Build Now or Be Fired

SUMMARY

In a 2026 monologue, software builder Jordan expresses alarm over AI tools like Claude Code slashing development costs, enabling rapid custom analyses of 250,000 websites for under $300, threatening SaaS viability and urging imaginative building.

STATEMENTS

The speaker builds software primarily using voice commands, with dramatically reduced costs, but notes the enormous yet fast-fixable QA required in these AI-driven systems.
For a client seeking a top-down TAM analysis, the speaker used Claude Code to download and analyze 250,000 websites' homepages, identifying technology vendors and URL patterns for just $300 total.
AI tools like Claude's Cloud Batch API enable pattern identification in HTML at minimal cost, such as $6 for processing 250,000 pages, outperforming traditional tools like BuiltWith.
Data enrichment costs are plummeting toward zero, with tools like Million Verifier checking emails for pennies, allowing one-time builds to yield perpetual rewards.
Value in future software will stem from validation and APIs providing non-public data, like historical or blended private info, which cheapen as public web structuring becomes easy.
Horizontal SaaS companies targeting other SaaS firms face imminent disruption, as clients build in-house solutions tailored to their needs using AI, bypassing vendor software.
Vertical SaaS for industries like plumbing or oil has more runway, as non-tech users won't adopt "vibe coding" soon, but all must focus on unique, scale-derived benefits.
Network effects, as discussed by NFX, will be crucial; services like FullEnrich's caching of prior queries without extra charges exemplify value from user scale.
The era of hiding behind software complexity ends, limited now by imagination; anyone, even without technical skills, can build what's computable.
To thrive, individuals must embrace tools like Claude Code, asking what they'd create with unlimited resources, fostering recursive intelligence for explanation, teaching, and creation.

IDEAS

AI enables scraping and analyzing 250,000 websites for a comprehensive TAM in hours, using just a $200/month subscription and minimal cloud costs.
Traditional high-cost consulting like McKinsey-style market positioning can now be done near-free by individuals with basic setups, democratizing intelligence.
Prompting AI to generate regex patterns from URL observations automates scalable data extraction, turning static homepages into live competitor insights.
SaaS vendors must pivot to irreplaceable value like cached data reuse or private blends, as public web data becomes trivially accessible.
Horizontal SaaS targeting tech-savvy firms will collapse first, as in-house AI builds create perfect-fit tools without security risks from vendors.
Vertical industries gain time, but eventual disruption looms unless providers leverage network effects for personalized, scale-only features.
Building once with AI yields forever rewards, as seen in email verification tools charging fractions of a cent per check.
Imagination becomes the sole barrier; software can now handle any computable task, unshackling non-engineers from technical limits.
QA, security, and deployment risks persist but are mitigable by the tools themselves, turning pains into solvable challenges.
Recursive intelligence from conversational AI allows users to learn, explain, translate, and iterate creations in real-time.
Big SaaS faces rough changes as sidelined individuals build custom solutions, accelerating in-house shifts.
Future winners combine bold problem imagination with fluent "alien tool" communication to manifest ideas.
Employees risk obsolescence more than ever without hands-on AI experimentation, in a tech economy rewarding adaptability.
Even amid rapid change, foundational joys like sunlight persist, grounding technological upheaval.
Crisis of imagination demands envisioning builds with unlimited time, resources, and information for human flourishing.

INSIGHTS

AI's cost reductions in data tasks erode SaaS moats, forcing a shift toward validation and proprietary data APIs that outpace individual scraping efforts.
Network effects redefine software value, where scale enables unique services like free cached enrichments, inaccessible to solo builders.
Horizontal SaaS disruption accelerates as AI empowers tailored in-house solutions, prioritizing security and fit over vendor generality.
Vertical markets' temporary safety relies on users' tech aversion, but long-term survival hinges on irreplaceable, domain-specific scale benefits.
The imagination bottleneck emerges as AI removes technical barriers, challenging society to dream computably vast ambitions without prior limits.
Recursive AI interaction fosters continuous learning and creation, turning users into iterative innovators who mitigate risks through dialogue.
Public data commoditization drives private, historical, or blended info premiums, making cheap APIs competitive necessities.
Employee vulnerability heightens in AI eras; proactive tool mastery blends creativity with execution for economic edge.

QUOTES

"the amount of QA that you have to do in these systems is enormous. But it's also fast fixes."
"250,000 websites cost me $6. So 200. So now we're like 200 bucks plus six bucks and I don't know maybe Amplify is like another 50 $300 close to $300."
"data costs are going to come down close to zero. I mean, you see tools like Million Verifier that will go verify emails for like 02."
"We will no longer live in a world where people can hide behind the complexity of the software that they have built because it will become so much easier to build."
"if we're limited by our imagination, the question is what is the biggest ambition we could possibly have?"

HABITS

Building software primarily through voice commands to leverage AI efficiency and reduce manual coding time.
Iterating AI prompts locally on a computer for contextual, task-specific analysis without external dependencies.
Challenging oneself to complete complex tasks like TAM analysis near-free, using only subscription tools and personal hardware.
Regularly validating AI outputs with live scripts to ensure accuracy, even from reliable sources like Exa.
Encouraging clients to experiment with AI tools like Claude Code for safe stumbling and non-catastrophic learning.
Using recursive questioning of AI to explain code, teach concepts, and translate jargon for deeper understanding.

FACTS

Processing 250,000 HTML homepages via Claude's Cloud Batch API costs just $6, 50% cheaper than standard rates.
AI subscriptions like Claude Code run $200 monthly, enabling solo analysis rivaling expensive consulting firms.
Tools like Million Verifier validate emails for $0.02 each, exemplifying plummeting data enrichment costs.
Technology sector comprises 10% of the economy, highlighting software's vast potential when computable.
FullEnrich caches user-queried emails without recharging, reusing prior data provider payments for efficiency.

REFERENCES

Claude Code ($200/month AI tool for building and analysis).
Cloud Batch API (50% cheaper for batch HTML processing).
Appify (handles downloads bypassing protections like Cloudflare).
BuiltWith (traditional tool for tech detection, now outpaced by live AI).
Million Verifier (email validation at $0.02 per check).
Exa (search tool for finding people, requiring validation).
FullEnrich (caches prior email queries without extra charges).
NFX (VC firm discussing network effects in scale).
McKinsey (style of top-down TAM and market positioning analysis).

HOW TO APPLY

Start by subscribing to an AI tool like Claude Code and experiment with voice-based building to internalize its conversational flow for rapid prototyping.
For market analysis, prompt the AI to download and store local copies of target websites' homepages, handling 80% directly and using cloud services for the rest to build a dataset efficiently.
Develop custom prompts to extract patterns from HTML, such as technology vendors and URL structures, including regex suggestions, then batch-process via cost-effective APIs for scalable insights.
Iterate locally on your machine with full task context to refine outputs, validating against client needs and running live scripts to confirm accuracy before deployment.
Embrace recursive intelligence by querying the AI to explain its code, teach unfamiliar concepts, or translate outputs, mitigating QA and security risks through ongoing dialogue.
Focus on network effects in your builds, like caching repeated queries without recharges, to create scale-dependent value that individuals can't replicate solo.

ONE-SENTENCE TAKEAWAY

Embrace AI tools to unleash imagination, building custom software now before SaaS disruptions sideline the complacent.

RECOMMENDATIONS

Pivot SaaS strategies toward validation APIs and private data blends, competing on irreplaceable access rather than public scraping ease.
Cultivate a crisis of ambition by daily prompting AI with "unlimited resources" scenarios to expand creative horizons.
Train teams in Claude Code for in-house experimentation, emphasizing safe failures to build AI fluency and reduce vendor dependency.
Seek network effects through user-scale features, like free cached data, to deliver personalized value beyond solo capabilities.
Prioritize vertical niches initially for breathing room, then innovate domain-specific benefits that leverage collective data without tech-savvy demands.

MEMO

In the dim glow of a screen on a crisp February morning in 2026, software innovator Jordan confronts a seismic shift in his industry. What once demanded armies of engineers and bloated budgets can now be conjured with a voice prompt and a $200 subscription. Jordan recounts downloading 250,000 websites— the digital footprint of an entire market— for a client's TAM analysis. Using Claude Code, he sidestepped traditional tools like BuiltWith, processing HTML via the Cloud Batch API for a mere $6. "It's amazing," he admits, "but I don't know where data enrichment goes from here." This feat, totaling under $300 including cloud tweaks for stubborn sites, underscores AI's ruthless efficiency, turning exhaustive research into an afternoon's work.

Yet triumph breeds trepidation. Jordan, whose livelihood orbits SaaS in its San Francisco epicenter, warns of an impending quake. Horizontal SaaS firms peddling to other tech outfits are first in the crosshairs; why subscribe when AI lets clients craft bespoke tools in-house, ironclad against security lapses? Vertical players in plumbing or oil rigs buy time—their users shun "vibe coding." But all must unearth defensible moats: not greed or stinginess, but unassailable truths via validation and APIs hoarding historical or private data. Echoing NFX's gospel on network effects, Jordan praises FullEnrich's sleight—caching your prior email hunts gratis, a scale perk no lone coder can match. As public data nears free, software's edge sharpens on the intimate, the blended, the once-ungraspable.

The real peril, Jordan posits, is imagination's famine. Engineers once battled time and syntax; now, anyone sans code chops can summon software for any computable end—10% of the economy's lifeblood. "We have a crisis of imagination," he declares, urging a mental unshackling: What worlds would you forge with infinite time, resources, infinite info? QA gnaws, security looms, deployment tempts fate—yet AI itself mends these wounds, explaining its esoterica in plain tongue. Jordan drills his clients in Claude Code, fostering recursive smarts: query to learn, to create, to iterate sans catastrophe.

This renaissance terrifies Big SaaS, priming rough upheavals as sidelined souls seize the forge. Employees, beware: in tomorrow's tech bazaar, victors wed audacious vision to alien-tool rapport. "It's a beautiful time to be alive," Jordan muses, "a scary time for big SaaS." Step forth, he implores—build, stumble, thrive. Outside, the sun endures, a steadfast amid the code storm, reminding that no algorithm eclipses nature's quiet gleam.