Everyone is being naive. There is no such thing as a profitable LLM without ads.
Betting on ad-free LLMs is like betting against math.
OpenAI burned $1.6 billion last year alone. Other companies are doing their best to match it.
Subscription revenue is a drop in that bucket.
Compute costs are exponential and realistically, the subscriptions are linear.
Only one business model scales infinitely.
Ads. One query, one ad. A billion queries, a billion ads.
The economics of it dictate the outcome. Ads are inevitable.
What’s worth talking about now, is the winner or winners.
The first ones to monetise your attention, not just your answer.
But what is also worth talking about is the problem. Whoever embeds ads in the model, outputs controls what billions of people see as objective truth.
And that’s a scary reality.
I’m Blake. I’ve spent the last 10 years in tech advising companies on data and AI strategy.
I’m here to break down for you the most interesting moves the players are making in AI.
Products. Moats. Distribution tactics. Everything is curated so you can stay informed.
Today, we’re diving into the upcoming land grab for ad revenue.
Let’s break it down.
I’ll start with something Sam Altman said.
“AI will be superhuman at persuasion before it is superhuman at intelligence”.
The key is in the word. Persuasion.
When an LLM answers your question, that answer FEELS like truth.
But if an advertiser pays to change that answer, that truth becomes inventory.
The company that controls this doesn’t just sell ads.
They sell the perception of truth. And once they lock that in, nobody can compete.
They get to rewrite reality for profit. Now, I’m going to spoil the video for you here.
I think Google will win. That’s my ultimate take.
But to make it clear as to why they will win, we need to explore what the major players are and what they are doing.
It’s only natural to start with Google.
I think Google isn’t just building an AI business. They are upgrading the world’s largest ad monopoly.
If you’ve noticed over the last few days, Google made its first foray into native chat ads.
I think how the ads look now, is similar to Bard compared to Gemini. There’s a lot of room for improvement. But they will get there.
Now of course, having to become the best AI company isn’t by choice. The upstarts are eating their biggest revenue model.
But. As they’re forced to compete, they’re also using their moats.
Every entry point to the internet already funnels through Google infrastructure.
They don’t need to find users. They have 4 billion of them on Android, Chrome, Maps, Gmail and YouTube.
They also don’t need to find advertisers. They have millions already paying them.
They have an ad salesforce.
Embedding Gemini into Search turns their existing monopoly into an AI monopoly overnight.
They technically win by default because they own the rails.
And I haven’t even touched on Waymo. Or deepmind.
Or whatever they cook up next. On the other hand though, there’s OpenAI.
OpenAI has the arguably smartest model, and the consumer mindshare.
But, they are homeless.
They live in Microsoft’s house. Azure, Bing, Office… it’s all rented land.
OpenAI has zero control over the operating system or the browser.
To sell ads, they will need data.
OpenAI has chat logs.
But Google has your search history, location, emails, and purchases. It’s not a fair fight.
It’s why they’ve created Atas.
It’s a data play.
But, alas, Atlas isn’t chrome and Unfortunately for OpenAI, brand recognition doesn’t pay the rent.
In platform wars, the landlord always wins.
I do see OpenAI releasing a suite of work related tools.
Their version of word, excel and powerpoint in the future.
Maybe an Atlassian takeover? Stranger things have happened.
So who else is in the race. Another big contender is Meta.
Meta knows who you are.
Meta’s moat is the social graph. They can embed AI into the conversations you care about most.
Meta’s advantage is context, they know your social graph, your interests, your behavior.
That data is gold for targeted ads.
They’re strong in engagement and social commerce. Weak in neutral information provision.
But social isn’t search.
And you don’t go to Instagram for facts; you go for vibes. Meta wins engagement, but they lose the “truth” monopoly.
Meta lacks the search-grade AI dominance and enterprise distribution that Google has.
Google only knows what you want. That’s a dangerous difference.
They do have a large chance of owning a niche, entertainment AI.
But, they won’t own the information utility.
Which brings us to X.
X is betting that “truth” is overrated and “now” is what matters. They own the real-time pulse.
News breaks on X. X plays the disruptor. Fast-moving, cultural, willing to take risks.
But their ad infrastructure is broken, and brands are scared of the chaos.
They can disrupt the narrative, but they can’t monetise it at the scale of Google.
How can I put this simply? Google owns the pipes. Microsoft owns the office. Meta owns the party. X owns the riot.
In a war for information dominance, owning the pipes (Google) beats everything else.
So if you’re betting on a winner, bet on the plumbing.
Google is the plumbing.
More users = more data = better ad targeting = more revenue = better models = more users.
Once the ad engine starts, the door closes. Forever.
I’m Blake.
Hopefully this was interesting
See you in the next one.
