Hacker News Reader

Comment by Mistletoe | Hacker News Reader

I’m constantly confused why Google wants me to use Gemini so much. Why would they want me to use an AI that doesn’t show ads, isn’t their search, and costs them money to generate the responses? The instant we see ads in LLMs we will stop using them, it will be as abrupt and as annoying as the vuvuzela example because using an LLM is a very intimate interaction. It will feel much more jarring when your intimate chat friend becomes a used car salesman to you. I get that AI=stock go up right now but it seems like they are poisoning their own well deeply down this path.

View on HN

Google thinks it is going to charge a monthly fee for Gemini so they are basically correcting their model to be user paid in the switch from search to LLM.

They are also trying to reduce workers in search, etc, somewhat through the trained LLMs so the other idea is that they have lower costs per user.

I don't think many people will pay monthly fees or will want to pay them for each platform they use which is why they all tend to do so many questionable integration attempts to try to get users to not want to use a separate LLM of their choice in a browser instead.

This.

Gemini Pro requires a monthly subscription.

Seems like a pretty straightforward business model.

How many would really pay though? I certainly wouldn't. Would you pay for Google search?

Their ads are very convincing, I'd pay up to 40¢ a month for Gemini to tell me how to recover with a recipe for tomato sauce I accidentally put sugar in, though only in months where I inexplicably add sugar to things.

> I’m constantly confused why Google wants me to use Gemini so much. Why would they want me to use an AI that doesn’t show ads, isn’t their search, and costs them money to generate the responses? The instant we see ads in LLMs we will stop using them, it will be as abrupt and as annoying as the vuvuzela example because using an LLM is a very intimate interaction.

I think with the last sub-clause (that I emphasized), you answered your question: because the conversation is more intimate, Google learns more about your "true interests", be it to make advertising to you more targeted, or for more sinister purposes.

If you're going to use an LLM instead of their search better to keep you on their estate. Presumably they're using responses to tune a personalisation layer, or in some way, they can use to modify their advertising algos?

> The instant we see ads in LLMs we will stop using them

People didn’t stop using Google Search, Facebook, Instagram, YouTube, TikTok and a myriad other services and products (and things like TV before those) because they got ads.

It's not just ads but plenty of dark patterns. At some point as a user you no longer have much control and no longer get what you were looking for and decide to leave. But the slow boiling leaves a lot of frogs in the pan, it will probably be a slow decay before death. To me browsing the net is unusable without ad blockers but there are many users out there who don't know what ad blockers are.

> The instant we see ads in LLMs we will stop using them

You mean like how the instant we see ads on YouTube we will stop using it?

Or like how the instant Netflix disables password sharing, we will stop subscribing to it.

> The instant we see ads in LLMs we will stop using them

But what if you don't see distinct ads? LLM advertising can just be paid-for bias in the generated content.

Moreover, the bias could be arbitrarily subtle. This is something that really worries me about black box AI systems.

I‘d argue that, at some point, this behavior will surface enough that it gets interesting for government to take a look at it. At least in my country, ads need to be labeled as such - I don’t think this is going to change.

The entire business model of tech is to fly under the radar of regulations. They can't even force the bastards to stop tormenting us with cookie pop-ups, i seriously doubt the government will do much.

What marketing department on earth is going to pay for an ad campaign so subtle that they can't even tell if it's running?

Ad budgets aren't bottomless. People making decisions about them have a lot of options for where to spend. They want provable attribution so they can tell which channel is giving them the most bang for the buck. If that exists, then the ads will be discernible.

Google has this pesky problem of carrying almost all their eggs in one basket: The moment their ad business doesn’t work anymore, the behemoth that Google is will come down to its knees. But due to the way the company is structured, they need search to drive users to the ads, and it’s all one big mess of entangled revenue streams that can’t be touched for fear of breaking something.

Now if AI turns out to be the next big thing, they can steer differently next time, sell subscriptions, and avoid all that entanglement with multi-sided markets and layered revenue strategies. At least that’s my take.

>The moment their ad business doesn’t work anymore, the behemoth that Google is will come down to its knees.

Which makes it all the weirder that they seem to be intentionally sabotaging it by nerfing Search into unusability.

The slow destruction of Google search is and will continue to be revenue positive for Google right up until the point where it isn't.

Eventually they will break the "trust thermocline" in their search results and that will blow it up but on the way they'll keep making more and more money from every damaging change they make.

Crucially, the people driving each incremental change get to celebrate the extra revenue they drove and get promoted for it.

> they need search to drive users to the ads

Why? The websites we visit can still be infested with google's ads, and so can our gmail accounts, and so can the youtube videos we watch, and they can push ads directly onto our cell phones 24/7. Google has plenty of ways to force ads into your life.

Google used to need search in order to build extensive dossiers on everyone. It told them what people were looking for online. What they were interested in. Now Google has their cell phones, their browser, and their DNS servers doing that for them. Most people are handing all of their browsing history to Google. Google doesn't need search, which is why it's been allowed to atrophy into uselessness.

This was true, and is why Google didn't come out with ChatGPT themselves 5 years earlier. But since OpenAI's come to take their lunch, they understand this predicament and are pivoting.

Until that pivot isn't even close to complete, this continues to be true. Take ads away now, and Google ceases to exist tomorrow. It's going to take a long time until that fundamentally changes.

I saw my first AI ad yesterday, in Microsoft Copilot, recommending a Jetbrains product.

Because Gemini has paid tiers. Also to prevent users from going to other providers.

Reply on news.ycombinator.com