Page 12 - GBC Spring 2026 ENG

Instead of simply indexing pages
and pointing users to them, LLMs
ingest massive amounts of text,
learn patterns in language, and
generate direct answers to
questions. In other words, they
don’t just find information; they
synthesize it.
This distinction matters. When
a golfer asks an LLM a question –
“What are the best public courses near
Niagara Falls?” or “Where can I book
a tee time this Saturday?” – they are
often given a complete answer
without ever visiting a website.
Search engines themselves are
moving in this direction. Google’s
AI Overview feature now reads
content from across the web and
presents a summarized answer at
the top of the search results, before
traditional listings even appear.
For many users, that answer is
“good enough,” eliminating the
need to click further.
For operators, the implication
is straightforward and uncom-
fortable: Less traffic is reaching
course websites, even when interest
in golf remains strong. This isn’t a
reflection of declining demand. It’s
a structural change in how
information is delivered. And it’s
only the beginning.
WHY LLMS CHANGE THE
ECONOMICS OF DISCOVERY
To understand why this shift is so
disruptive, it’s important to
recognize what LLMs actually
“know”.
Traditional search engines
store summaries and signals about
pages, such as keywords, links,
metadata, etc. LLMs, by contrast,
are trained on vast amounts of
language. They effectively absorb
the words on websites, articles,
reviews, and documentation across
the internet and learn how concepts
relate to one another. That means
your website content is no longer
12
Golf Business Canada
Instead of telling a golfer where they might book a tee time, an AI agent
can actually navigate websites, check availability, log in, and complete a
reservation.
just something a human reads. It’s something machines actively consume
and reuse to answer questions.
From a golfer’s perspective, this is incredibly convenient. Instead of
opening multiple tabs and comparing options, they receive a single
response. From an operator’s perspective, however, it creates a new
challenge: the website is no longer guaranteed to be the moment of
engagement.
The carefully designed homepage. The compelling photography.
The thoughtfully written experience descriptions. All of it may be
bypassed. This doesn’t mean websites no longer matter. But it does mean
their role is changing – from being the destination to being a source.
THE RISE OF AGENTIC AI
If LLMs answering questions marks the first phase of disruption, Agentic
AI represents the next – and more consequential – phase.
Agentic AI refers to systems that do not just provide information but
can take action on behalf of a user. Instead of telling a golfer where they
might book a tee time, an AI agent can actually navigate websites, check
availability, log in, and complete a reservation. That application might
reside on your phone or desktop computer, and it might run within an
advanced form of a browser. In practical terms, this changes the flow from:
User → Browser → Website → Booking
to something much closer to:
User → AI Agent → Booking Completed
Some early versions of this already exist. AI-powered browsers can
scroll pages, click links, evaluate options, and present availability back to
the user. Major players in the AI and LLM space are already making the
leap from text-based chatbots to action-focused Agentic Browsers.
ChatGPT and Perplexity launched their browsers - Atlas and Comet - in
2025. Smaller startups, like Fellou and Sigma AI, are also launching their
own Agentic Browsers. As integrations improve, these agents will
increasingly be able to complete transactions end-to-end.