Skip to main content

Lots of Talkers, Very few Doers

··2126 words·10 mins·
Author
Brian Ritchie
Table of Contents

(This article was originally published in 2012 and has been updated with new insights on AI’s impact on this challenge)

BLUF: While you’re heads-down building something real, AI-powered “experts” are flooding the conversation with polished content they never implemented. By the time you come up for air, you’re forced to choose: spend precious time fighting for recognition or get back to building the next thing. I’ve lived this paradox, and I’m sharing what I’ve learned about surviving it.

The Problem With Today’s Startup Ecosystem
#

The modern tech scene has a credibility problem. While scrolling through my LinkedIn feed last week, I counted no fewer than fourteen “thought leaders” making grand proclamations about technologies they’ve never actually implemented. Their posts – suspiciously polished and garnished with all the right keywords – had thousands of engagements. Meanwhile, the people actually building revolutionary tech were nowhere to be found in my feed. They were probably, you know, building.

Every local tech meetup now seems to feature the same pattern: a crowded room where 95% of attendees are furiously networking and “ideating,” while the handful of people actually building things are huddled in the corner, quietly comparing notes on real problems they’ve solved. That’s if they even made it to the meetup – most are too consumed with solving actual technical challenges to attend.

In this article, I’ll examine the growing disconnect between talking and doing in startup communities, how AI tools have supercharged the problem, why authentic builders are increasingly disadvantaged, and what we can do to shift our culture toward valuing demonstrated ability over performative knowledge.

Key Finding:

Research from Stanford University shows that technically focused founding teams can more quickly reach market milestones, from design and prototype completion all the way to product launch.

The Anatomy of a Tech Scene Talker
#

We all know them. They’ve “advised” seven stealth startups (none of which ever launched). They claim to be “technical enough to be dangerous” (translation: once edited a config file). They’ve pivoted so many times they’re practically spinning in circles.

What exactly characterizes these startup scene participants who talk more than they do?

Definition: The Tech Scene Talker

A community participant who builds perceived expertise primarily through networking, jargon fluency, and association rather than through demonstrated technical or business execution. Often has impressive social media presence but limited portfolio of completed projects. Increasingly augmented by AI tools that help create the illusion of deeper knowledge than they possess.

The modern tech talker typically exhibits these characteristics:

  • Claims expertise across improbably broad domains (“AI, blockchain, quantum computing, and no-code expert”)
  • Résumé filled with advisor roles but few operational positions
  • Uses cutting-edge terminology without contextual understanding
  • Portfolio consists mainly of concept decks rather than shipped products
  • Deflects specific technical questions with vague, high-level responses
  • Measures success by networking metrics rather than product outcomes
  • Publishes suspiciously frequent and polished content
  • Has more time for personal branding than anyone actually building something should

I’ve witnessed full-room presentations where someone speaking authoritatively about machine learning deployment couldn’t answer basic questions about model evaluation metrics. It’s the intellectual equivalent of Instagram filters—the appearance of depth without the substance.

(And I’m not immune to occasionally slipping into talker mode myself. We’ve all done it. The difference is recognizing when you’re operating beyond your actual expertise.)

The AI Amplification Effect
#

The rise of sophisticated AI writing and content generation tools has dramatically worsened this problem. What used to require genuine expertise – crafting thoughtful, informed articles or presentations – can now be approximated with a few prompts to an AI assistant.

In 2025, the tools available to “talkers” have become extraordinarily powerful:

  • AI writing assistants that generate seemingly expert content on any technical topic
  • Auto-generated case studies with fabricated metrics and “lessons learned”
  • Presentation creators that transform bullet points into visually stunning decks
  • Content schedulers that maintain consistent social presence without actual effort
  • Voice cloning technology that creates podcast appearances requiring minimal input

Key Finding:

85% of marketers believe generative AI will transform content creation in 2024, with 60% of those using generative AI content expressing concerns it could harm brand reputation due to bias, plagiarism, or values misalignment

The result is a massively amplified signal-to-noise ratio, where those most willing to leverage AI tools for self-promotion can create the appearance of 10x more output and insight than those focused on actual implementation. For every genuine technical insight shared online, there are now dozens of AI-polished, keyword-optimized imposters.

The cruel irony? The very tools being breathlessly discussed by “talkers” are simultaneously being used to amplify their voices above those who are actually building those tools.

The Builder’s Paradox - Build or Be Noticed
#

And here lies the central paradox faced by genuine builders in today’s ecosystem: while you’re heads down solving real problems, the conversational landscape about your very domain is being shaped by those who aren’t doing the work.

Consider this scenario that plays out constantly:

  1. A talented engineer spends a couple of years building a novel solution to a complex problem
  2. During those years, they’re largely absent from the conversation, focused on execution
  3. Meanwhile, non-builders use AI tools to produce dozens of “thought leadership” pieces about the problem space
  4. By the time the actual solution launches, the builder finds their voice drowned out by established “experts” who’ve never implemented anything
  5. The builder faces a choice: divert precious time to self-promotion or continue focusing on the next technical challenge

I experienced this firsthand after spending years architecting and building specialized data processing pipelines, privacy-by-design CDP substitutes, and niche analytics platforms. When I finally came up for air and tried to share our learnings, I discovered the conversation had been dominated by consultants who had created impressive-looking frameworks and methodologies – all without ever having built a production system. Their AI-assisted content creation had established them as the “go-to voices” despite their lack of hands-on experience.

The painful reality for builders? By the time you’ve solved a hard problem and have genuine insights to share, you’re entering a conversation that’s already cluttered with noise, shaped by those who specialize in talking rather than doing. And now you must somehow find the time and energy to distinguish your voice while simultaneously moving on to the next building challenge.

Spotting Authentic Doers in a Sea of Talkers
#

How do you identify the people actually building valuable things in your tech community? Here’s a comparative framework that’s become even more relevant in the age of AI-amplified noise:

Characteristic Talkers Doers
Portfolio Concept decks, wireframes, “coming soon” projects Shipped products, completed implementations, public repositories
Knowledge sharing Speaks in abstractions and generalizations, content suspiciously consistent and frequent Offers specific insights with contextual nuance, publishes irregularly based on actual progress
Problem focus Discusses trendy problems that attract attention Tackles unsexy problems that need solving
Failure stories Vague references to “pivots” and “learnings” Specific technical or business failures with detailed lessons
Technical claims Broad, expansive expertise claims Clearly defined areas of strength with acknowledged limitations
Time allocation Primarily networking and presenting Significant time spent building and implementing
Questions Asks high-level, impression-managing questions Asks detailed, problem-solving oriented questions
Content output Suspiciously high volume and consistency Sporadic, tied to actual project milestones
Technical depth Repeats conventional wisdom with polished language Provides counterintuitive insights from hands-on experience

The most reliable indicator has always been evidence of completed work. Not perfect work—sometimes quite flawed work—but actual functioning implementations that solved real problems.

The true experts I’ve met can immediately dive into the thorny details of their domain. Ask a real machine learning engineer about their most challenging feature engineering problem, and you’ll get a thoughtful, nuanced response that reveals both expertise and limitations. Ask a talker the same question, and you’ll get an impressively fluent response that somehow manages to avoid any specific implementation details.

Building Genuine Credibility Without Compromise
#

After years of wrestling with this problem, I’ve finally confronted the central question facing every builder: How do I gain the recognition my work deserves without abandoning the very work that makes me valuable?

Through trial and error (mostly error), I’ve realized that authentic credibility in an AI-amplified world requires a deliberate approach that I’m still learning to implement:

  • Building in public, but selectively: Share meaningful milestones rather than trying to maintain constant presence
  • Demonstration rather than claims: Letting working implementations speak louder than polished explanations (though of course there’s the risk of them stealing it as their own, can’t win everything)
  • Signature depth and focus: This is hardest for me given my vast experience but trying to concentrate my public communication on areas where I’ve solved problems others haven’t
  • Create artifacts that prove implementation: Open source components, technical documentation, or detailed case studies
  • Finding “builder-recognizes-builder” channels: Communities that value demonstrated work over performative expertise
  • Allocating promotion time: Scheduling limited, focused time for sharing work rather than letting it consume building time
  • Leverage builder-specific signals: I’ve learned that GitHub contributions (or equivalent) often speak louder than unnecessary LinkedIn posts to those who matter

Key Finding:

According to a study published in 2025 by researchers from IE Business School, University of Lausanne, and Harvard Business School, startups that engage with open source communities show “a substantial increase in the likelihood of being funded.” Their difference-in-differences models revealed that early-stage startups engaging with open source communities on GitHub increased their likelihood of receiving funding by at least 36%

The builders I admire who manage this balancing act typically allocate about 80% of their professional time to actual building and 20% to thoughtful, focused communication about what they’ve built. They’re selective about where and how they share, prioritizing depth over reach.

“I’d rather be known by 100 people who actually understand the problem than 10,000 people who just like how I talk about it.” I’ve taped this quote to my monitor metaphorically speaking as a daily reminder.

What I’ve Learned About Culture Change
#

After battling this frustration for years, I’ve come to believe that shifting from a talking culture to a doing culture requires both personal adaptation and community-level changes, especially as AI tools make talking increasingly frictionless.

What I’m trying to practice personally:

  • Reserving 70-80% of my “tech scene” time for building, 20-30% for communicating
  • Setting concrete, public/semi-private completion goals rather than just sharing concepts
  • Contributing tangible work before seeking recognition
  • Getting comfortable saying “I don’t know enough about that yet” when appropriate
  • Using AI tools to make my building more efficient, not just my talking more prolific

What I’ve started advocating for in communities I’m part of:

  • Designing events around demonstrated work rather than networking
  • Creating showcase opportunities that require working implementations/demos
  • Establishing mentor networks based on verified technical experience
  • Recognizing and amplifying the voices of proven builders
  • Implementing “show, don’t tell” policies for presenters and speakers
  • Creating dedicated spaces where builders can connect with other builders

The tech communities I’ve seen thrive are those that gradually shifted their status mechanisms from “who can talk most impressively” to “who has built most successfully.” This doesn’t mean ignoring the value of communication—it means grounding that communication in demonstrated capability. I’m trying to be part of this shift, even when it feels like swimming upstream.

Key Takeaways
#

  • The tech ecosystem suffers when performative expertise is valued over demonstrated capability
  • AI tools have dramatically amplified the ability of “talkers” to create the appearance of expertise
  • Authentic builders face a paradox: while they’re building, the conversation about their domain is shaped by non-builders
  • Even with limited time, builders need strategic approaches to gain recognition without abandoning their core work
  • Both individuals and communities can take concrete steps to foster action-oriented cultures that value implementation over impression management

The next time you’re at a tech event and someone is holding court about their revolutionary approach to [insert trending tech here], it’s worth politely asking: “That sounds fascinating—what have you built with it so far?” The response will tell you everything you need to know.

After all, in a world where AI makes talking easier than ever, the ability to transform concepts into reality remains the scarcest and most valuable skill of all.


(Original post)

There’s a few things that have started to annoy me in the local startup scene, A LOT. One thing in particular is the fact that there seems to be alot of Talkers but very few Doers.

People that have no real credibility claiming geek cred for things they clearly aren’t good at  doing, and in the process have succeeded in leading the blind.

If you really are who you say you are, show your relevant track record in the open and be honest about what you can and can’t do. People respect you for that in the long run. Be clear and honest. Just be a Do-er and Stop Talking.