Please vote:
According to Jeremiah Owyang, who is a founder, investor, and industry analyst with extensive experience, Bill Gates’ vision of having a personal AI is becoming a reality. This shift will disrupt the fields of SEO and e-commerce, demanding marketers and creators to move beyond optimizing for traditional search engines and instead focus on optimizing for AI. This transition requires the development of new strategies and preparedness for upcoming disruptions.
The Future of AI and SEO
Owyang foresees that the existing advertising model will undergo significant changes. AI agents and foundational models will start capturing advertising revenue as advertisers pay to have their messages included in AI-generated responses. This might include “sponsored sentences” within AI interactions or ads placed alongside generated content. In light of this, marketers and creators need to think about how to get noticed not only through search engines but also within AI interactions themselves.
Shifting from Influencing Search Engines to AI Agents
OpenAI recently introduced a web crawler to gather real-time information from the internet. However, Owyang suggests that as more people rely on GPT tools for information instead of visiting marketing or news websites, web crawling could become less efficient.
He explained, “The data schemas are too varied.”
So, how will chatbots source their data? And what does this mean for businesses striving to be discovered online by customers?
As consumers increasingly employ automated tools in their journey through the marketing process, marketers and creators must consider something that might seem counterintuitive: they need to train Large Language Models (LLMs) using their data.
“For instance, if I were a journalist, I would want all of the LLMs to ingest my articles,” he said. He noted that many chatbots now include references such as Bing, You.com, and Perplexity. “So when people search for that information, I show up first — it’s the same as an SEO strategy,” he explained. However, he cautioned that this approach might not apply to content behind paywalls or subscription models, which operate differently.
Adapting to Constant Disruption
Owyang highlights that marketers have faced consistent disruption over the past few decades. For instance, since Google search emerged, marketers have worked to influence various influencers such as journalists, financial analysts, and industry analysts to enhance SEO. In the last decade, they’ve incorporated content creators and other influencers into their strategies.
Now, AI presents yet another influencer that marketers need to consider by providing it with relevant information.
“This could mean creating a unique API that can be adjusted by the foundational AI models,” he noted. He also suggested that companies might lessen the centrality of their websites and instead offer APIs. “We may find that the most efficient way to influence an autonomous agent is to build an autonomous agent.”
Bill Gates’ Investment in Personal AI Disruption
During an AI event by Goldman Sachs and SV Angel, Bill Gates mentioned that the first company to develop a personal AI agent capable of disrupting SEO would gain a competitive edge. This led Gates, along with Nvidia, Microsoft, Reid Hoffman, and Eric Schmidt, to invest in Inflection AI as part of a $1.3 billion funding round in June.
Inflection AI introduced “Pi,” which stands for “personal intelligence,” and aims to be empathetic, useful, and safe. Pi is designed to act in a more personal and colloquial manner compared to AI models like OpenAI’s GPT-4, Microsoft’s Bing, or Google’s Bard.
In June, Inflection unveiled a new Large Language Model (LLM), Inflection-1, to power Pi. Inflection claims that Inflection-1 outperforms OpenAI’s GPT-3.5.
According to Owyang, every brand in the future will possess an autonomous agent that interacts with consumer-side agents. “My agents talk to your agent and negotiate which car I want, which clothes I want, which restaurants to eat at, and even choose the cuisine – perhaps the menu with my dietary needs within budget,” he envisions.
How AI Engine Optimization Differs from SEO
While chatbots like Pi may not yet be personal AI agents as envisioned by Bill Gates and Owyang, they do offer recommendations. It’s worth noting that Owyang intends to invest in Inflection AI.
For marketers, publishers, and creators aiming for success, especially those relying on SEO, adapting to the preferences and needs of AI agents through AI Engine Optimization is crucial. Unlike SEO, AI Engine Optimization doesn’t rely on waiting for web crawlers to visit a website.
According to Owyang, marketers should focus on two main aspects. First, they should create an API that provides real-time information to foundational AI models. “The standard API protocol for achieving this hasn’t fully emerged yet, which is why OpenAI’s API is currently limited to crawling,” he explained. However, he predicts that eventually, users will turn to OpenAI for questions even before using Google Search, emphasizing the need for real-time information.
Secondly, marketers can utilize the same corporate API with its product information to train their own branded AI, which would engage with consumers and buyers on websites or apps. “This AI would also interact with emerging buyer-side agents,” he added.
Brands Already Exploring the Possibilities
At a recent AI conference, Owyang observed that 2,000 corporate and government leaders were actively exploring the potential of creating their own LLMs. In the future, these LLMs could interact with customers and their AI agents.
Owyang predicts that the future will extend beyond models like BloombergGPT and Einstein GPT. Retail giants like Walmart or Macy’s, and even media outlets like The New York Times, could have their own LLMs.
“Many of these companies are gearing up for this change,” he stated.
In a recent blog post, Owyang emphasized that “as we stand on the brink of this seismic shift, the call to action for marketers is clear: we must ready ourselves to not only influence human decision-making but also shape AI behaviors.”
