Consensus 2024 just wrapped up in Austin, Texas, and over three days in late May, the decentralized AI community and Morpheus ecosystem participants discussed their vision on what’s being developed and how it will look in the decades to come.
Here's the breakdown.
Watch all the videos from DeAi Day Austin here
A Call for Decentralization in AI
David Johnston is one of Morpheus’s 225 core contributors.
Morpheus is neither a company nor a foundation, but instead an efficient way to build a network of decentralised AI. It brings together a permissionless community whose focus is rewarding individuals for adding value, while bypassing the usual centralized bureaucratic processes.
David’s journey with Morpheus is all about the world we are creating for generations to come, and he made that clear at DeAI Day.
“The story hasn’t been written yet. Are there going to be free markets? Freedom of speech? Freedom of thought? It has to be,” he claimed. He’s been working tirelessly to foster the networks that will bring those values to center stage as the 21st century progresses.
Ryan Condron is also one of Morpheus' core contributors. A software engineer with more than 20 years of experience, Ryan began learning about Bitcoin–especially through mining–in 2012, and hasn’t looked back since. His Bitcoin trail has taking him to found Lumerin.io, a Web3 routing firm.
According to Ryan, losing control over their population is one of the main reasons that governments are interested in AI.
“Government's survivability is based on its ability to control its citizens. The moment a society decides that the government shouldn’t exist, that instantly becomes the enemy number one for them,” he said. “So what would a government have to do with AI? It’s because they can control it, or allow it in a pseudo fashion where they can control it if they need to.”
For this dystopian future to never be realized, both Ryan and David agree that decentralization is key. Morpheus is vital to ensure freedom, privacy, and unbiased access to artificial intelligence.
Pioneering Real World Solutions
Venice is the living embodiment that open-source can be on par with centralized AI.
The privacy-centric and censorship resistant ChatGPT alternative will use inference supplied by Morpheus ecosystem participants. Venice is capable of chatting, writing code, and generating images, always protecting its users personal data.
As Bitcoin OG Erik Voorhees said during the fireside chat at DeAI Day, “the machines right now treat you like you are some kind of child, putting guardrails up for you. Venice is decentralized, permissionless, private, and easy-to-use, so people can enjoy the benefits of generative AI without censorship and without being spied on.”
But to ensure a bright future, Erik sees one big challenge the open-source community has to overcome: efficiently training open-source AI models in a decentralized environment.
Because for now, he said, it remains extremely cost-inefficient, especially compared to its centralized counterparts.
Building Roads for Empowering People with AI
David and Ryan imagine a future in which decentralized AI empowers individuals, protects their privacy, remains censorship-resistente, and automation lives at their fingertips.
Ryan sees a world where a user tells an AI: “OK, here is $10,000. Now make me as much money as you can over the next seven days, and then come back to me.” According to him, "The model will have all the access needed to complete the task,” said Ryan.
He explained that the only way to achieve that future is through smart systems that can talk to each other. Part of our mission, he claimed, is building the web of interconnectivity for these smart agents to be able to crawl across.
AI Plus Crypto Is The Perfect Match
Symbiotically, AI and crypto are a match made in heaven.
David sees AI as an easy-to-use UI for crypto, whose main issue is the lack of a proper user-friendly interface: “At this stage, crypto is like the Internet without search engines, where you had to manually input every IP address you wanted to visit.”
Mike Anderson (6079), Jonathan Jaranilla (Exabits), Imgesu Cetin (Genie AI), Trevor Harries-Jones (Render Network), Anna Kazlauskas (Vana), and Willy Ogorzaly (Nounspace) delved deeper into this matter emphasizing on the relevance of ownership, not just for data, but also compute power and intellectual property (identity).
They all agreed that this is how the future of intelligence is being built.
For them–as well as Erik–the goal is to have individual computers that are malleable and usable for AI training. However, this is not a challenge that can be accomplished individually but rather through collaboration in the ecosystem.