Much of course has been made of the possible business models around LLM AI technologies, be they the premium subscription revenues of OpenAI’s consumer facing ChatGPT Plus at $20/subscriber (reportedly over nine figures for OpenAI), or the Amazon AWS style API usage fees paid to OpenAI by businesses.
WHAT’S OF NOTE: But the other Real Money with LLM AI services, is of course using the technologies to serve online Ads. These are the real potential Juggernaut businesses.
It’s hard to believe, but Google has so changed the world over the last quarter of a century. As Nilay Patel at the Verge reminded us in a must read piece a few days ago (link below),
“Google’s grand promise was to organize the world’s information, but over the past quarter century, an enormous amount of the world’s information has been organized for Google — to rank in Google results. Almost everything you encounter on the web — every website, every article, every infobox — has been designed in ways that makes them easy for Google to understand. In many cases, the internet has become more parseable by search engines than it is by humans.”
It’s happening again. This time with AdTech for AI.
And as one might expect, both Google and Facebook/Meta had back to back announcements in this context this week.
Google plans here were highlighted by CNBC
“The company has given the green light to plans for using generative AI, fueled by large language models (LLMs), to automate advertising and ad-supported consumer services, according to internal documents.”
And plans by Facebook/Meta on the same front.
“Social media giant Meta Platforms Inc (META.O) joined the generative AI product race on Thursday, saying it would begin testing artificial intelligence-powered ad tools that can create content like image backgrounds and variations of written text.
A select group of advertisers will be invited to experiment with the tools in a "testing playground" that the company is calling the AI Sandbox, Meta executives said at a press event in New York.”
Notable also this week was Google/YouTube head Neal Mohan confirming some revenue metrics for his business for the first time, of course powered by earlier versions of AI/ML.
“Google’s YouTube has generated nearly $40 billion in revenue over the past 12 months, according to YouTube Chief Executive Officer Neal Mohan.
“Monetization for us, first and foremost, is driven by advertising,” said Mohan, who gave an overview of the video platform’s business at the MoffettNathanson technology conference on Thursday. He added that YouTube would remain focused on the advertising opportunities across its fastest growing products, including Shorts — the company’s answer to the massive popularity of ByteDance Ltd.’s TikTok — and connected TVs.”
Remember TikTok owned by Bytedance, is one of the foremost examples of AI Algorithms both for consumers, and in serving ads to consumers.
As these businesses evolve, the best engineering talent tends to migrate to the advertising side from the consumer side. We are at the beginning stages of this evolution. Stay tuned.
To understand better how Google Search upended the world the last time around, here is the link to the story referred to above. Worth a read.