---
title: AI Made Products Easier to Build, But It Did Not Make Entrepreneurship Less Risky
description: AI has simplified product creation, but entrepreneurship remains risky. Discover how the shift in development processes impacts founder confidence and market fit.
image: https://www.medicalofficemarketing.org/hubfs/ChatGPT%20Image%20Sep%2026%2c%202026%2c%2012_03_28%20PM.png
---

[Skip to content](https://www.medicalofficemarketing.org/blog/ai-made-products-easier-to-build-but-it-did-not-make-entrepreneurship-less-risky#main-content)

[![LOGO Design with White Background JPEG](https://www.medicalofficemarketing.org/hubfs/Logos/LOGO%20Design%20with%20White%20Background%20JPEG.jpg)](https://www.medicalofficemarketing.org/)

[Medical English Courses](https://www.medicalofficemarketing.org/Medical-English-Courses)

[Contact Us](https://www.medicalofficemarketing.org/contact)

- [Home](https://www.medicalofficemarketing.org)
- [Business Strategy Services](https://www.medicalofficemarketing.org/business-strategy-financial-modeling)
  
  Show submenu for Business Strategy Services 
  
    - [Financial Forecast Modeling](https://www.medicalofficemarketing.org/business-strategy-financial-modeling#modeling)
    - [Concierge & Private Practice Models](https://www.medicalofficemarketing.org/business-strategy-financial-modeling#private)
    - [For Regional Anesthesiologists](https://www.medicalofficemarketing.org/business-strategy-financial-modeling#regional)
    - [Grant & Funding Strategy](https://www.medicalofficemarketing.org/business-strategy-financial-modeling#grants)
- [Patient Experience Services](https://www.medicalofficemarketing.org/patient-experience-care-optimization)
  
  Show submenu for Patient Experience Services 
  
    - [Healthcare Consumer Journey Mapping](https://www.medicalofficemarketing.org/patient-experience-care-optimization#mapping)
    - [Patient Experience Consulting](https://www.medicalofficemarketing.org/patient-experience-care-optimization#consulting)
- [About Us](https://www.medicalofficemarketing.org/about-us)
- [Blog](https://www.medicalofficemarketing.org/blog)

- [Home](https://www.medicalofficemarketing.org)
- [Business Strategy Services](https://www.medicalofficemarketing.org/business-strategy-financial-modeling)
  
  Show submenu for Business Strategy Services 
  
    - [Financial Forecast Modeling](https://www.medicalofficemarketing.org/business-strategy-financial-modeling#modeling)
    - [Concierge & Private Practice Models](https://www.medicalofficemarketing.org/business-strategy-financial-modeling#private)
    - [For Regional Anesthesiologists](https://www.medicalofficemarketing.org/business-strategy-financial-modeling#regional)
    - [Grant & Funding Strategy](https://www.medicalofficemarketing.org/business-strategy-financial-modeling#grants)
- [Patient Experience Services](https://www.medicalofficemarketing.org/patient-experience-care-optimization)
  
  Show submenu for Patient Experience Services 
  
    - [Healthcare Consumer Journey Mapping](https://www.medicalofficemarketing.org/patient-experience-care-optimization#mapping)
    - [Patient Experience Consulting](https://www.medicalofficemarketing.org/patient-experience-care-optimization#consulting)
- [About Us](https://www.medicalofficemarketing.org/about-us)
- [Blog](https://www.medicalofficemarketing.org/blog)

![Hamburger](https://22206477.fs1.hubspotusercontent-na1.net/hubfs/22206477/list-bold-1.svg)

![Close Icon](https://22206477.fs1.hubspotusercontent-na1.net/hubfs/22206477/x-bold-1.svg)

# AI Made Products Easier to Build, But It Did Not Make Entrepreneurship Less Risky

[Marion Davis,](https://www.medicalofficemarketing.org/blog/author/marion-davis)  26 September 2026

![ Illustration of a founder reviewing blueprints beside a glowing app on an unfinished foundation, symbolizing the business groundwork still needed for AI-built products.](https://www.medicalofficemarketing.org/hubfs/ChatGPT%20Image%20Sep%2026%2c%202026%2c%2012_03_28%20PM.png)

Recently, I spoke with a founder who did not understand what I meant by the "pre-marketing" phase of product development. That conversation stuck with me because even a few years ago, product development usually involved much more friction.

One of my earliest professional projects as a curriculum designer involved creating an e-book that helped entrepreneurs understand what developing a product actually cost. At the time, startup competitions were giving out grants in the tens of thousands of dollars because bringing a product to market could require substantial capital before the founder knew whether anyone would buy it.

For physical products, there might be years of expertise behind the product itself. Think about a professional furniture maker who spends years developing a craft, then designs a product, builds prototypes, sources materials, works out production costs, and eventually figures out how to sell it. Software had its own barriers. Building even a relatively simple digital product required developers, time, money, and technical expertise.

Because development was expensive, there was a strong incentive to ask questions before building. Who wants this? What problem are we solving? How do potential customers describe that problem? What will they pay? That questioning process, sometimes called pre-marketing, meant founders did not necessarily wait until the product was finished to discover whether anybody wanted it.

## **What Changed**

AI has changed that sequence, and the change is measurable, not just anecdotal. McKinsey has found that generative AI can improve developer productivity by 35 to 45 percent and can meaningfully cut the time needed for code documentation and refactoring, which lowers both the cost and the speed of getting from idea to working product. A founder can now use AI to build an app, write a website, create a brand, generate marketing copy, and develop an investor deck, often in a fraction of the time and cost that used to be required.

Something strange happens when creating the artifact becomes that much easier. The product can look more mature than the business behind it, and sometimes the founder's own confidence has not caught up to what is sitting in front of them.

## **Two Kinds of Founders**

I have started to notice a pattern in how this shows up in conversation. A founder who generated an app largely from a prompt will often preempt me with a small, deflating comment before I have said a word about the product. They will say something like "I know it's basically just AI" or "it's nothing special, I just had it built for me." Compare that to a founder who built something from the ground up, even if AI enhanced parts of it along the way. That person tends to walk me through the thing with visible pride, explaining the decisions they made and what they tried that did not work, because they lived inside that process.

## **What Confidence Training Taught Me**

Part of my background is in confidence training, which is a structured approach to building genuine self-efficacy rather than just telling someone to feel more confident. I worked with a client who was launching a certified nursing assistant school. I helped her secure the school's registration and then built custom confidence training material for the CNA students themselves, because confidence is closely tied to skill retention in clinical training. Once the program had trained CNAs who could demonstrate their competence, I led the marketing campaign that sold exclusive staffing contracts to local long term care homes, and the graduates' confidence was part of what made that pitch credible.

Confidence training works by giving someone repeated, structured opportunities to perform a skill, get honest feedback, correct mistakes, and eventually succeed at something that once felt hard. Psychologist Albert Bandura's research on self-efficacy identified the strongest source of genuine confidence as mastery experience, the felt sense of having personally overcome a real difficulty. Watching someone else succeed, or being told you will do fine, helps a little, but it is a distant second to having done the difficult thing yourself.

That is a useful lens for the product itself, not just for the CNAs I trained. If AI generated most of the features, the logic, and the design decisions from a short prompt, the founder skipped the mastery experience. They did not personally wrestle with the hard parts and come out the other side knowing they could solve the next hard part too. What they have is a finished artifact without the internalized competence that normally accompanies one, and on some level they seem to sense that gap even when they cannot name it. A founder who built the foundation themselves, then used AI to move faster on top of it, usually does not have that problem, because they still went through enough of the difficulty to own the outcome.

## **The Product Exists Before the Questions Are Answered**

The product now often exists before the founder has answered the questions that should have come first, and without the confidence that would normally have been built along the way. The traditional sequence was never universal, but it often looked something like expertise, idea, research, validation, investment, development, audience building, and launch. Increasingly, I encounter something closer to idea, build, product exists, now find customers.

That changes the conversation I have as a marketer. I can be speaking with somebody whose app is ready for beta while asking questions that ordinarily should have appeared much earlier in the process. Who is this really for? Why would someone trust you with this information? What have potential customers actually told you?

## **What the Failure Data Shows**

The data on why businesses actually fail backs up why that sequencing matters. CB Insights has analyzed hundreds of startup post mortems and consistently finds that the largest single cause of failure, at roughly 42 to 43 percent, is poor product market fit, essentially building something the market did not sufficiently want. That figure sits well ahead of running out of cash or getting outcompeted, and researchers who revisited the analysis with four times the original sample size found the number barely moved.

The Bureau of Labor Statistics data that the Small Business Administration draws on tells a similar story. Roughly one in five new businesses close within their first year, and about half do not make it to the five year mark. None of that risk disappeared because building got cheaper. It just moved further downstream, arriving at the marketer's desk instead of being resolved before a single line of code was written, and often arriving alongside a founder who is not entirely sure the thing they built deserves to succeed.

Over the past year, I have noticed more prospective clients asking me to provide certainty about things I cannot possibly know yet. They want a forecast before there is acquisition data, an exact scope before they have supplied enough information to define it, and increasingly detailed strategy before starting a paid engagement. Some of that reads less like ordinary business caution and more like a founder looking for someone else to supply the conviction that a mastery experience would normally have given them.

## **The Soft Skills Prediction**

There was a widely shared prediction that as AI automated technical and production tasks, distinctly human skills would rise in value. The World Economic Forum's Future of Jobs Report 2025 lines up with that idea on paper. It surveyed more than a thousand employers and found that skills like analytical thinking, creative thinking, and resilience, flexibility, and agility remain among the most sought after capabilities even as technical skills grow fastest in relative terms. Upwork's own research points the same direction, reporting that a large share of business leaders say they are willing to pay a premium for independent talent who bring creative or innovative thinking to an engagement.

The prediction may have been right about demand. I am less convinced it was right about valuation.

My work involves listening to people, identifying what they are actually trying to accomplish, synthesizing complicated information, and creating structure where there was previously ambiguity, much of what confidence training does for a person, just applied to a business instead. There is extraordinary demand for those capabilities, but they are sometimes treated as the free interpersonal layer around the transaction rather than as part of the professional service being purchased.

That dynamic shows up clearly in commission only arrangements. I have seen an increase in founders looking for marketers willing to work on commission, through affiliate revenue, or through some other heavily performance dependent structure, often framed around wanting marketers who are confident in their own work. It is a strange ask coming from a founder who cannot always articulate confidence in their own product.

I am confident in my work. What I cannot be confident about is every variable inside someone else's company, including whether the founder will follow the strategy, whether the product will retain customers, or whether the market will validate a product that was never properly validated before I arrived.

There are legitimate performance compensation arrangements, where a substantial base fee combined with incentives tied to metrics the marketer can meaningfully influence aligns interests well. That is very different from asking a marketer to prove belief in their own skill by risking unpaid labor on someone else's unvalidated business. If a founder is confident enough in the product to expect that, it is worth asking why they are not confident enough to risk money hiring the marketer instead.

## **Where AI Adoption Actually Stands**

It is worth checking whether the picture I am describing–cheap building outpacing earned confidence–holds up outside the United States because AI adoption is genuinely uneven across countries, though not always in the direction people assume. Microsoft's global AI diffusion research for the second half of 2025 found that the United Arab Emirates and Singapore lead the world in generative AI use, while the South and Central America region had 14 of 24 countries with adoption rates below 20 percent, including Mexico, Venezuela, and Brazil. Interestingly, the United States itself ranks well outside the top tier of adoption, at around 28th globally, despite housing most major AI companies and the most advanced AI infrastructure, with Chile standing out as Latin America's clear leader in the space.

So the slower adoption in much of Latin America is real, but it is not simply a story of a region lagging behind. It is closer to a region where AI adoption and entrepreneurial activity are moving on two different tracks entirely.

## **A Region That Builds Businesses Without Much Infrastructure to Support Them**

Colombia is a useful example of that split. The country has historically ranked among the most entrepreneurial in the world by the Global Entrepreneurship Monitor's measures, at one point placing fourth globally for highest entrepreneurial activity, and Latin America as a region consistently posts total early-stage entrepreneurial activity rates well above the global average.

What Colombia and several of its neighbors do not have, at the same scale, is the small business infrastructure to match that appetite for starting something. Colombia's own central bank has reported a national labor informality rate of 56 percent, and more recent reporting places Colombia's informality among the highest in Latin America, ahead of both Chile and Brazil. Among new Colombian entrepreneurs specifically, just over a third are formally registered, and only about half of established entrepreneurs are.

Colombia is not alone in this pattern. Ecuador and Guatemala show a similar combination of extremely high entrepreneurial activity and high informality, with Ecuador, Guatemala, Chile, and Panama posting the four highest entrepreneurial activity rates in the region in recent GEM surveys, and Guatemala reporting roughly one in three adults starting or running a new business. Research on Latin American self-employment has flagged Bolivia, Colombia, the Dominican Republic, Ecuador, Panama, and Peru as countries where informal self-employment runs substantially higher than their income level alone would predict.

That combination, a lot of entrepreneurial energy paired with comparatively thin formal business infrastructure, is exactly the kind of environment where deep clinical or technical competence can exist without the surrounding scaffolding that would normally translate it into confidence within a different system.

## **The Real Barrier for Colombian Clinicians**

That is the pattern I see up close with the Colombian clinicians I have worked with. The barrier is rarely clinical skill. It is that they have had no cross-border training on how to operate inside US professional culture, so a genuinely excellent clinician can walk into that gap and read it as a personal shortcoming rather than what it actually is, a lack of exposure to a different system's norms and expectations.

What makes this particularly striking is what their own training actually built in them. Colombian general physicians typically complete a rural health year after medical school where they have graduated responsibility handed to them under supervision. The clinicians I work with speak about that year with a specific, earned kind of confidence, because it was the mastery experience Bandura's research points to, a real difficulty they overcame themselves, not a credential handed to them.

Their holistic, patient-centered approach comes from the same place. In a lower-resource setting, using nutrition when medication was not available, or relying on a patient's own narrative when diagnostic tests were not accessible, was not an aspirational buzzword. It was a practical necessity, and it made them resourceful in a way that shows.

## **The Confidence Gap Runs Both Ways**

Some Colombian clinicians tell me they doubt they can compete with US healthcare professionals. I push back on that directly. I have seen specific competence gaps in US medical settings myself, and some Colombian generalists who have worked in lower level roles like medical assistants in US offices have noticed the same thing independently, specifically that US doctors are often not practiced at troubleshooting because they are used to the tools thinking for them and to blaming the patient or the equipment when something does not work.

That is really the same story as the AI-built product, told from the opposite direction. A founder who generated a product from a prompt often lacks confidence because they skipped the mastery experience that would have earned it, even though the finished product looks impressive. A Colombian clinician who spent a supervised year making real decisions in a resource-scarce setting has already had that mastery experience in full, and often more of it than their US counterparts, yet a different kind of gap, exposure to a foreign professional culture and its status signals, can leave them just as unsure of themselves. In both cases the visible product, whether it is an app or a credential, is a poor proxy for what actually earns confidence, and my job in both settings ends up being the same, helping people see where the real competence already sits.

## **AI Did Not Eliminate Entrepreneurial Risk**

AI has made many things dramatically cheaper. It can reduce development costs, accelerate research, and let an individual test an idea that might previously have required an entire team. Those are genuine advances.

But AI did not eliminate uncertainty, product-market fit, customer psychology, competition, or the need for judgment, and it certainly did not eliminate entrepreneurial risk. It changed where some of the costs appear, including the emotional cost of not having earned confidence in what you built.

When building becomes cheap enough, a founder can reach the marketing stage without having made the financial, professional, or emotional investment that historically preceded it. That is not inherently bad since cheap experimentation is one of the real advantages of the technology.

The problem begins when the remaining uncertainty, financial and psychological alike, gets pushed downstream, onto whoever is hired next to help sell what already exists. A good marketer can investigate that uncertainty, design experiments around it, and change direction when evidence contradicts the plan. What we cannot do is manufacture the mastery experience after the fact.

The entrepreneur still has to make a decision without knowing everything that will happen next, and still has to build the kind of confidence that only comes from having done the hard part themselves. AI can build the product faster. It cannot take that responsibility away.

I’ve had US physicians and US and European HealthTech businesses alike find my blog articles on this site and point out the extensive work put into analyzing systems; identifying healthcare gaps; and reviewing, synthesizing, and expanding on relevant research. Rather than working to address their own competence and product gaps, they often asked me for the shortcut of building upon my platform, such as wanting me to post an article they wrote about their product for free. Without fail, it was always a product that was lacking in quality and ability and did not align with the complexity of solutions that I address in my articles. These small business owners wanted me to absorb the gap and fill in the holes via my professional reputation and competence. What they seemed unwilling to do largely was face the fact that trying to take shortcuts in marketing was a downstream effect of having failed to create a better product and/or service at an upstream point.

What ties these experiences together is the attempt to borrow credibility without doing the work that makes it warranted. A marketer can help a founder test assumptions, strengthen an offer, and communicate its value. That work has value of its own, and it requires the founder to remain accountable for what the evidence reveals. If the product needs improvement, that becomes part of the work. If demand is uncertain, investigating it requires investment. AI can lower the cost of getting started, but building a business still means taking responsibility for what you put into the world and giving people a reason to trust it.

## **Ready to build confidence in what you offer?**

If you are a clinician or healthcare founder ready to invest in that work, I help connect your expertise with unmet patient needs through business strategy, positioning, and patient experience consulting. Together, we can identify where your offer is strong, where gaps need attention, and what it will take to reach the people you can meaningfully serve.[Explore my business strategy services](https://www.medicalofficemarketing.org/business-strategy-financial-modeling) to take the next step toward building a business with credibility supported by what it delivers. You can also schedule a [complimentary 15-minute introductory call](https://calendly.com/mariondavis/complimentary-introduction-call) to discuss your business needs with me.

Obtain a competitive advantage by addressing hidden markets in healthcare. Build a durable brand by offering value that lasts.

[Subscribe](https://www.medicalofficemarketing.org/subscribe)

- <https://www.facebook.com/SpinalCSFLeakSupport>
- <https://www.linkedin.com/company/medical-office-marketing>

Copyright © 2026 | SupportedSuccess, LLC | All Rights Reserved.

 |

```json
{
  "@context" : "https://schema.org",
  "@type" : "BlogPosting",
  "author" : {
    "@type" : "Person",
    "name" : "Marion Davis",
    "url" : "https://www.medicalofficemarketing.org/blog/author/marion-davis"
  },
  "dateModified" : "2026-09-26T16:06:53.814Z",
  "datePublished" : "2026-09-26T16:06:53.000Z",
  "headline" : "AI Made Products Easier to Build, But It Did Not Make Entrepreneurship Less Risky",
  "image" : [ "https://www.medicalofficemarketing.org/hubfs/ChatGPT%20Image%20Sep%2026%2c%202026%2c%2012_03_28%20PM.png" ],
  "mainEntityOfPage" : {
    "@id" : "https://www.medicalofficemarketing.org/blog/ai-made-products-easier-to-build-but-it-did-not-make-entrepreneurship-less-risky",
    "@type" : "WebPage"
  },
  "publisher" : {
    "@type" : "Organization",
    "logo" : {
      "@type" : "ImageObject",
      "url" : "https://www.medicalofficemarketing.org/hubfs/Logos/LOGO%20Design%20with%20Black%20Background%20JPEG.jpg"
    }
  }
}
```