AI Agents Will Change the Internet

It's just a matter of time—a few years, in fact—before humans will no longer need to visit websites in the same way they do today. Why? Because AI agents will be doing it for us.

That's according to Jeremiah Owyang, a general partner at Blitzscaling Ventures, who made the prediction during his presentation at the GenAI Summit 2024 in San Francisco this week.

"During the Industrial Revolution we traded muscle power for mechanical power," Owyang said. "Now we're replacing cognitive power with compute power."

Owyang's talk focused on the rapid growth and changes in the AI industry and the importance of business models over technical prowess for AI startups.

"We've never seen growth this fast," he said. "We've never seen this many startups so fast. We've never seen so many enterprise companies become AI companies. We've never heard AI, or any trend mentioned in the boardroom. And you can look at the all the analyzes on what boardrooms are talking about in their annual reports. And certainly, every single big tech company wants to be an AGI leader."

Owyang backed up his predictions with some persuasive statistics.

"There are over 12,000 AI projects in existence," he said. "And again, this market is so young, not even two years old. A thousand new AI projects launch per month. That means two projects will launch during the course of my presentation. The cost is very low. People were doing it all around the world. You just need SaaS access. You just need a few dollars and a great idea. It's very easy to get an AI startup to go." He added that San Francisco, which he called "the cradle" of generative AI, saw 145 AI-related events last month alone. "There's usually two events per day and three-to-five events per night on Saturdays and Sundays. And there's usually several hackathons per month, with 100 to 300 developers per hackathon working overnight."

The business opportunity Owyang sees in this rapidly shifting landscape is the estimated that 80% of the world's data that's secured behind firewalls and stored in private clouds, corporate clouds, government clouds, and military clouds.

"Most of the world's data has not been shoved into an AI and indexed and trained," he said. "That's where the business opportunities are. In fact, in our investment thesis, we're looking for the companies that have access to exclusive data."

He also pointed to the money-making potential of synthetic data—the data being generated by AI—and he cited Adobe Firefly and Behance, Adobe's social media platform for showcasing creative work, as an example of an organization that's monetizing that data.  

Owyang drew parallels between foundational models in AI and operating systems, and he categorized these models into two types: proprietary and open source. His list of notable proprietary models included OpenAI's GPT and Google's Gemini, while popular open-source models such as Meta's Llama 3 and UAE's Falcon lead the pack among approximately 40 open-source versions.

He emphasized the strategic advantage for sophisticated AI startups and enterprises in adopting a model-agnostic approach, allowing them to swap out the best and most cost-effective solutions. He noted that several vendors at the summit were showcasing products that enable this flexibility.

But it's the applications layer is the focus of the majority of AI-driven companies today, Owyang said. Approximately 12,500 of the 12,800 extant AI companies focus on apps, he said, ranging across geographies and use cases, including customer care, HR, sales, and even niche markets such as religious texts. Owyang identified this layer as having significant growth potential.

Looking ahead, Owyang discussed the emerging category of AI agents, which are poised to transform internet interactions. Unlike traditional AI applications with real-time copilots, AI agents operate autonomously with minimal or no prompts, conducting tasks and navigating the internet independently. This shift is expected to reduce the need for users to visit websites directly, fundamentally changing how the internet functions.

Owyang shared an example of a consumer-level AI agent from a company called MultiOn, which autonomously books flights and compares prices without user intervention. This development underscores the potential for AI agents to revolutionize everyday tasks and interactions.
The business opportunity Owyang sees in this rapidly shifting landscape is the estimated that 80% of the world's data that's behind firewalls, private clouds, corporate clouds, government clouds, and military clouds.

Highlighting the importance of business models for AI startups, Owyang pointed to key success factors such as access to exclusive data, network effects, and product-led growth. He referenced insights from the book, Blitzscaling: The Lightning-Fast Path to Building Massively Valuable Companies, written by Chris Yeh and Reid Hoffman. Yeh is a founding general partner of the company, and Hoffman, who co-founded LinkedIn, is an advisor. The two taught the first "Blitzscaling" class at Stanford University (along with John Lilly and Allen Blue), on which the book is based.

Owyang advised startups to focus on niche markets to avoid being overwhelmed by the rapidly changing landscape. He also mentioned the potential for AI to reduce market competition by providing the same deals to all consumers. Owyang also pointed to the potential of AI to level the playing field in market competition by offering the same deals to all consumers. However, he suggested that innovative marketing strategies could help companies maintain a competitive edge.

Concluding his talk, Owyang cited Bill Gates's bold prediction that AI could disrupt dominant platforms like Google Search and Amazon by winning the personal AI agent market. With Gartner forecasting a 25% decrease in search engine traffic to Google, the race for AI supremacy is set to redefine the digital landscape.

This year's GenAI Summit, underway this week in San Francisco's Palace of Fine Arts, is the second annual event organized by GPT Dao, a global generative AI community. According to event organizers, this year's summit drew an estimated 10,000 attendees and 300 exhibitors. The list of exhibitors at this year's conference includes Microsoft, IBM, and Amazon. (A complete list is available on the conference website.)

("Dao" stands for Decentralized Autonomous Organization. Dao's operate based on smart contracts executed on a decentralized network, typically a blockchain, which allows for decentralized decision making and control over the organization’s assets and operations.)

GPT Dao provides a range of services to its community, including Web3 and AI project incubation, GPT investment research education, and AI infrastructure services. It also provides a platform for community governance designed to allow members to propose, discuss, and vote on changes, new features, and/or initiatives. It also provides a decentralized crowdfunding platform that allows anyone to invest in or contribute to a project, regardless of their location or financial status.

About the Author

John K. Waters is the editor in chief of a number of sites, with a focus on high-end development, AI and future tech. He's been writing about cutting-edge technologies and culture of Silicon Valley for more than two decades, and he's written more than a dozen books. He also co-scripted the documentary film Silicon Valley: A 100 Year Renaissance, which aired on PBS.  He can be reached at