Neo
Do you always look at it encoded?
Cypher
Well you have to the image translators
work for the construct program but
there's way too much information to
decode the Matrix. You get used to it.
I, I don't even see the code.
All I see is blonde, brunette, redhead.
At the dawn of Internet commerce, Berners-Lee invented a simple document exchange and reader. Microsoft, then the gatekeeper to computer users with its flawed Windows 3.1 along with slow dial-up birthed CSS/JS, which turned the browser into a massive local executable to handle rendering and state. That historical accident created every front-end dev career.
Built for ambiguity, human-centric computing—web GUIs, touch screens, the Matrix of our present reality—will be going away.
Buckle your seatbelts Dorothy, because Kansas is going bye-bye.
The future has no need for your front-end frameworks, UI libraries, design systems, and accessibility overlays. When that Matrix disappears, your hard-earned visual presentation layer skills will become worthless.
The fast-coming future is a high-utility, agentic system with the conversational agent becoming the singular interface to reality.
- Why render a webpage when an agent only needs structured data?
- Why build a GUI when intent can be expressed directly?
- Why maintain CSS animations when the endpoint is pure information exchange?
When firms start exposing capabilities through high-utility data streams built on negotiation protocols, webpages and the HTML/CSS/JS stack becomes worthless.
Today's human-driven, request-response endless click loop gets replaced by tomorrow's agent-to-agent instant, zero-friction negotiation:
Agent A: "Purchase flight under $500 for Thursday"
Agent B: "Options at $380, $420, $495. Preferences?"
Agent A: "Book $380, window seat"
Your personal agent—the digital projection of your mental self—will interact with the world not through buttons and forms, but through executable intent.
From that moment, the web browser ceases to be the gateway. Agents communicate directly. Marketing clutter disappears—content generates only when requested. Visual design becomes optional—clarity and structure dominate.
The future UI is conversational. Audio becomes the primary interface not because screens disappear, but because higher-utility conversation beats clicking. Why scroll through menus when you can ask? Why watch canned videos when your agent can generate bespoke explanations in real-time?
"Hey Agent, negotiate with the car dealership for that EV."
[Agent negotiates via structured protocols]
"Deal secured at $38,500. Want me to generate a 2-minute summary video of the features?"
"Just audio, while I drive."
[AI generates bespoke audio explaining the deal, features, delivery]
The future content is bespoke, on-demand AI-generated video and audio. Gone are pre-recorded sales videos, marketing explainers, generic tutorials.
"Agent, explain quantum entanglement to me like I'm 15"
→ AI generates 3-minute audio tailored to your knowledge level
"Show me how to fix this leak"
→ AI generates 60-second video of YOUR specific plumbing setup
"Summarize today's market movements in my portfolio"
→ AI generates personalized financial briefing
Marketing becomes permission-based discovery. Education becomes personalized tutoring. Entertainment becomes on-demand generation. The entire content industry—built on creating once, distributing many—shifts to generating uniquely for each request.
Not the death of media, rather it is the personalization of media at scale. And it requires systems built for deterministic performance, not visual decoration. The languages and protocols that win will be those that enable reliable, predictable agent negotiation—not those that render the prettiest buttons.
The Agentic Economy: 10-Year Horizon
❌ All of this is going away within 10 years:
- Web servers (like FTP and Gopher before those)
- Web browsers
- Front-end frameworks
- JavaScript SPAs
- Visual design systems
- "User experience" as we know it
✅ Replaced by:
- Agent negotiation protocols
- Structured data dialects
- Intent-based APIs
- On-demand media generation
- Zero-friction commerce
- Pure utility interfaces
The Technical Implication: Back to Basics
This future requires deterministic, minimal systems—not more abstraction layers. Languages like Hare and Odin (not React/TypeScript) become relevant because such langs:
- Expose machine physics (deterministic execution)
- Minimize abstraction (clear resource management)
- Enable efficient agent runtimes (predictable performance)
The future belongs to systems that do one thing well and communicate clearly.
The Human Implication: Sovereignty Returns
With agents handling negotiation:
- Your time is reclaimed from browsing
- Your attention is no longer monetized
- Your intent drives outcomes directly
- Your data stays local unless negotiated away
The visual web's complexity never has been about serving humans—it has been about capturing attention. The agentic future will free us from the Big Tech Matrix. Gone will be the attention economy.
Building the Future
This is not speculation—it is inevitable physics:
- Agents are more efficient than humans at parsing structured data
- Visual interfaces are inefficient for machine communication
- Complexity has maintenance costs that simple protocols avoid
- Deterministic systems outperform probabilistic GUIs for automation
The question is not "if" but "who builds it." And it will not be built with today's web stack. It will be built with systems that respect machine physics, enable deterministic behavior, and prioritize utility over decoration.
The choice is binary: Continue decorating the Matrix, or learn to read the code. The blonde-brunette-redhead interface is dying. The deterministic, agentic future is being built right now—in languages like Hare, Odin, and Zig, with systems thinking that starts from machine physics, not visual design. The collapse is coming. Will you be looking at blonde-brunette-redhead when it arrives, or will you see the code?
Acknowledgments
This piece exists because @sylwia-lask challenged me to expand a conversation into a full article. Her exact words: "Your view is unique, and it deserves its own space."
About me
As a Systems Ontologist, my work exists at the intersection of:
- What exists — current technological, economic, and psychological realities
- What could exist — feasible, high-probability futures
- How to get there — practical engineering and strategic pathways
- What might break — systemic risks and unintended consequences
I run a skunkworks focusing on agentic architecture—the protocols, runtimes, control systems, the incentives and payoffs that will replace the visual web.
I consult, build, and write about the deterministic systems that will define the next era—where agents, not browsers, mediate reality. I do not predict the future; I extract the signals from today's emerging technologies that are already deciding it.
Top comments (3)
This was such a fun read - I absolutely love your enthusiasm and how boldly you explore these ideas. Even if I don’t fully share your apocalyptic vision of HTML and UI disappearing anytime soon 😄, I really appreciate the perspective and the passion behind it.
And honestly, I’m glad you turned your thoughts into a full post - they definitely deserve their own space, not just a comment thread. Keep them coming!
Thanks—I appreciate you engaging with the ideas. Let's define terms concretely, because timeline disagreements often mask definition disagreements.
I'm forecasting functional irrelevance of the visual web stack for new systems within 5-7 years. Not that government sites disappear, but that zero new economic value is created with HTML/CSS/JS after that point. Like COBOL today: still runs, but no startups choose it.
AI doesn't just generate PWAs faster—it collapses the skill moat. When one person + AI can output what required 10 frontend devs, the economic justification for those roles evaporates. This isn't about technology adoption curves; it's about labor arbitrage at AI scale.
If your entire career has been inside the browser-as-runtime paradigm, the transition feels unimaginable. But consider:
I remember installing software from floppies, using BBSs, navigating Gopher. The shift from local executables → browser-as-runtime happened in ~8 years (1995-2003). We're now at the browser-as-runtime → agent-as-interface inflection.
Most consumed video already bypasses the browser stack. Binary → native player. Why? Performance matters more than platform loyalty.
WoW, Steam, consoles—all thick clients with lightweight sync protocols. They solved the 'local/remote' problem 20 years ago without browsers.
The compression isn't linear—it's parabolic. AI advancing at 2× Moore's Law means capability doubles every 9 months. HTML/CSS/JS being 'easy to gen' is actually the death knell: when the output is commoditized, the specialization becomes worthless.
'The browser will never replace native applications.' That's 1998 talking.
Sound familiar? We're at that exact inflection—but this time moving from browser to agent. The people who said 'anytime soon?' about native apps in 1998 were Netscape employees. Where are they now?
What would have to be true for you to shift your 'anytime soon' estimate forward? What observable signal would change your mind?"
@sylwia-lask Have you seen this?