OpenAI Closed, Goertzel Open: Why the AGI Pioneers Are Betting the Future on Blockchain and Decentralized Networks?

Written by: Boaz Sobrado

Compiled by: AididiaoJP, Foresight News

"I don't want to hand over control of my work to venture capitalists," Ben Goertzel said on the (On The Margin) podcast, "because I believe AGI is too important to do that."

Goertzel popularized the term 'Artificial General Intelligence' (AGI) through an influential book, led the team developing the Sophia robot, and has spent decades trying to build a thinking machine. Now, his bet runs counter to almost everyone racing toward the same goal: the most critical software in human history should not be owned by a single company. "I firmly believe that the core AGI code responsible for thinking should be free and open-source," he stated.

Note: Ben Goertzel is the founder of SingularityNET, one of the leading researchers in AGI (Artificial General Intelligence), often referred to as the 'Godfather of AGI.' He strongly advocates that AGI must be open-source and decentralized, opposing the monopolization by companies like OpenAI and Anthropic. He believes AGI is too important to be controlled by venture capital firms or a single company.

This belief drives his blockchain project SingularityNET, and the broader Artificial Superintelligence Alliance. The alliance, formed in 2024 by Fetch.ai, SingularityNET, and Ocean Protocol, combines their tokens into FET (Ocean Protocol exited by the end of 2025). While OpenAI and Anthropic raised hundreds of billions and locked their strongest models behind closed doors, Goertzel has connected his AGI work to a crypto network owned by users.

Open-source code isn't enough; decentralized computing is also needed.

His argument starts from a problem that open-source itself cannot solve. He says just releasing the code doesn't help much if no one can afford to run it.

"If the code is open-source but the data needs a server to store it, and you need a massive server to use it, then open-source code is not very useful," he said. "What you really want is to deploy the first AGI on a decentralized network controlled by ten thousand different people across fifty different countries."

In his view, this is the whole point of putting AI on the blockchain. He acknowledges it might make some people nervous, as bad actors could fork the system and create harmful things, but he's willing to accept that trade-off.

"I believe there are more good people than bad in the world, and they would be willing to host AI systems," Goertzel said. "We will be better off with open and decentralized AI, rather than getting caught in an AGI arms race where only a few major countries possess AGI and use it to destroy each other."

Those laboratories that have shifted from open-source to closed-source

Goertzel is outspoken about companies that once held open ideals like him but later diverged. He points to the litigation records between Elon Musk and Sam Altman, clearly showing how quickly OpenAI's founding mission has shifted.

"If you look at all the court hearing records between Elon Musk and Sam Altman, it seems Sam Altman didn’t hold out for long," Goertzel said. "He quickly turned to try to make it proprietary. The records show he shifted quickly to acquiring hardware, which required them to be closed and proprietary."

He is harsher on the origins of Anthropic and unrelenting about Musk's transition from an AI doomsayer to an AI builder.

"Dario Amodei's stuff was closed and proprietary from the start," Goertzel said. "Elon Musk's relationship with openness has always been complicated. In 2015, he first said AI was summoning demons, and no one should build it. Then he probably started trying to build it himself. But now what he's doing at xAI is also closed AI development."

Goertzel acknowledges that the closed path is indeed simpler: raising venture capital, locking the model, and moving towards acquisition. He insists that the open path is not impossible, just harder. "Linux and the Internet itself are proof: it can be open-source, globally decentralized, and still serve as a foundation for many people to make money," he said.

How token founders make money

Currently, his business runs on cryptocurrency. "I have SingularityNET, which is a utility token, so it's a blockchain project," Goertzel said. "We have node operators running SingularityNET, hosting AI processes. We essentially make money through the tokenomics on this network."

He expects this situation to change. The plan is to keep AGI code open-source while selling polished products on top of it, hiding the blockchain in the backend. "Our project might slightly shift from Web3 towards Web2, starting to offer products purchasable with regular currency," he said. "The backend is still our crypto network, but for the end user, it's just an AI service."

Goertzel says that SingularityNET will roll out a paid tier for enterprises and heavy users sometime next year. "We're going to launch something like Claude Pro or ChatGPT Pro but smarter," he said. "It runs on a decentralized blockchain backend, but the reasoning capabilities and creativity are much stronger than today's chatbots." He won't chase the retail chatbot market. "I’m not going to try to launch a retail product like ChatGPT because those guys are losing money," he said.

The Agent economy he's building

Goertzel's pitch to regular users is that the next wave of advantage will belong to those who can command fleets of AI Agents, not to labs.

"The next wave of victories will likely belong to small groups that can effectively organize larger teams of AI Agents to get things done," Goertzel said. "The key lies in how you educate and command your beneficial Agent army."

Other builders in the crypto space are heading towards the same vision: Agents not only answer questions but also spend money. "The next step - which has actually already begun - is AI Agents starting to trade on your behalf," said Varun Kabra from the identity chain Concordium in the same podcast. "They will make payments, register services, and possibly handle your financial transactions."

This assigns a risky task to software. "The essence of an Agent is outsourced purchasing, and anyone who has experience with outsourced purchasing knows it comes with trade-offs," said Nitya Subramanian from the crypto wallet company Para in the same podcast. Goertzel's answer is: running this economy on an open network rather than a corporate cloud.

Why this matters

Goertzel still believes human-level AGI is coming soon. "I think we can achieve it by 2029," he said. "If it happens in 2027, I won’t be shocked. If it’s 2030, I wouldn’t be surprised either." He is more concerned about who will be left behind by the machines rather than what the machines will do.

"If there is a gap in understanding between the top-tier folks and the bottom-tier folks about what’s happening with AGI, that gap will exacerbate the unequal speed of growth," he said.

It's still unproven whether a crypto network can outpace companies with hundreds of billions. Goertzel's first test is set to launch in weeks. "We will be releasing the first downloadable version of our new Agent Omega Claw in a few weeks," he said. "We'll have the opportunity to teach our own personal Agents to help us manage our lives and assist us in making money."