Google is rethinking its advertising model to adapt to generative AI chats. For months, the company has been testing injecting ads directly into conversations with Gemini AI. Google is expected to roll out these ads more widely in the coming months.
Did you also find it strange that Google hadn't already included ads in your conversations with Gemini? Don't worry! The company has already been on the case for months. And according to a report from our colleagues at Bloomberg, ads will indeed be introduced into conversations with its AI – at least for users who don't pay a subscription… but also, perhaps, other groups of paying users.
According to our colleagues, the injection of ads into conversations has indeed begun on the sites iAsk, Liner and other platforms whose service relies on Gemini's API (application programming interface). And it seems that the firm is at this stage quite satisfied with these ad impressions, which therefore have every chance of being integrated into a more global monetization strategy.
Google should soon insert ads into your conversations with Gemini
A Google representative confirms that “AdSense for Search is available for websites that want to show relevant ads in their conversational AI experiences”. The arrival of advertising in AI-powered chatbots is obvious in itself. Generative AI services are extremely difficult to monetize.
They currently make it easier for the large firms that publish them to lose money than the other way around. OpenAI, a pure player in the sector, for example, does not expect to reach financial equilibrium before 2029. With services like Deep Search causing exorbitant costs in terms of computing resources.
This is one of the reasons why these services are also seeking to make the execution of their models less burdensome on the infrastructures they rent from players like Microsoft Azure. At the same time, the entire sector is in a race for market share. A race that prevents setting prices that are too high – even if reaching such price points would make the operation of the models more economically sustainable.
0 Comments