This article was originally published on LinkedIn.

Lately I've been watching the chaos unfolding in the AI chat-client space, and I keep coming back to the same thought: this all feels a lot like the browser wars of the not-too-distant past. The patterns, the moves, the land grabs - it's almost a replay, just with a different cast of characters.

The real fight isn't about the app. It's about the platform.

Netscape and Internet Explorer weren't battling over who had the nicer UI. They were battling for the right to become the layer everything runs on.

AI clients are doing the exact same thing.

Behind the friendly chat bubbles, every vendor is trying to position itself as your:

  • reasoning engine
  • search engine
  • workflow engine
  • writing engine
  • planning engine
  • personal assistant
  • OS for getting things done

The browser became the application layer. Now AI clients are becoming the cognition layer - the place where work actually happens.

And that's a much more valuable position to occupy.

The war is won with distribution, not just features.

Microsoft didn't win because IE was the best browser. It won because it came preinstalled.

We're seeing the same playbook now:

  • iOS and macOS now integrate ChatGPT inside Apple Intelligence
  • Google is infusing Gemini across every app and surface it owns
  • Microsoft is weaving Copilot into Windows, Office, Edge, and everything in between
  • Meta is dropping its assistant into Instagram, Facebook, and WhatsApp

You don't need to be the best model. You just need to be the default doorway.

And the company that owns the default gateway owns the ecosystem.

Standards are once again a battlefield.

If you built for the early web, you remember the chaos: ActiveX, VBScript, HTML that only worked in one browser, JavaScript dialects, inconsistent CSS… it was wild and frustrating.

AI clients are repeating that mess:

  • different agent frameworks
  • different tool/plugin ecosystems
  • different API schemas
  • different prompt formats
  • incompatible memory systems
  • divergent UI metaphors

It's fragmented, fast-changing, and developer-hostile - just like the early web.

Eventually we'll get convergence. But right now, we're firmly in the chaos era of the AI standards war.

The pace of change is punishing, and it's creating real anxiety.

The browser wars were volatile. Every month something broke.

AI clients evolve even faster:

  • new models every few months
  • UI and UX overhauls every few weeks
  • frameworks rearchitecting themselves mid-release
  • capabilities that exist one day and disappear the next

Developers feel like they're standing on quicksand.

An entire ecosystem is trying to build a stable future on top of platforms that haven't even agreed on what stability means.

Hype is creating noise, confusion, and misalignment.

The late 90s had a "we need a web strategy" mania. Everyone built portals, homepages, toolbars, and web plugins nobody actually needed.

AI is in that same hype phase now:

  • "We need a chatbot."
  • "We need an AI assistant."
  • "We need our own model."
  • "We need to be doing something with agents."

There's more fear of missing out than real strategy. More urgency than understanding.

The signal is there - but so is a lot of noise.

The long-term winner becomes the gravitational center of software.

We saw this play out with Chrome. Once it won, it quietly became the true operating system for most people's computing lives.

AI clients will repeat this outcome.

The AI client that wins - or consolidates - will:

  • replace search
  • replace many apps
  • own your personal data/knowledge graph
  • store your memory
  • automate your workflows
  • provide your interface to the world

This is bigger than the browser ever was.

The browser reorganized how we access information. AI reorganizes how we think and work.

"Open vs closed" has returned with a new cast of characters.

In the browser wars, open standards fought proprietary systems.

In AI, we have:

  • open-weight models (Llama, Mistral, etc.)
  • open frameworks
  • open agent systems

…opposing tightly integrated, closed ecosystems from OpenAI, Anthropic, Google, Apple, and Microsoft.

It's the same philosophical divide. And we're watching the same tension play out - innovation vs control, freedom vs reliability, openness vs polish.

History may not repeat, but it certainly rhymes.

Lock-in is deeper this time, because the system actually knows you.

Switching browsers was inconvenient. You lost bookmarks, maybe a few plugins.

Switching AI clients means losing:

  • your memory
  • your preferences
  • your voice
  • your work history
  • your saved workflows
  • your custom agents/GPTs
  • your problem-solving patterns

Your AI isn't just a tool - it's quickly becoming your co-pilot for cognition.

That makes switching exponentially harder.

Everyone is racing toward full-stack control.

At the end of the browser wars, every browser had its own:

  • rendering engine
  • JavaScript engine
  • extension store
  • default integrations

AI clients are pushing even further:

  • model
  • tokenizer
  • client app
  • agent runtime
  • tools
  • execution environment
  • plugin ecosystem
  • persistent memory
  • app marketplace

They're not just building a product. They're building a universe.

The browser reshaped software. AI reshapes work.

This is the part that feels most significant.

The browser abstracted away the old desktop app model.

AI is abstracting away:

  • apps
  • screens
  • interfaces
  • navigation
  • decision trees
  • complexity

AI clients don't just help you use software. They become the interface to all software.

They're not rendering documents. They're rendering intention → output.

That's a much bigger jump.

The déjà vu is real, but the stakes are much higher.

We're not in another browser war. We're in a battle for the new default mechanism of human-computer interaction.

A battle for the layer where our thinking, planning, analysis, and creativity will live.

The browser changed the internet. AI will change work itself.

And whether you're developing software, implementing ERPs, building automations, or just trying to understand where the industry is headed, it's worth paying attention - because the patterns we're seeing now will determine the direction of the next decade.