How AI Will Disrupt Incumbents and Erase Traditional Margins

Artificial Intelligence isn’t just another wave; it’s a tidal shift that will unbundle industries and redistribute value at speed. Much like the personal computer and mobile revolutions before it, AI will unlock massive new participation in markets. But here’s the paradox: while customers and activity will surge, traditional incumbents may find their margins evaporating. Why? Because of a dynamic economists have long understood, Jevons Paradox; applied now to digital work.

Jevons Paradox: More Efficiency, More Demand

In 1865, William Stanley Jevons observed something curious: as coal-fired steam engines became more efficient, coal consumption didn’t go down; it went up. Efficiency drove demand.

Today’s AI models represent an unprecedented leap in cognitive efficiency and a rapid commoditization of digital labor. Tasks that once needed specialized teams; content writing, customer support, data analysis, prototyping — can now be done faster, cheaper, and at scale by anyone with access to an LLM. As the cost of intelligence drops, demand for intelligent services will explode. But this demand won’t consolidate with big players — it will fragment.

The Disaggregation of the Incumbent

Mainframes: Centralized, scarce computing - single app. PCs: Personal software begins to scale - dozens of apps. Mobile: App explosion and niche use-cases flourish - thousands of apps. LLMs: Every task becomes a microservice. Every workflow is up for automation, personalization, or replacement.

This isn’t the era of a few giants growing ever larger. It’s the age of fragmentation. AI disassembles monolithic business units into modular services. Each piece can be optimized, resold, or replaced; often by startups or solo builders using open tools and APIs.

Think of a top-tier law firm. AI tools now handle contracts, research, and case law. That $200/hr associate? Being outperformed by a $20/month LLM. Now apply that across consulting, finance, healthcare, logistics, and media. The message is clear: scale is no longer a moat.

Customers aren’t just asking if your enterprise solution is better than your competitor’s. They’re asking why your $500k platform can’t match a $10 app in speed, simplicity, or results.

From Premium to Precarious

Take Adobe Photoshop, once the gold standard in content creation. Its edge came from power, flexibility, and ecosystem lock-in. But now:

Mobile apps offer AI-assisted editing, filters, and composition for a fraction of the cost. AI video generators handle editing, scriptwriting, and animation, disrupting studios and newsrooms. Solo creators are producing broadcast-quality content using cheap, powerful tools.

In 2023, the AI content creation market hit $10.5 billion. Video generation alone is projected to grow from $534 million in 2024 to over $2.5 billion by 2032. Even Adobe, despite launching Firefly, saw a 10% stock drop; a sign of investor anxiety.

Photoshop isn’t disappearing; but its dominance is. “One big suite fits all” is giving way to “many focused tools for each job.”

More Users, Squeezed Margins

AI will broaden the user base; but reduce revenue per user. Just as Google saw CPMs decline as ad inventory exploded, or Uber faced margin pressure despite scale, AI commoditizes what was once premium.

Margins don’t scale in this new environment; participation does.

Media companies will face infinite AI-generated content flooding their niches. Education providers will compete with personalized, interactive AI tutors. Consulting giants will lose ground to niche AI agents solving problems on demand.

This is not a “winner takes all” landscape. It’s a “many take slivers” world. And the slivers are getting thinner.

The Next Generation of Buyers

A silent shift is underway: the buyers have changed.

A new generation is entering the workforce; digital-native, AI-fluent, and allergic to legacy overhead. They aren’t planning decade-long ERP rollouts. They’re building with no-code tools, launching with AI copilots, and assembling workflows with public APIs.

Traditional enterprise sales relied on senior execs buying from other senior execs. But what happens when tomorrow’s CTO or COO is a 30-year-old ex-product manager who built three tools over the weekend?

We’re already seeing it:

A founder using ChatGPT for product specs, customer emails, and investor updates. A growth hacker plugging together Notion, Slack, and LLMs instead of buying CRM software. A high school student building a GPT-powered tutoring platform that outperforms legacy edtech tools. A junior developer publishing a niche project management app that overtakes enterprise software in user reviews. A designer using generative AI to create brand kits, pitch decks, and websites over a weekend — then monetizing via Substack or Gumroad.

Enterprise vendors who wait for RFPs may be bypassed entirely. The next enterprise isn’t bought — it’s assembled.

What Incumbents Must Do

Incumbents have two strategic paths:

1. Reinforce the Enterprise Moat - The Traditional Defense

Productize internal expertise as AI-augmented services. Microsoft embeds copilots in Office. ServiceNow enhances workflows with generative AI. Salesforce adds assistants to its CRM. Intuit uses AI in TurboTax and QuickBooks. Break down offerings into APIs or modular tools. Microsoft Azure exposes cognitive services. SAP’s BTP offers domain-specific APIs. Oracle turns insights into composable tools. Leverage brand and ecosystems. Adobe’s creative community remains sticky. Salesforce’s AppExchange locks in users. IBM holds trust in regulated sectors. Collaborate with specialists. Microsoft with OpenAI. SAP with NVIDIA. Adobe integrates Firefly into Figma. Oracle aligns with Cohere. Meet new decision-makers with API-first onboarding. Atlassian’s Forge platform. HubSpot’s AI copilots. Intuit’s developer tools.

This play is about staying central in a fragmented world. Own key workflows. Plug into emerging ones. Win through infrastructure, not interface.

2. Go AI-Native - The New Offense

Play like the disruptors; and beat them at their own game.

Incubate and spin out AI-native startups.

Instead of defending the core, incumbents can nurture internal AI-native ventures and spin them out. This avoids bottlenecks, maintains optionality, and allows innovation to thrive.

Amazon incubated AWS internally. Alphabet launched X and DeepMind. Shopify created Sidekick as a commerce assistant. Comcast is building AI-native personalization engines. Siemens’ Next47 funds AI startups in automation and energy.

The best way to compete with disruptors? Build one and let it move faster than your core ever could.

AI isn’t just a productivity tool. It’s a market multiplier and a margin destroyer. Like water finding cracks in a dam, AI will seep into every inefficiency and overpriced model. The flood is coming.

And it won’t just be technology driving it it’ll be people with new thinking. Tomorrow’s decision-makers will choose tools that feel like extensions of thought, not legacy-mandated systems.

Originally published on LinkedIn