LLM Search Disrupts Google’s Business
Key Points
- Large language models (LLMs) are now delivering search experiences that can shift substantial value away from Google, offering ad‑free, highly actionable results.
- Demonstrations with an LLM (referred to as “O3”) showed it can instantly provide detailed ticket information, flight options, booking strategies, and logistical tips—features that Google’s standard search and services don’t bundle together.
- The “bias for action” of this model is unusually strong, delivering step‑by‑step recommendations that go beyond simple answers, making it more useful for planning tasks.
- Although currently accessed via a paid Pro account, the presenter predicts that similar LLM capabilities will become widely available within months, threatening Google’s ad‑driven revenue model.
- By handling diverse queries (sports tickets, travel, health advice) in a unified, strategy‑focused interface, LLM‑based search could fundamentally disrupt many of Google’s core search verticals.
Sections
- LLM Search Challenges Google - The speaker demonstrates an ad‑free, highly actionable LLM‑powered search experience that provides detailed results like ticket information, arguing it could shift significant search value away from Google’s traditional model.
- LLMs vs Google for Complex Queries - The speaker argues that large language models retrieve obscure, nuanced information—like identifying an 80s movie with a robot kid and SR‑71—more reliably than traditional Google search, and reflects on using longer, natural‑language prompts for AI learning and search integration.
- LLM Search Disruption Threat - The speaker warns that advanced LLM-powered search could rapidly erode traditional search market share, posing a looming competitive threat to Google.
Full Transcript
# LLM Search Disrupts Google’s Business **Source:** [https://www.youtube.com/watch?v=dlwODfg4eE0](https://www.youtube.com/watch?v=dlwODfg4eE0) **Duration:** 00:08:14 ## Summary - Large language models (LLMs) are now delivering search experiences that can shift substantial value away from Google, offering ad‑free, highly actionable results. - Demonstrations with an LLM (referred to as “O3”) showed it can instantly provide detailed ticket information, flight options, booking strategies, and logistical tips—features that Google’s standard search and services don’t bundle together. - The “bias for action” of this model is unusually strong, delivering step‑by‑step recommendations that go beyond simple answers, making it more useful for planning tasks. - Although currently accessed via a paid Pro account, the presenter predicts that similar LLM capabilities will become widely available within months, threatening Google’s ad‑driven revenue model. - By handling diverse queries (sports tickets, travel, health advice) in a unified, strategy‑focused interface, LLM‑based search could fundamentally disrupt many of Google’s core search verticals. ## Sections - [00:00:00](https://www.youtube.com/watch?v=dlwODfg4eE0&t=0s) **LLM Search Challenges Google** - The speaker demonstrates an ad‑free, highly actionable LLM‑powered search experience that provides detailed results like ticket information, arguing it could shift significant search value away from Google’s traditional model. - [00:03:56](https://www.youtube.com/watch?v=dlwODfg4eE0&t=236s) **LLMs vs Google for Complex Queries** - The speaker argues that large language models retrieve obscure, nuanced information—like identifying an 80s movie with a robot kid and SR‑71—more reliably than traditional Google search, and reflects on using longer, natural‑language prompts for AI learning and search integration. - [00:07:06](https://www.youtube.com/watch?v=dlwODfg4eE0&t=426s) **LLM Search Disruption Threat** - The speaker warns that advanced LLM-powered search could rapidly erode traditional search market share, posing a looming competitive threat to Google. ## Full Transcript
All right, I want to do one of those
rare Nate shows the screen videos
because I want to talk about something
that I think is getting solved that
hasn't really been properly solved until
now. We've talked about chat GPT and
other large language models disrupting
search, but I feel like we are at the
point where you can reasonably shift a
lot of the search value out of Google
and that is a big big deal. I'm not
saying tomorrow that Google has no value
and I'm not saying Google doesn't
respond. I'm well aware that the cost
curve on serving LLMs is dominated by
Google. Google has some very sharp minds
working on LLMs, but the cost to disrupt
their own core search experience is not
something I'm sure they're willing to
bear because as you can see here, this
is an LLM search results page that has
no ads. That does not suit Google's
business model and yet it's very useful.
I just threw a few uh queries we're
going to run through. I'm a Michigan
fan. Can you find me Michigan tickets
for next season? I'd love to see the
Ohio State game. So, I see the game
time, the date, the venue. It's going to
be at home at Michigan Stadium. It gives
me the average price. It then sort of
goes beyond that. This is a very 03
thing to do. Tells me how to lock in the
tickets. Gives me different
strategies. Gives me tips to go to the
game in Ann Arbor. gives me a pick list
of next steps. I tell you what, 03 has
the highest bias for action of any model
I've seen. It's higher than my bias for
action. And that is saying
something. And I say, great. I love
this. Uh, can you get me Smart Flights
from Seattle for this game? What are my
options? Now, this is really interesting
because I'm deliberately trying to push
it. I would not say this is as good as
Google Flights by any means, but I think
it's enough that it is going to disrupt
some of the top offunnel search as chat
GPT awareness begins to spread. You have
to realize, you know, I get 03 now
because I'm paying for pro, but in 6
months, everyone's going to have lots
and lots of O3. And this kind of
experience is going to be possible for
everybody at
scale. And with that in mind, this kind
of strategic here's your options to
think about is super useful and frankly
in some ways easier to sort through than
the wall of flights that Google
provides. Again, no ads here. deeply
disruptive of Google's business. So then
it gives me the smart booking moves. It
gives me the strategy. Google doesn't do
any of this. It gives me ground and
logistics tips. Again, Google's just
trying to sell me rental cars to to buy
at this
point. Um, and then I I deliberately
switch it up. I'm like, "Okay, let's try
a health one. What are some quick
exercises I can do to improve my back
health?" Because this is one of those
things where like I don't just want to
fall into the trap of being like, "Oh,
yeah. booking flights. That's why people
use Google. People use Google for lots
of other stuff. So, I wanted to give a
real rant here. So, I think the the
weakness here where I think that OpenAI
can and probably will improve is there's
no like demo video, no diagram. Now,
Google's a bit disorganized, but I will
say their image search and so on will
get you diagrams of all these exercises
in the first click. And that's something
that's missing here that I anticipate
OpenAI is going to close, which is the
other reason to do an exercise like this
periodically. It basically gives you a
cheat sheet into what OpenAI is going to
be thinking about from a product
perspective because Sam is very vocally
going after Google's cake um and wants
to grab that search
spot. But even without the without the
diagrams, this is already very useful. I
have clear names. I have reps. I have
quick form cues, which is actually very
difficult to get on Google. I have a how
to use it in one clear place. It's that
clarity of like a single answer that's
opinionated that I think that Google is
really missing. And I know they're
trying it with the Google summaries.
It's just not as
effective. All right, I now go to the I
can't remember this and it's complete my
knowledge kind of query. This is my
favorite version. Uh, I can't remember a
movie I was watching. It had a kid in it
that was like a robot. He flew in like
an SR71 or something across the US and
the kid was part human. It was really
old from the 80s. It's Daryl. D A R Y L.
Um, and yes, I did remember the plane
correctly, which shows you how much of a
nerd I am. Uh, it just retrieves it like
that. This kind of query is kind of
funky for Google to retrieve. Sometimes
you get it, sometimes you don't. I do
get better results reliably with LLMs
because it just pattern matches across
their pre-training data with natural
language in a way that like the standard
Google search just doesn't do as well
typically. Then moving on, last query I
think I threw up here. What's my best
approach to getting into the whole AI
vibe coding space if I am a
beginner? Now, you'll notice as I go
through here that I am using what I
would call Nate prompts. I'm doing a
little bit more language here. I It's
hard for me to stop myself doing that at
this point. Like I know how LLMs work. I
know that works better as a query. The
jury is very much out on whether people
are going to
tolerate longer queries to get results
that are more meaningful to
them. I know that people are using Chad
GPT a lot. My guess is that is going to
bleed into search as openAI is able to
make search more relevant and meaningful
and seamless a part of the LLM
experience. As an example, I get people
who are asking
me why does chat GPT not understand chat
GPT's own plans? Well, the answer is the
pre-training data doesn't include
current models and it would have to
search. But that's a very easy search
thing and it would feel really seamless
and it would be one of those examples of
search being really useful in the model
for people who are typing very quick
queries like what's in chat gpt
free. All right, moving on. Uh it gives
me a practical path to coding. It talks
about the stack. This gets in over my
head. So let's just assume I am much
more of a beginner than all of this.
Um, so I'm going to actually do this
live here. Let's assume I am 9 years
old. Uh, please dial back and give
me a
beginner friendly approach. Now, this
gets at the one weakness that I think
they're going to have to
fix. This is slow. It's good search, but
it tends to be slow. This is the fastest
response I had in this whole query
string. And I think it's because it's
reasoning off of a pre prior response.
So it's
quicker. So it gives me like specific
recommendations.
Um which I think are really great
actually. Like it's super helpful. Might
use this with my kid. Um gives me
support and safety. Encourages me to
have a grown-up, which sounds great. Um
I love this tip. If something feels
confusing, say pause and ask for help.
That's what real engineers do every day.
I think that's just a fantastic little
moment.
And uh it gives me encouragement that
spending just 15 minutes really
helps. So my point here is that search
is at the point in LLMs where it
is potentially disruptible if they can
bring this kind of quality of search
down across the free tier. And I want to
flag it now. I want to get what your
guys' thoughts are. I would be worried
if I were Google. I don't think it's
going to be easy to compete with OpenAI
at this point. They have all the
momentum. They have the ability to
distribute to their whatever 6 7 800
million I lose track. It's going to hit
a billion by the end of the year active
users. And that search experience if it
starts to peel away search volume, it
could look like a sigmoid curve where it
peels away a little bit initially and
then it just falls off a cliff. uh if
they can actually crack the nut on
product adoption for basic everyday
searches which I guarantee you there are
teams at OpenAI working on. What's your
thought? Are we at a point where we are
close to a tipping point on