7 Highlights From Our DeAI Webinar Featuring Akash, The Lifted Initiative, And Swarm

Lou Kerner
Quantum Economics

--

On July 30th, CryptoOracle Labs held a webinar titled, “The Obstacles and Opportunities of Decentralized AI.” Our three esteemed panelists included Greg Osuri, the founder of Akash, Eric Bravick , CEO of The Lifted Initiative and Head of AI for CryptoOracle, and Gregor Zavcer, Director of the Swarm Foundation.

You can view the hour long replay of the insightful webinar:

Or you can read our seven highlights:

#1 — AI Related Tokens We’re among The Best Performing Sectors YTD Through June

Per a recent Grayscale research report, AI related tokens have been among the best performing sectors in all of crypto YTD thru June, 2024: :

#2 The Battle for Control Decentralized AI In The US Is Playing Out on the Capitol Hill

The battle for control of Decentralized AI in the U.S. is currently on display in the Senate and Congress. The battle is between open-source and closed-source AI. Both sides are ramping up their lobbying efforts. Closed source companies like Microsoft and Open AI argue that AI is fundamentally dangerous if not controlled, and that China will overtake U.S if we open-source our knowledge. The open source camp makes a compelling argument that China steals both open and and closed source tech, and open-sourcing enables more counter-threats. In addition, while the validity of an existential long term threat from AI is debatable, few see any imminent near term threats.

The consensus among the panelists is that neither the House Committee for Commerce and Trade nor the House Financial Services Committee sees an imminent AI threat to justify Congressional involvement today. Despite this consensus, the panelists appreciated that there are some in D.C., including the Biden administration, who see an existential threat to be addressed. As a result, the Biden administration recently passed an Executive Order to address those threats.

#3 — Llama 3’s Release Signifies A Seminal Point In Decentralized AI

The first two models for Meta’s Llama 3 were released on April 18, 2024. The 8B and 70B pre-trained models support a broad range of use cases and set new standards for performance across a wide range of industry benchmarks and offers improved reasoning. In his opening remarks, Greg stated that he believed that Llama 3.1 marks an inflection point for AI now that the most powerful model is open sourced vs. previous models from OpenAI that were closed source.

#4— The UX Needs Improve For DeAI To Scale

This is a general problem across all crypto sectors, and DeAI is no different. Getting people into DeAI involves users creating a wallet, learning how a wallet functions, and learning how cryptocurrency trading functions. That takes time and it’s complicated. Getting into CeAI doesn’t require any of that complicated crypto knowledge. So, our panelists all agreed, that for DeAI to scale, all the crypto-related complexity needs to be abstracted away and fade into the background. Similar to how Apple abstracted away all the complexity in using a pocket computer/iPhone.

#5 —Availability, Reliability, And Price Will Drive Demand For DeAI, Not Data Privacy Or Data Sovereignty

AI is trained on data, generally scraped from, or handed over from, an internet site, with little knowledge by most internet users. If asked, most users would opt out of their data being scraped. But most people don’t even care enough to take the time to go to the settings page to disable the scraping of their data. So while many in crypto we’re drawn here because of data privacy and data sovereignty, the vast majority of the planet could care less.

Per Greg, wDeAI demand will be driven by simply providing reasonably priced, and reliable, on-demand access to GPUs that are otherwise not available for most people, and/or are prohibitively expensive.

#6 — Data Provenance Is Critical For Data To Evolve Into A User Owned Asset

In his opening talk, Gregor noted that decentralized storage is a key aspect of decentralized data provenance, which enables a fair distribution of value for data. A fair distribution of value enables data to be an asset that can be user owned, and leveraged for new user owned foundational models.

#7 Alignment And Control of AI Systems Remains An Issue, And Crypto Is The Likely Solution

Even though we are increasingly encountering AI In our homes, offices, and on the street (e.g. Waymo), alignment and control of AI systems remain issues to be solved. Eric believes that the major question to answer with regards to alignment is “Aligned with who?”. And the answer to that question is likely to be found in crypto given our unparalleled experience with decentralized governance.

If you got at least 0.00000001 Bitcoin worth of value from this post please “Clap” below so others will see the post.

This content is for educational purposes only. It does not constitute trading advice. The author of this article may hold assets mentioned in the piece.

--

--

Lou Kerner
Quantum Economics

Believe Crypto is the biggest thing to happen in the history of mankind. Focused on community (founded the CryptoOracle Collective & CryptoMondays)