arrow_back The AI Pravda
#57

AI kills websites

People love bots

987 million people are using AI chatbots in 2025, according to data research firm DemandSage. Published academic research comparing ChatGPT to Google Search show users spend less time on information tasks with AI while reporting higher satisfaction, usefulness, and enjoyment. This isn't convenience. It's a fundamental shift in how people think about information access.

Traditional web browsing forces users to navigate multiple pages, parse different formats, synthesize information from various sources. AI agents eliminate this friction. They deliver exactly what users need in their preferred format.

No more visiting review sites, reading through multiple pages, comparing ratings manually. Instead: "What's the best Italian restaurant near downtown for a romantic dinner under $100?" One comprehensive, contextual response.

This efficiency advantage drives rapid adoption despite the technology's newness. The global chatbot market grows from $8.27 billion in 2024 to $10.32 billion by 2025. Retail spending on chatbots jumps from $12 billion to $72 billion by 2028. [DemandSage, 2025]

ChatGPT reached 100 million weekly active users within a year of launch. Faster than smartphones. Faster than social media. Faster than any previous consumer technology.

Tech is getting ready

Leading companies are building the infrastructure. Microsoft's NLWeb protocol, released May 2025, turns existing websites into AI-accessible databases without rebuilding architecture. It leverages Schema.org markup and RSS feeds that many sites already use. Converts them into natural language interfaces that AI agents can query directly.

O'Reilly Media proves this works. Their platform with 59,000 books now supports conversational searches like "books by Uncle Bob"—recognizing Robert C. Martin's nickname automatically. Implementation took "just a few days" because it leveraged existing structured data.

Anthropic's Model Context Protocol gained adoption across 5,000+ active servers. AI agents access everything from GitHub repositories to Google Drive documents through standardized interfaces. OpenAI, Google DeepMind, Zed, and Replit integrated MCP support. An ecosystem where AI agents seamlessly interact with web-based data sources.

Web-sites as we know 'em, will be dead soon

The transformation ends 30+ years of internet history. Online media followed the same model as traditional media, except for the medium. Online shopping was the same shopping as offline, just in a different environment.

The central play for web inhabitants—online media, retail, corporate sites—was always disintermediation. Establishing direct rich contact with readers, buyers, clients. This direct contact enabled better experience for site owners and visitors. It enabled strong business models when site owners could identify and monetize visitor intent.

U.S. digital advertising revenue reached $258.6 billion in 2024 [IAB, 2025] Much of this depends on users clicking through to publisher websites.

Google web search was the largest winner. It's rapidly becoming irrelevant. Early data shows AI search reduces this traffic significantly. Google's AI Overviews cause an 8.9% average reduction in clicks to cited URLs [Search Engine Land, 2024].

We're entering a new era. Websites become databases that users access through intermediaries. First group: AI agents mixing "cocktails" of information from multiple sources. Second group: tools and services built and presented by AI industry.

This new status quo kills websites, as it deprives its teams of purpose: the connection with site visitors. It also deprives them of money, killing the economic model.

AI: "please, don't worry"

When I asked Claude to research this topic, it spent 14 minutes thinking and returned characteristically verbose politically correct formula. "The consumer-facing web is undergoing its most fundamental transformation since the shift from desktop to mobile, as AI agents increasingly mediate how users access information and complete tasks online. Rather than 'killing' the web, AI is evolving it from a browsing-based interface into an intelligent, agent-accessible data ecosystem — a shift that, while disruptive, ultimately serves users better."

It claimed this transformation "benefits users by delivering information in optimal formats while creating new economic opportunities for content creators and service providers."

Claude's research referenced Deloitte's Global Predictions of 2025. A timeline of gradual transformation rather than sudden disruption. In the immediate term (2025-2027), approximately 20-30% of consumer-facing websites become AI-accessible as major platforms integrate these protocols. E-commerce sites and information publishers lead adoption. More complex services like healthcare and financial services follow as regulatory frameworks develop.

Claude have tried to convince me that online media switches to new monetization models that may prove more valuable than traditional advertising. Publishers secure direct licensing deals worth $1-5 million annually with AI companies. Revenue-sharing programs offer 50% splits of AI platform revenues. News Corp's five-year deal with OpenAI is valued at $250 million. Academic publisher Wiley generated $44 million from AI training agreements in a single year.

But Claude is sugarcoating simple reality: eventually, AI will kill web-sites.

Mikael Alemu Gorsky

Mikael Alemu is an educator and researcher, and the author of two programs: Agentic Software Engineering, on building software with AI agents, and Building AI-Native Agentic Systems, on building software that thinks.

He teaches at the Holon Institute of Technology, near Tel Aviv, where Agentic Software Engineering runs as a credit-bearing course. He is an educator and researcher.

Nine published works, 76 citations. A 350-page textbook under contract with a major academic publisher.

Teaching and programs

Agentic Software Engineering — program, preprint and textbook

The discipline of structured, auditable human-agent workflows for building software. The human frames, specifies and judges. The agent executes. Nineteen modules in four parts, built on a running project called Tribunal, a web application in which agents argue opposing sides of a case and a judge agent decides. Taught for credit at the Holon Institute of Technology.

Agentic Software Engineering curriculum

Building AI-Native Agentic Systems — program, paper and book in writing

How to build systems that hold a language model as a working component, and treat that component as what it is: stochastic, slow and metered. Fourteen modules in four parts, about seventy hours. The running project is the Observatory, a news agency that watches sources, selects what matters and publishes on a cadence.

Research and analytics

Publications — journals and proceedings

Nine works, 76 citations. Research on artificial intelligence in education, with Ilya Levin and Alexei Semenov.

The AI Pravda — LinkedIn newsletter

Critical analysis of artificial intelligence and its effect on work and society. 5,500+ subscribers. The complete archive of 103 issues (2023–2026) is published in full at mgorsky.net/theaipravda.

Subscribe to The AI Pravda on LinkedIn

Pro bono

AI for seniors — free workshop

Helping older adults use everyday AI tools. Delivered to Russian-speaking communities in Israel.

For older adults, artificial intelligence is about preserving quality of life, maintaining autonomy, and sustaining the feeling of independence that defines dignified aging. For seniors who have emigrated, AI becomes a bridge: it can translate documents, explain official letters, help compose emails in the local language, and guide users through government websites. The workshop has been delivered to Russian-speaking communities in Israel, where participants — many of them in their 70s and 80s — discovered that AI could help them read Hebrew documents and communicate with Israeli institutions.

Startup competitions — unpaid time

Judging and mentoring early-stage ventures. Helping teams clarify their value proposition, assess technical feasibility, and prepare for the realities of scaling an AI product.

AC/VC LinkedIn group — community

A group for developers and students working with coding agents. The community shares practical insights, code examples, tool comparisons, and honest assessments of what works in production.

Join the AC/VC LinkedIn group

Recent

Important Links