Given how much AI costs to train and run as well as the sort of company Google is, I can't really say I am surprised. This also slightly feels like shoehorning AI into it for clicks. Google will...
Given how much AI costs to train and run as well as the sort of company Google is, I can't really say I am surprised.
This also slightly feels like shoehorning AI into it for clicks. Google will force sponsored results in any overview they can get away with.
I was expecting there to be ads inside the Gemini app or within the conversation. Like, it would say, “while I generate that image, enjoy this ad.” That would have been way worse. This is just...
I was expecting there to be ads inside the Gemini app or within the conversation. Like, it would say, “while I generate that image, enjoy this ad.” That would have been way worse. This is just more ads on the search page.
This article is extremely light on detail. All I can say for certain is that they are showing ads in the AI Overview section of the search results now. How are those ads generated? Via AI? I can’t imagine a brand worth anything just saying, “Sure, lets give the AI a shot at writing our search ads!” And if the ad gives false info about a sale or something, who bears responsibility for that? The advertiser or Google?
If they aren’t AI generated, then how are they any different from normal ads, other than appearing under a new heading?
This concept has raised an interesting question though: How will ads impact conversational AI? For example, if I ask for an itinerary for a three day trip to Paris, will I get the “best” information from Gemini or will I get information about a shitty hotel that paid for an ad. There’s all sorts of weird trust issues to work out with ads and these new “AI” features.
This is the same Google that gave us life stories in recipe blogs and the stereotypical Youtuber video format. I'd hate to see what their version of this for AI will be. Based on what their SEO...
This is the same Google that gave us life stories in recipe blogs and the stereotypical Youtuber video format. I'd hate to see what their version of this for AI will be.
Based on what their SEO favors now I can only a guess a goal where they want to maximize your engagement to Gemini as long as possible so you don't bounce. Dole out the actual results so you're forced to enter more prompts to find the specific information you're looking for. Then sprinkle links to sponsored prompts that return a fast canned answer crafted and vetted by the sponsor.
Tacking on to this, even if there were some legal obligation for transparency (there's probably not in the US), it would be super hard to figure out. For example, instead of just dropping in the...
Tacking on to this, even if there were some legal obligation for transparency (there's probably not in the US), it would be super hard to figure out. For example, instead of just dropping in the text, the AI could be shaped by coding and training to favor certain commercial interests in its answers. That's kind of dystopian to me, and really difficult to detect. Consumer interests are super hard in a capitalist society, and it seems now more than ever.
SEO already ruined good search results and the same is going to happen with AI models. If there is a financial incentive to be part of the data training an AI model, someone will try to game it.
SEO already ruined good search results and the same is going to happen with AI models. If there is a financial incentive to be part of the data training an AI model, someone will try to game it.
They're already planning on it by putting up content where the AI searches go for data. SEO experts will be going to go beyond posting blog posts cluttering up searches to cluttering places like...
They're already planning on it by putting up content where the AI searches go for data. SEO experts will be going to go beyond posting blog posts cluttering up searches to cluttering places like Reddit even more.
This is what Bing does right now with its LLM responses. Sponsored links are included if you ask about stuff like plane tickets or clothing.
I was expecting there to be ads inside the Gemini app or within the conversation. Like, it would say, “while I generate that image, enjoy this ad.” That would have been way worse. This is just more ads on the search page.
This is what Bing does right now with its LLM responses. Sponsored links are included if you ask about stuff like plane tickets or clothing.
Given how much AI costs to train and run as well as the sort of company Google is, I can't really say I am surprised.
This also slightly feels like shoehorning AI into it for clicks. Google will force sponsored results in any overview they can get away with.
I was expecting there to be ads inside the Gemini app or within the conversation. Like, it would say, “while I generate that image, enjoy this ad.” That would have been way worse. This is just more ads on the search page.
This article is extremely light on detail. All I can say for certain is that they are showing ads in the AI Overview section of the search results now. How are those ads generated? Via AI? I can’t imagine a brand worth anything just saying, “Sure, lets give the AI a shot at writing our search ads!” And if the ad gives false info about a sale or something, who bears responsibility for that? The advertiser or Google?
If they aren’t AI generated, then how are they any different from normal ads, other than appearing under a new heading?
This concept has raised an interesting question though: How will ads impact conversational AI? For example, if I ask for an itinerary for a three day trip to Paris, will I get the “best” information from Gemini or will I get information about a shitty hotel that paid for an ad. There’s all sorts of weird trust issues to work out with ads and these new “AI” features.
This is the same Google that gave us life stories in recipe blogs and the stereotypical Youtuber video format. I'd hate to see what their version of this for AI will be.
Based on what their SEO favors now I can only a guess a goal where they want to maximize your engagement to Gemini as long as possible so you don't bounce. Dole out the actual results so you're forced to enter more prompts to find the specific information you're looking for. Then sprinkle links to sponsored prompts that return a fast canned answer crafted and vetted by the sponsor.
Tacking on to this, even if there were some legal obligation for transparency (there's probably not in the US), it would be super hard to figure out. For example, instead of just dropping in the text, the AI could be shaped by coding and training to favor certain commercial interests in its answers. That's kind of dystopian to me, and really difficult to detect. Consumer interests are super hard in a capitalist society, and it seems now more than ever.
SEO already ruined good search results and the same is going to happen with AI models. If there is a financial incentive to be part of the data training an AI model, someone will try to game it.
They're already planning on it by putting up content where the AI searches go for data. SEO experts will be going to go beyond posting blog posts cluttering up searches to cluttering places like Reddit even more.
This is what Bing does right now with its LLM responses. Sponsored links are included if you ask about stuff like plane tickets or clothing.