The Dumbest Move in Tech Right Now: Laying Off Developers Because of AI
AI-assisted coding doesn’t mean fewer developers—it means more ambitious, higher-quality products
Buggy Software and Poor UX: Time to Raise the Bar
Let’s be real about the state of software today. Most products are, at best, “good enough” —unintuitive, buggy, and frustrating to use. Whether it’s consumer apps, enterprise software, or, even worse, developer tools, the experience is far from perfect despite all the claims about user-centric design, delightful experiences, empathy for the user, blah blah.
As a product manager, I've lived with the same painful reality for years: engineering is always the critical bottleneck. That new feature your users keep asking for—the one you know would drive revenue? Still marked as “coming soon.” That brilliant UX improvement that could increase adoption or engagement? "Definitely next quarter... probably." That bug driving your users crazy? "We'll triage it right after this sprint—promise.”
This isn't a criticism of developers—quite the opposite. It's simple math: there are roughly 29 million software developers worldwide serving over 5.4 billion internet users. That's one developer for every 186 users, each with unique requirements and preferences, all increasingly dependent on software for every aspect of their lives.
Now, with AI-assisted coding, we have an unprecedented opportunity to invest more (artificial) resources to dramatically improve software quality and user experience. Yet headlines are filled with executives viewing these emerging AI capabilities primarily as cost-cutting measures—a chance to achieve current output with fewer developers. This mindset fundamentally misunderstands AI’s true potential, which isn’t to maintain the status quo (of low quality products), but to amplify output by an order of magnitude.
Forward-thinking companies should use AI to transform each developer into a "10x developer," significantly boosting overall productivity and ultimately product quality. This isn't a time to replace teams with AI; it’s time to invest in them, to empower them with AI.
Laying off developers is a management and vision issue more than a reflection of AI advancement. Here is a challenge: show me your product and I guarantee I can point out ways to simplify its UX, highlight the bugs that frustrate users, suggest growth-driving features, add viral loops, and flag the unnecessary bloat holding it back. AI should be the catalyst for these improvements, not an excuse to cut costs.
Better products isn't a luxury—it's a business imperative that drives higher user engagement, premium feature adoption, lower churn, and ultimately increased revenue. But AI can't make these strategic product decisions alone and then execute them. It requires skilled, insightful developers collaborating with visionary product managers to deliver truly exceptional experiences.
Your mission as a leader isn't to maintain current output with reduced headcount—it's to leverage AI to deliver 10X value with your existing teams. The question isn't "How many developers can we cut?" but rather "What can we finally build now that the constraints have been lifted?"
Pent-Up Demand Across Every Vertical: The Untapped Opportunity
In 2021, I was an Entrepreneur in Residence at Khosla Ventures, exploring where the next major shift in software might come from and how to potentially incubate companies around it. This was before the AI boom, ChatGPT hasn't launched yet, but I had access to GPT-3 and early version of Codex, the first coding model from OpenAI.
The first two slides of my research deck captured what remains even more relevant today:
The enormous pent-up demand for specialized software across every conceivable vertical. Most businesses were—and still are—stuck with one-size-fits-all tools, siloed workflows, and fragmented data.
The severe constraint imposed by the relatively small number of developers compared to the number of business users constantly demanding specialized tools and workflows, let alone the massive number of end-users expecting deeper personalization and customization:
The demand exists at every level:
Workflows that remain manual and very difficult to automate with existing no-code tools
Vertical markets that operate on decades-old systems or, even worse, completely unserved
End-user experiences that expect customization and personalization to simplify their interaction with software.
My design work envisioned a new kind of platform combining AI-generated UIs, natural language logic, seamless access to business data, and early code generation via Codex to create a collaborative environment where developers and business users could work together. While the abstraction and simplicity of use of this platform was primarily targeting business users—clearly an underserved group—this was never about replacing developers. it was about increasing everyone's impact:
Helping engineers move faster and focus on core, deeply technical problems
Empowering business users to contribute directly to implement the workflow they need with the interfaces that best suits them
Giving end-users control over how they experience those tools in ways previously impossible
While this platform doesn't exist yet in full yet, the components of this vision—AI-driven UI generation, instant database integration, and coding assistants—are finally emerging. Tools like Vercel v0, Supabase, Cursor/Windsurf, and Bolt/Lovable/Replit are creating a world where more people can participate meaningfully in software creation—but only if developers remain central to the process.
Laying off engineers now, at the moment their leverage is finally compounding, is exactly backwards. If you think AI reduces your need for talent, you’ve missed the opportunity entirely.
The companies that hold onto their talent and use AI to amplify them—not replace them—are the ones that will capture vast, untapped software opportunity.
Reimagined AI-native Products: Win or Lose—Your Choice
Since OpenAI released ChatGPT in 2022, we've seen remarkable advances in AI technology. But let's be honest: genuinely innovative, human-centric products remain rare. Most companies just bolt superficial AI features onto existing workflows, delivering incremental improvements instead of true transformation. The classic “lipstick on a pig”.
Current AI models allow us to fundamentally rethink how products are imagined, designed, and built. True human-centric AI design demands a radical shift from creating isolated tools to deeply enhancing natural user experiences. New AI-native products won't just incrementally improve existing workflows; they will radically transform them. Every workflow, app or service can be redesigned from scratch—faster, simpler, more intuitive. Rather than forcing users to adapt to the rigid structures of traditional software, AI allows products to adapt to the way people naturally work, automate complex steps and eliminate unnecessary friction.
I experienced this firsthand while building a few AI-driven productivity tools. Multi-step processes that traditionally required navigating several screens and interfaces were collapsed into simple, intent-based, context-aware interactions. The result wasn't just a faster way to do the same old things—it was an entirely new way of working that felt almost magical to users.
This kind of reinvention is not only possible but necessary across virtually every category—from legacy enterprise software to consumer apps, from underserved verticals still running on spreadsheets at best, to entire industries yet to be meaningfully digitized or automated.
The implications extend far beyond incremental improvements. We're witnessing the early stages of a fundamental shift in how humans interact with technology. Just as the smartphone revolution rendered many web-first companies obsolete, this AI revolution will create winners and losers based on who truly understands the paradigm shift.
Look at what's happening to search, a market monopoly that seemed unassailable for decades. For the first time in 20+ years, traditional search is declining as users discover new ways to access information through AI assistants rather than keyword-driven search engines.
This pattern will repeat across countless industries, with incumbent leaders suddenly finding themselves fighting for relevance.
The strategic imperative is clear: this isn't the time to cut development teams—it's time to double or triple down on transformative AI-native products. Companies must embrace this opportunity to rebuild their offerings from the ground up—not merely adding AI as a feature, but integrating it as the foundation. The question isn't whether these new products and experiences will emerge, but who will create them.
If your company isn't proactively engaging in this shift, competitors—established players or agile startups—surely will. The barriers to entry have never been lower, with powerful models accessible to nearly anyone with development skills. A small team of AI-fluent developers can now create products that would have required hundreds of engineers just a few years ago.
The companies that will dominate the next decade aren't the ones optimizing for short-term cost savings by reducing headcount. They're the ones reimagining what's possible with AI at the core, investing aggressively in their development capabilities, and racing to deliver experiences that will make current offerings seem as outdated as flip phones in the age of smartphones.
In short, yes, it’s time to build.