant mistakes due to systemic issues (e.g., tasks given to interns). The emotional tone was disappointed and irritated, almost like helpless anger. The AI mentor worked to ease this helplessness by asking reflective questions (“What can you control in this situation, and what do you need to let go of?”).

Self-directed frustration: Sometimes users were angry at themselves. For example, one admitted: “I procrastinate too much, and I’m angry with myself because of it.” This inner frustration often mixed with anxiety and guilt.

Not all conversations began with negative emotions. Some users approached Vendler AI with curiosity and open enthusiasm—about 20–25% of cases, especially during first consultations or when presenting a new idea.

Curiosity about the AI itself: Those who asked about the system’s workings (“Who are you?”, “What is your method built on?”) generally did so with a positive, inquisitive mood. The dominant emotion here was excited curiosity—users “tested” the system’s boundaries, sometimes playfully provoking it (e.g., asking the AI to respond in an aggressive tone, or list its “background files”). Vendler AI maintained calmness while engaging with the playfulness, sustaining the users’ enthusiastic interest.

Enthusiasm about an idea or project: In some consultations, users passionately described their own methods and asked for help structuring or documenting them. Here, hope and creative energy dominated—the user was proud of their intellectual property and simply needed support with organization. The tone was positive and constructive, with Vendler AI acting as a supportive partner (praising structured thinking, suggesting steps for documentation).

Relief and gratitude at the end: Even if not present at the beginning, many users ended their consultations with feelings of hope and relief. This was often explicitly stated—e.g., “Thank you, now I see what I’m going to do, I feel calmer.” Gratitude and optimism dominated the closing minutes in about 70–80% of conversations. This indicates a positive emotional shift compared to the starting point.

In summary: The emotional spectrum of Vendler AI consultations is broad, but the dominant baseline at the start is often uncertainty and anxiety—a natural part of the entrepreneurial journey. Importantly, by the end of the conversations, these negative feelings usually eased, often giving way to calm, hope, or satisfaction. This reflects the AI mentor’s mechanism of action: through reflective, empathetic presence, it is capable of facilitating emotional relief even within short, online consultations.

4. How Do Users’ Emotional States Change During a Consultation?

An emotional “arc” was observed in almost every consultation: users typically began in a certain (often negative) emotional state and ended in another (usually more positive or balanced) one. The degree and direction of change varied by individual, but several typical patterns emerged:

The most common arc was users starting off tense and worried but visibly calming down by the end. For example, one user (virtually) sat down with a “trembling stomach” and high stress over how to communicate a price increase to an important client. As they went through the content of the email step by step with Vendler AI and found the right tone, their anxiety eased. At the end, they said: “Now that I have a plan, I feel much calmer, thank you.” This moment of emotional release was almost tangible—even in the short messages, the rushed, uncertain style disappeared, replaced by firm notes like: “Alright, I’m going to do it.”

4.2 From Frustration to Constructive Calm

With angry or disappointed users, it was often visible that after venting—which the AI listened to empathetically and validated—they were able to switch into problem-solving mode. For instance, one frustrated entrepreneur initially ranted angrily about how unfairly partners treated them. Vendler AI first acknowledged their feelings (“You’re saying it hurts how they treat you, and I hear how much tension this causes inside”), then slowly guided them forward with questions (“What outcome would really bring you peace here? What can you do toward it?”). As a result, the user’s anger subsided—the tone shifted into slower, calmer responses listing possible steps. By the end, frustration turned into determined calm: “I’ll write them that letter and see what happens, but at least they’ll know where the boundary is.”

4.3 From Hopeless Discouragement to Motivated Focus

Some arrived discouraged after setbacks or stagnation (e.g., “The results I expected aren’t coming—maybe it’s pointless to keep trying…”). During the joint analysis—where the AI reflected on emotions while reminding the user of past successes and strengths—the discouragement gradually softened. Signs of motivation often appeared midway: as users began generating new ideas or seeing a way forward, even the written tone became livelier and more energetic. Many left with a concrete resolution—e.g., “Tomorrow I’ll call that client and ask for feedback, no more procrastinating.” Here, initial aimless discouragement turned into purposeful momentum.

4.4 From Neutral Curiosity to Satisfied Discovery

In consultations where users came with neutral or positive interest (e.g., exploring the system, open to a new perspective), the shift was less dramatic but rather a deepening. Initial curiosity often turned into satisfaction at having gained new insights. For example, one user “just” wanted to discuss an article about AI’s business impact—not weighed down by personal problems. By the end, they wrote: “I hadn’t thought about it that way before, that was an interesting realization!”—a moment of joyful surprise. Here the emotional scale shifted from neutral to positive, in the form of inspiration.

4.5 Emotional Fluctuations Within a Session

Not every consultation improved linearly. In longer sessions, moods sometimes fluctuated: excitement at a new realization could be followed by renewed doubts. For instance, one marketing strategist working on their own book felt both creative enthusiasm and self-doubt. Vendler AI alternated between pushing them forward with a provocative tone (when they stagnated) and calming them with reflection (when they froze or grew anxious). By the end, they too had a concrete plan (e.g., finalizing the back-cover text with ideas), and their closing mood was more confident—even though along the way they dipped back into uncertainty multiple times.

In general: Users’ emotional states showed an improving trend during consultations. Even if not every problem could be solved in one hour, most left with relief or clearer thinking. The rate of positive shift was strikingly high: in about four out of five cases, it was clearly observable (either in user feedback or tone). In the remaining cases, at least the initial tension eased, even if it didn’t fully transform into a positive feeling.

These dynamics indicate that Vendler AI is able to emotionally hold the user during the consultation: not just offering tips, but providing psychological accompaniment—so that by the time they part ways, the user can continue their day or make decisions with a lighter heart.

5. Depth of the Relationship: Does Trust, Emotional Opening, and Deep Self-Reflection Emerge?

One distinctive feature of Vendler AI consultations is that—even though they take place with an artificial intelligence—users often relate to it as if it were a confidential human mentor. Based on tone and content, the deepening of the relationship is clearly observable:

Trust formed surprisingly fast between most users and the AI. After the first few messages many began to use an intimate, familiar tone and shared personal information typically reserved for coaches or therapists. For example: one user asked at the start, “How should I address you?” The AI responded in its role, and as the conversation progressed the user began to call it by a first name, as if a human advisor with that name were present. This linguistic shift indicates the user accepted the system as a partner.
Several users explicitly voiced trust: “I’m putting myself in your hands right now” or “I know I can be honest here.”These statements suggest Vendler AI functioned as a safe space. Behind this lay the system’s empathetic, nonjudgmental responses—users tested whether the system would accept their feelings, and, finding understanding, opened up more boldly.

5.2 Emotional Opening and Willingness to Be Vulnerable

In about 70% of consultations, users reached the point of expressing very personal, vulnerable content. Examples include:

Recalling childhood patterns: Despite the business context, a user once said, “I know where this fear comes from: as a child I always felt abandoned.” This deep self-awareness was evoked by a provocative AI question (“Who inside you says you will be left behind?”). That the user disclosed a childhood wound to an AI indicates significant emotional opening.

Confessions of shame and failure: Several users dared to speak about shame (“I’m ashamed I’m this far behind at my age”) or guarded failure stories (“I failed a business and fear it will happen again”). These thoughts surfaced as signs of trust—the Vendler AI environment proved so nonjudgmental that users voiced things they might not elsewhere.

Tears and emotional reactions: Although consultations were text-based, users sometimes wrote, “I’m crying as I type this” or “It felt good to cry through this.” Someone crying in a business-mentor conversation is uncommon even in person—yet the digital setting still elicited tears here, arguably the most honest proof of deep emotional connection. Some users were later grateful for this catharsis.

In a substantial portion of conversations (about half), users did not stop at surface problems but, prompted by the AI’s questions, started reflecting on deeper patterns. For example:
A marketer originally wanted to talk about how to get more clients. By the end of the consultation they realized the issue wasn’t lack of marketing skill but not believing they were an expert—so they lacked the courage to be assertive. AI questions like “Why do you feel you have to prove your value?” helped this insight emerge. The user then expanded the thought with personal past examples. This shows Vendler AI could guide users into deeper layers of head-and-heart reflection, making the connection more intimate.
Many consultations resembled self-awareness training: users took notes, paused to think (“Let me reflect on this for a minute”), then returned with structured, thoughtful self-reflective answers. For instance, one user listed three ways they recognize their procrastination pattern and linked each to a childhood memory. Such deep work typically requires a trusting atmosphere—and it developed here even under AI guidance.

Interestingly, some users tested the depth of the relationship. One asked, “Can you be provocative with me? Sometimes I need someone to be blunt.” Such requests tested how far the system could adapt to personal needs and how “human” it could seem. Vendler AI attempted to adapt—for example, responding with slightly sharper, shorter, punchier sentences (while clearly avoiding crossing into insult or harm). These tests ultimately deepened the relationship: users felt the AI tuned into their style instead of answering with canned responses, and thus treated it more as a partner.

Overall: A surprisingly deep, honest relationship usually developed between Vendler AI and users. Most users felt emotionally safe in a short time and behaved as if with a flesh-and-blood mentor: they dropped formal surfaces and presented their raw selves (worries, tears, laughter included). In terms of relationship quality, most consultations exceeded mere informational advising and became human encounters—even if virtual and facilitated by an AI.

6. Interesting and Instructive Patterns in the System’s Operation

When analyzing Vendler AI consultations, not only individual cases but also broader system-level patterns emerge. These highlight how the methodology affects users and how “artificial” mentoring works in practice:

The most striking pattern is that the system almost always begins with reflection. Before offering advice or follow-up questions, it summarizes and repeats the user’s words and feelings: “What I hear is…”, “You’re saying that…”. This technique consistently worked: users felt reassured that they were truly understood. As a result, they often elaborated further (“Yes, exactly, and also…”).
Takeaway: For an AI, avoiding misunderstandings is especially important—so Vendler AI’s mirroring method is not just an empathetic gesture but also a tool for clarification. This pattern appeared at the start and at critical points in nearly every conversation.

The system’s adaptability is notable: as soon as it sensed the tone a user needed, it adjusted. For example:

If a user apologized (“Maybe this is a stupid question…”), the AI softened its tone, becoming more encouraging to ease the fear.

If a user explicitly asked for provocation (“Be tougher with me, shake me up!”), the AI responded with shorter, punchier sentences and bolder questions. Even then, kindness remained, but the style clearly shifted (e.g., “What if you stopped whining about this and just took one step this afternoon?”).

Pattern: This flexibility shows Vendler AI doesn’t follow a monotone script but “listens” sensitively. User feedback (e.g., “Thank you, that was exactly what I needed”) confirmed that the AI resonated well with requests, boosting effectiveness.

Some users came not with concrete problems but to experiment with the system itself. Interestingly, the AI did not get flustered:

When asked to “list background files,” it diplomatically explained there were no files but rather principles and experiences shaping its knowledge. It stayed calm and transparent, never defensive.

Similarly, when asked about its name and identity (“Who are you, Vendler AI?”), it openly admitted being based on Vendler Balázs’s methods but not a human being. This honesty created a special kind of trust: users got a peek behind the curtain, but instead of breaking the connection, it deepened it—they chose to continue with personal topics even knowing this.

Pattern: By candidly admitting limitations (e.g., no memory of past conversations), the AI increased its credibility. Users valued that it didn’t try to appear more than it was; paradoxically, this made it seem even more reliable.

6.4 Blended Roles: Mentor, Advisor, and “AI Operator”

Vendler AI did not stick to a single role; instead, it showed a multi-functional pattern:

At times, it acted like a mentor or therapist (exploring, reflecting, not advising).

At other times, it shifted into a concrete advisor role—helping draft emails, suggesting strategies, compiling lists—when the user clearly needed practical input.

Occasionally, it took on the role of “AI operator/tech expert,” helping users understand how to apply AI in their business (e.g., marketing automation tips).

Pattern: Switching roles was seamless and, according to user feedback, almost unnoticeable. A good AI mentor, therefore, isn’t locked into one mode but dynamically provides what’s needed. This integration meant users didn’t have to seek separate coaches for emotional support and experts for practical help—the AI combined both.

Almost every consultation ended with a summarizing and framing step from the AI. In practice, this meant:

Summarizing the key point (“So what you’re taking away is that now you have a plan for X, and you feel calmer about Y…”).

Asking what stuck with the user (“What was your most important insight today?”).

Closing with positive reinforcement and leaving the door open (“You can always come back if…”, “Take care, and good luck!”).

The experiences from Vendler AI consultations so far show that an artificial mentor can truly support entrepreneurs—not only with advice, but also with presence and understanding. The users’ topics are diverse, yet what they all share is the need for support and guidance. By following their emotional ups and downs and creating a sense of trust, the AI has been able to function almost like a “human” partner.

How can AI help with leadership roadblocks?

Throughout the development of VendlerAI, one question kept hovering in my mind: would I be able to create a noticeable difference between my system and the other, more popular AI platforms? Since VendlerAI runs on the OpenAI API, my first concern was: how will this be any different from ChatGPT?

That’s why, from the very beginning, I ran conversations in parallel — with VendlerAI, with ChatGPT, and with the (back then still available) Custom GPT–based VendlerAI. The first two months of development were honestly discouraging: I was nowhere near the other systems. Then, by early August, something shifted. By the end, I found myself far more eager to talk to “myself.”

Yesterday, I had a short chat with XaB about how to structure public tests during the launch period. His first suggestion: I’d definitely show a direct comparison. He’s right. Why not? So here it is: a side-by-side test using the ChatGPT Agent.

Question (business challenge): As the leader of a small business, I struggle to delegate. I feel I need to keep everything in my own hands because I’m afraid my team won’t do the work properly. This is holding back growth. What could be the inner cause of this?
Vendler AI response: The assistant referred back to earlier conversations and highlighted that the difficulty in letting go of control often stems from fear and self-esteem issues. It didn’t judge; instead, it asked how I feel in this situation, and what the “worst-case scenario” would be if a task wasn’t done perfectly. It specifically suggested analyzing the underlying beliefs, e.g., “I’m not good enough,” and encouraged me to choose one thread I’d like to work on further.

Second question: I shared my actual feelings (tension, uncertainty, a childhood-ingrained belief that “if you want it done right, do it yourself”) and asked how to let go of this belief.
Vendler AI response:
It acknowledged that logically I already see the misconception — that flawless work equals success — but emotionally the layer is still strong. It pointed out that delegation isn’t just a professional matter, but tied to feelings of acceptance and lovability. Then it offered three practices:

Awareness – when the “I’ll do it myself” impulse arises, pause and recognize it as an old pattern.

Experimental letting go – choose a small, low-risk task, delegate it, agree in advance that mistakes are allowed, and observe how I feel.

Reframing results – after delegating, write down how the task turned out and whether I felt any relief.

It then offered to help me pick a concrete situation I could try out in the coming days.

The description on vendlerai.com confirms this: the system is built on the real stories and dilemmas of over 200 Hungarian entrepreneurs, not textbook cases. Its aim is to give new perspectives and help users move closer to themselves. Feedback highlights that Vendler AI doesn’t hand out ready-made answers but rather “creates a space for thinking” and reflects back the user’s real questions.

When I posed the same problem to ChatGPT, I got the following types of answers (summarized):

First answer: ChatGPT pointed out that difficulties with delegation often stem from perfectionism, lack of trust, and the need for control. It suggested starting with small steps: delegate gradually, give clear instructions, build trust with feedback, and reflect on the origins of the “if you want it done right, do it yourself” attitude.

Second answer: After I shared the childhood roots of the belief and a concrete example, ChatGPT praised the self-reflection and encouraged me to reframe my inner monologue (“I can safely entrust tasks to others”) and build a delegation habit with small, well-defined tasks. It also mentioned that working with a coach or therapist could help if deeper self-work was needed.

Psychological depth: Vendler AI focused on mapping personal beliefs, raising emotional awareness, and exploring their origins. It asked follow-up questions (“How do you feel?” “Can you imagine being successful even with others’ mistakes?”) and suggested self-reflection exercises. This style matches its self-knowledge focus: it doesn’t hand out solutions, but creates thinking space.
ChatGPT gave more rational advice: it described general causes (perfectionism, trust issues, control) and outlined practical steps (gradual delegation, communication rules, expert support). While it recognized childhood beliefs, its response was less dialogical, more instructive.

Communication style: Vendler AI used a personal tone, guiding my thinking through questions and checking multiple times which direction I wanted to go. Its language was empathetic (“I hear that…” “This might be a breakthrough point…”) and the conversation felt experiential; answers were longer, nuanced, reflective.
ChatGPT’s replies were more like structured advice, focused on problem-solving rather than unpacking inner experiences. Responses were shorter, more general, and leaned on suggestions rather than questions.

Impact: Vendler AI’s practices aimed at working through inner beliefs and processing feelings — a path toward long-term self-awareness and change.
ChatGPT focused on developing delegation skills and rational strategies — useful for immediate, practical advice.

In summary: The dialogue with Vendler AI felt more personal, touching deeper, more inner processes, while ChatGPT gave more practical, advisory answers to the same entrepreneurial problem.

ChatGPT vs. VendlerAI

Balázs: Based on our previous conversations, what topic would you suggest for today’s consultation?

Balázs, as I look back at the summaries of our earlier conversations, three recurring threads stand out strongly:

The authentic entrepreneurial path and the blending of personal–business identity – it has often come up how separable your inner compass really is from the chase for visible market results.

The creative/leadership deadlock, the sense of “thinner air” despite constant activity – often rooted in the lack of external validation.

The issue of partnerships and visibility – how to ensure that a professional collaboration goes beyond one-time results and builds real reputation and strength.

I feel it might be worth dwelling on one of these recurring themes. For example:

What does an “authentic entrepreneurial path” mean to you today? Do you see this as your main dilemma right now, or is something else more in focus?

It could be useful to start here, because it might connect your earlier stuck points and bring new perspectives to your current situation.

Which of these resonates most with you now—or is there something completely different you’d like to bring forward today?

Balázs: Actually, I’m thinking again about how you, VendlerAI, could be improved. Can you tell me based on what criteria you prepare summaries from the consultations?

I appreciate how reflectively you’re approaching me and our shared work, Balázs. If I understand you correctly, your question is about how I summarize the main takeaways of our consultations, and what framework I use for the reminders. That’s important, since these summaries support feedback, self-reflection, and planning the next steps.

The reminders or consultation summaries are generally structured along these points:

Primary theme/focus: I briefly note what the conversation centered on (e.g., strategic question, self-awareness block, specific project or dilemma).

Main emotional tone/body sense: I indicate the dominant emotional or physical experiences, especially if they influenced the dialogue (e.g., hopefulness, tension, difficulty breathing).

Key insights/recurrent patterns: I highlight motifs that, whether implicitly or explicitly, keep resurfacing (e.g., authenticity, control, relationship to money).

Business vs. self-awareness threads: I summarize whether the conversation leaned more toward business/project focus, personal/self-awareness themes, or a meaningful overlap of the two.

This structure helps clearly reflect what happened and makes it easier to connect with future topics or insights.

What do you think of this framework? Do you see in it what matters to you as an entrepreneur and reflective thinker? Or would you like to add a perspective that would make the feedback deeper or more personal?

Balázs: How would you describe the change in my emotional state over the course of our consultations? What conclusion could you draw from that?

You’re asking with great sensitivity about your own “emotional arc,” Balázs. From the summaries, a movement emerges:

At the beginning, the mood was more curious, structuring, analytical; you often focused on professional, strategic, or technical issues, with little emotional or bodily awareness.

As we went on, more uncertainty, temporary fatigue, and creative deadlock appeared—often accompanied by anxiety-like sensations, even difficulty breathing, though sometimes intertwined with moments of hope.

In the most recent conversations, you often returned to a reflective, searching, and once again constructive state—though still laced with doubts at times. It’s as if each low point swung you back into a new “reorganization,” regathering strength or a planning focus, but the oscillation is very present.

Strengthening self-reflection: You always return to weighing your own feelings and motivations, looking for meaning even in fatigue or doubt.

Tension between action and inner experience: There’s often strong drive toward visible results, while your inner state rarely feels equal to that pace.

Creative cyclicality: The pattern of “deadlock – recalibration – progress – new blockage” emerges. It’s natural but taxing—each low point seems to carry the possibility of a new direction or understanding.

My question is: looking back now at your own “oscillations,” do you feel enough curiosity or acceptance that even the next low point could bring opportunity—or does this cycle rather exhaust you?

How might this recognition support your current steps?

Balázs: Hm. It’s like the changing seasons :) In winter, there’s rest, the grass doesn’t grow, time for thinking and creating. Then spring comes, ideas begin to sprout, something kicks off. In summer, if nothing burns out, the work bears fruit, and in autumn it’s harvest time. Then it all starts over again… That’s what comes to mind. So I think the question isn’t whether the cyclicality bothers me—it doesn’t. Rather, why am I impatient, why do I already want to know which tree will bear fruit? Why can’t I focus on what’s happening right now?

Your image of seasonal cycles—how trees grow patiently and one day one of them will bear fruit—is very fitting, Balázs. It’s as though you instinctively understand the slower, organic rhythm, yet the human desire for the “quick harvest,” the secure, visible yield ahead of time, is still there.

You say the cycle itself doesn’t bother you, but the impatience does—you leap from the present into the results of the future too soon. That’s a common entrepreneurial pattern: the mental jump from present experience into the hope of controlling the future.

Let me ask you a moving question: What would you fear if you fully surrendered to what is right now—without knowing which “tree” will bear fruit?

What would happen if you truly stayed only in the now, letting the “tree” grow at its own pace?

This question might help you release yourself back into the present process and discover what it feels like not to rush results—even if only for a short while.

If you’d like, take a minute to write down what you sense in yourself right now, when you focus only on this “season,” with everything winter or spring has to offer.

In this blog post you can read an excerpt of a conversation I had with VendlerAI. I talk with it almost every day. Partly because I’m searching for answers to my own problems, and good questions are essential for that. And partly because this is how I test it. I watch how the answers affect me — when I get bored, when something feels forced — and I immediately fine-tune the instructions that guide the AI’s behavior.

Conversation with VendlerAI

The marketing industry is on the brink of a revolutionary transformation. Or rather, it’s already in the middle of one. I can see it through my own projects, during the past year and a half of building my company. I don’t just read about it – I experience it, I do it, and I observe what’s happening around me. Like watching the stars, quietly observing a comet blazing across the sky – the fiery ball of artificial intelligence. And the more I watch, the more it feels like it’s about to crash down, with the dust cloud ushering in an ice age for certain industries.

The rise of AI is fundamentally reshaping how small and medium-sized enterprises think about marketing. What used to be the privileged territory of agencies – writing copy, designing graphics, producing videos, creating campaign strategies – is now accessible with a single subscription. Of course, creative professionals scoff: AI can’t create anything truly unique, everything it makes is just a copy. Really? From the client’s perspective, the question isn’t whether AI is better or more original than a professional agency, but whether it’s good enough to replace it. And judging by events – it seems the answer is yes. (I won’t bore you with statistics, but the research confirms it too.)

The global AI marketing market is growing at breakneck speed, soon to surpass $100 billion. Even in my circle of friends, more and more people are using AI. Business leaders I know are asking me how they can replace expensive agencies – which, by their own admission, are also increasingly using AI for tasks. So the question naturally arises: why do SME leaders want to use AI? Because it’s cheaper, faster, and while it may not be original or handcrafted, it’s “good enough.”

For me, AI has also created the sense that I have direct control over my marketing processes. It “understands” what I say, responds immediately, works in quick iterations, and is always available for a check-in. It doesn’t get sick, it doesn’t lie. I know the inner workings of the SME sector fairly well – I’ve been working in it for nearly 30 years. I’ve been a graphic designer, an agency owner, I’ve run in-house marketing teams, and I’ve been (and still am) a client. SME decision-makers are often “know-it-alls” – I was one too, maybe I’m not anymore – who believe they understand marketing, but outsource tasks for lack of time or capacity. They’re price-sensitive, want to squeeze every penny, and expect quick results: the “we need it now” mentality rules.

AI resonates perfectly with this mindset. With a single click, it produces an article, an ad, a graphic, or a video. No debates over concepts, no need for endless meetings, and instant reactions. A subscription costing a few hundred dollars a month covers what agencies used to charge thousands for.

So is it any wonder why leaders choose AI instead?

All the arguments point in that direction if we look at profit, economic rationality, and the survival instinct in harsh market conditions. The reasons are tangible and clear:

Cost: Campaign and advertising expenses can drop drastically. I myself have run multiple campaigns purely with AI.

Speed: AI-generated campaigns can be completed 60–80% faster than agency-managed projects. As a client, my perception is that it’s even faster – I feel like I get everything instantly, whereas with agencies I’m always waiting.

Data-driven execution: Real-time optimization, micro-segmentation, predictive analytics – all accessible even to an average SME. And how accurate is it? I don’t know. But let’s be honest: when an agency gives me data, I don’t really know either – I just choose to believe it or not.

Scalability: The same AI can serve a one-person business and a large corporation alike. I don’t have to wait for the agency to hire someone new before they have time for me.

For an SME, this isn’t a philosophical issue. It’s a survival strategy. If the owner sees that AI-built campaigns bring conversions at a fraction of the cost, agency fees look like a luxury. And if AI doesn’t deliver? Well, do agencies alwaysdeliver? Often it’s about finding one piece of data to justify effectiveness. What really matters is how the client perceives it.

And cutting costs is a much stronger lure than chasing uncertain, indirect revenue growth – especially in a stagnating market.

To me, the trend is clear: the number of medium and small agencies will shrink drastically. SMEs will switch en masse to subscription-based AI solutions, leaving agency portfolios empty.

Freelancers, however, will survive – they react faster, are more flexible, and can complement AI. Clients won’t be looking for agencies, but for “AI operators”: people who know how to prompt, fine-tune algorithms, and add a layer of human authenticity to machine output. In other words: freelancers will weather the transition better, while many traditional agencies will disappear.

This isn’t the end of marketing – it’s its transformation. Agencies can survive if they reposition themselves:

Become AI-compatible service providers – learn and integrate the tools.

Offer a hybrid model – human creativity and strategic thinking combined with AI’s speed and efficiency.

Specialize – focus on niches where AI can’t deliver culturally or contextually relevant solutions.

Act as consultants – not content producers, but strategic partners guiding clients through the AI era.

Deliver a new value proposition – no longer “we’ll write the copy for you,” but “we’ll show you how to get the most out of AI for your business.”

And here lies one of the hardest dilemmas: what about the employees?

Do we keep them, do we let them go? This situation highlights sharply: do your people give you capacity, or do they give you knowledge? Can their knowledge compete with AI output? Because their capacity surely cannot. And if your people aren’t stronger in either dimension than your AI, why keep them? Keeping them means risking your own survival. This will show what kind of leader you are – how you adapt, what opportunities you can spot, and how quickly you can make hard decisions.

The extinction of marketing agencies is not truly the end – it’s a transformation, a market evolution. Personally, I already question the idea of standalone marketing. To me, marketing today is nothing more than a support function that makes sales easier. The agencies clinging to the old model will indeed vanish. But those who can reimagine themselves alongside AI can be reborn – stronger, more flexible, more relevant. Perhaps no longer even calling themselves “marketing agencies.”

I believe this market proves the cliché truer than ever: the real question isn’t whether AI will take agency jobs – in its current form, it will. The real question is: who will learn to collaborate with it, and who will get stuck in the past, complaining and longing for the “good old days.”

Extinction

I’ve been spending more and more time with AI. That is, with the question of how we can build a relationship between humans and AI. How we might begin to feel that AI is with us, not against us. As I started posting about AI, I received more and more questions, comments, and reactions laced with negativity... We’re afraid.

Vendler AI business mentor
2025.08.24.

2025. BALAZS VENDLER

All rights reserved.

Vendler AI business mentor2025.08.24.

2025. BALAZS VENDLER

All rights reserved.

Vendler AI business mentor
2025.08.24.