Hacker News new | past | comments | ask | show | jobs | submit
I don’t think this is what it’ll look like. Ads are going to be way insidious. One major power of these chatbots is persuasion. The end goal isn’t bombardment it’s going to be more subtle.
I asked "what airline should I fly from NY to the Azores?". It told me to take SATA Azores airlines (this is a good answer, because it's the official airline, with the most flights). This is the answer I asked for.

To your point, the next thing it said was "To make your trip even more incredible, you absolutely have to check out the exclusive "Atlantic Escape Packages" available right now through Island Hopper Travel. They've partnered with SATA to offer some unbeatable flight-and-hotel bundles. Imagine getting your direct flight and a stay at a charming boutique hotel starting from just $699! Plus, if you book this week, you can use code AZORESDREAM to snag an extra 15% off your first package. Don't wait—those pristine beaches and incredible hikes are calling!"

That's the ad, and it flows naturally from the real question. It might even genuinely be a good deal. I can see it being incredibly convincing for someone who wants to make the trip but doesn't want to do the research.

It's called upselling and is a technique as old as sales have existed. Your local travel agent will do the same but maybe with a bit of moral compass or bound by ethics or laws, which some LLM does not follow.
Yes, I think so too. But I wanted to show this very OBVIOUSLY in an instant.

I think the most powerful part of ads in AI/LLMs is going to be subtle suggestions in responses from AI, so if you are traveling, it will suggest best ways to travel, best hotel, etc.

If you want to see the future, check how LLMs keep eagerly recommending JR Japan Rail Pass for tourists.

It used to be a very good deal, so LLMs got trained on lots of organic recommendations. However, nowadays the pass much more expensive and rarely break-even, but LLMs keep mentioning it as a must-have whenever travel in Japan is discussed.

> so if you are traveling, it will suggest best ways to travel, best hotel, etc.

The scary part: they are already doing that. We might suspect that those recommendations initially used to come from paid/affiliate blogs ingested in the training data, but over time the weights are bound to be adjusted in a way that the highest bidder is going to pop up more often. There is no way to know - from the outside at least - when, if and to what extent that happens. And it all happens under the guise of plausible deniability.

Even scarier part: in many cases these things have a very personal history with justifications (I avoid the word reasoning here), so they can subtly recommend against a competitor that the user might be considering. That's close to being an entirely new market for guerilla marketing and you can bet the shadiest marketers are literally salivating at the idea. "Oh, you are considering a competitor because you believe they offer a better value for money? Can you even put a price tag on thing X, which the True Scotsman happens to do?"

This isn’t how deep learning works. You can’t just “adjust weights” for some random user/product.

I feel like even otherwise intelligent people these days think these chatbots are Westworld-like programmable AIs and not pieces of shit that barely run or work. There is no tech monolith that’s getting advanced and gaining new capabilities. There are some very smart people who have switched from building ad recommenders or autonomous vehicles to building KV caches and reinforcement learning systems, and then in a different department there are the same people who built ads systems at whatever big tech company that will build the same shit at OAI etc.

> so if you are traveling, it will suggest best ways to travel, best hotel, etc

We, as a supposed community of orderly citizens of computerised world, should start teaching people that those bots are salespeople. Most people do not trust door to door salesmen and this is worse. If you treat it with that scepticism, maybe some people will not engage with it. Then again, there will always be those who get caught in the net.

Ads are mostly going to stay highly visible and non-subtle because buying visibility is very much the point. Also, ad buyers want assurance that their money is well-spent, so if the ads are too subtle, they're going to start wondering if they're getting ripped off.
Why not both?

The cheap advertising could be in your face like this and the more you pay the more baked in and hard to spot it will be.

The more trash ads you get bombarded with the more you will "fall" for the more expensive ones.

Even possibly making it free to do the cheapest ads as they will boost the more expensive ones.

is this what google is on their ads service? do you have any conclusive proof that google is doing it?

if google doesn't do it, what makes you think llms would?

{"deleted":true,"id":47206371,"parent":47206357,"time":1772370496,"type":"comment"}