Zinger Key Points
- True utility, not speculation, is the biggest hurdle for AI agent adoption in mainstream applications.
- Open-source LLMs and decentralized inference can coexist with centralized AI players like OpenAI.
- Get the Strategy to Trade Pre-Fed Setups and Post-Fed Swings—Live With Chris Capre on Wednesday, June 11.
Humayun Sheikh, CEO of Fetch.ai FET/USD, dismissed the notion that AI agents forming collectives or decentralized autonomous organizations (DAOs) is imminent, labeling it an overhyped crypto narrative.
In an interview with Benzinga, Sheikh argued that the concept of AI agents autonomously creating DAOs lacks a clear purpose and is driven by speculative fundraising rather than real-world utility.
"I don't agree with that. I don't think that will happen. I don't think it will catch up," Sheikh said in response to whether AI agent collectives or DAOs are near or still science fiction. "It's one of the crypto narratives which people want to go use and raise money around. No, I'm not sold on it."
Sheikh emphasized that the idea of AI agents forming DAOs is an example of over-engineering, a common issue in the crypto industry where complex solutions are built for non-existent problems.
"You're over-engineering, which is a typical problem in this industry. We kind of create some very complex solution. But the problem doesn't exist," he stated.
Instead, Sheikh envisions a future where AI agents simplify interactions across industries like e-commerce, finance, and logistics by acting as universal interfaces.
He predicted that within the next 12 months, every website, business, and service will be represented by an AI agent, reducing the need for traditional apps.
"Everything will be represented by agents. Everything. Other machine learning models, humans, businesses, websites, everything will be done via agents," he said.
On real-world adoption, Sheikh highlighted user interface complexity as a surmountable challenge, with language models enabling seamless interactions, such as transferring tokens between wallets.
However, he stressed that the bigger obstacle is creating true utility beyond speculation.
"Most people don't speculate, only crypto bros do. So speculation is a crypto thing. We need to go beyond this," he noted, urging the industry to focus on practical applications like restaurant bookings or grocery ordering via agents.
Sheikh also addressed the role of decentralization in AI development, arguing that open-source large language models (LLMs) and decentralized inference, like Fetch.ai's ASI1 platform, don't need to compete with centralized players like OpenAI.
"I don't think we need to compete. We just need to be available as well," he said, pointing to open-source models' accessibility as a key advantage.
Regarding regulation, Sheikh believes existing e-commerce regulations are sufficient for AI agents executing tasks like financial transactions or bookings, cautioning against connecting agents to critical systems like power or medical infrastructure.
"I don't think we need to worry too much about over-regulating it because there is no need," he said, advocating for small, incremental steps in agent autonomy.
On interoperability, Sheikh noted that agent-to-agent protocols already exist, citing Fetch.ai's infrastructure as proof.
"Agents can speak to each other right now. So you don't need to be overly complicated. It exists," he said, emphasizing that adoption will grow as awareness of AI agents increases.
Read Next:
Image: Shutterstock
Edge Rankings
Price Trend
© 2025 Benzinga.com. Benzinga does not provide investment advice. All rights reserved.
Trade confidently with insights and alerts from analyst ratings, free reports and breaking news that affects the stocks you care about.