Sam Lehman: What the Reinforcement Learning Renaissance Means for Decentralized AI

Sam Lehman: What the Reinforcement Learning Renaissance Means for Decentralized AI

Join Tommy Shaughnessy from Delphi Ventures as he hosts Sam Lehman, Principal at Symbolic Capital and AI researcher, for a deep dive into the Reinforcement Learning (RL) renaissance and its implications for decentralized AI. Sam recently authored a widely discussed post, "The World's RL Gym", exploring the evolution of AI scaling and the exciting potential of decentralized networks for training next-generation models.

The World’s RL Gym: https://www.symbolic.capital/writing/the-worlds-rl-gym



🎯 Key Highlights


The three phases of AI scaling: Pre-training, Inference Time Compute, and the RL Renaissance.

How DeepMind's novel RL approach (using GRPO) created powerful reasoning models with minimal human data.

Understanding "reasoning traces" and how models learn to "think" longer and more effectively.

The potential downsides of human preference data potentially inhibiting model creativity, drawing parallels to AlphaGo.

Exploring the "World's RL Gym" concept: Decentralizing RL through open environments, diverse tasks, and verified data.

Why open, collaborative RL environments might outperform closed-source labs in generating diverse AI strategies.

The critical role of high-quality base models for successful RL fine-tuning.

Future AI architectures: Continuous learning and the potential of modular Mixture-of-Experts (MoE) models.

Current landscape: Open-source vs. proprietary AI, the challenge of model lock-in, and the role of crypto networks.

Debunking recent claims that "RL is dead" and understanding its true impact.



💡 Want to stay updated with the latest in crypto & AI? Hit subscribe and the notification bell! 🔔



🧠 Follow the Alpha


Tommy's Twitter: @Shaughnessy119

Sam's Twitter: @SPLehman

Symbolic Capital’s Twitter: @symbolicvc



🔗 Connect with Delphi


🌐 Portal: https://delphidigital.io/

🐦 Twitter: https://twitter.com/delphi_digital

💼 LinkedIn: https://www.linkedin.com/company/delphi-digital



🎧 Listen on


Spotify: https://open.spotify.com/show/62PR1RigLG2YN5Pelq6UY9?si=18ac7ccf36ab4753

Apple Podcasts: https://podcasts.apple.com/us/podcast/the-delphi-podcast/id1438148082

Youtube: https://www.youtube.com/channel/UC9Yy99ZlQIX9-PdG_xHj43Q



Timestamps


00:00 - Introduction: Sam Lehman, Symbolic Capital & "The World's RL Gym"

01:30 - History of AI Scaling: Pre-training Era

03:30 - Phase 2: Inference Time Compute Scaling

09:30 - Phase 3: The RL Renaissance & DeepMind Moment

14:30 - How DeepMind Trained R1 without Human Preferences

16:30 - AlphaGo Analogy: Human Data Inhibiting Creativity?

20:30 - Generalizability of RL Training: How Far Does It Go?

22:30 - The "Aha Moment": Models Learning to Think Longer

25:30 - Concept: Decentralized RL & The World's Gym

31:30 - Why Decentralize RL? Open Collaboration vs. Closed Labs

35:00 - Understanding Reasoning Traces

39:00 - Current Decentralized RL Projects (Prime Intellect, General Reasoning)

41:30 - Future Architectures: Continuous Improvement & Modular Models

46:30 - Open Source vs. Proprietary AI: Landscape & Challenges

50:30 - The Lock-In Problem with Foundational Models

52:30 - Is AGI Here? Experiences with GPT-4o

56:30 - Investment Focus in Decentralized AI

59:00 - Modular MoE Models & Jensen's HDEE Paper

1:03:00 - Debunking "RL is Dead" Claims

1:06:00 - Importance of Performant Base Models for RL



Disclaimer


This podcast is strictly informational and educational and is not investment advice or a solicitation to buy or sell any tokens or securities or to make any financial decisions. Do not trade or invest in any project, tokens, or securities based upon this podcast episode. The host and members at Delphi Ventures may personally own tokens or art that are mentioned on the podcast. Our current show features paid sponsorships which may be featured at the start, middle, and/or the end of the episode. These sponsorships are for informational purposes only and are not a solicitation to use any product, service or token.

Avsnitt(468)

Pascal Gauthier, President of Ledger: The Future of Crypto Custody

Pascal Gauthier, President of Ledger: The Future of Crypto Custody

Pascal Gauthier, President of Ledger, the provider of the popular crypto hardware wallet to update us on the future of the company. Topics include Komainu (Ledger partnered with Nomura to create a full-service institutional crypto service), Ledger Vault and the future of the company.   - Ledger has sold 1M+ hardware wallets; whats next on the roadmap and functionality.   - A discussion on Ledger Vault the institutional custodian offering; custody can unlock billions of inflows in our opinion.   - How ledger Vault differentiation vs Gemini, Coinbase, DACC, Kingdom Trust and competitors.   - Why ledger is the most secure institutional custody provider.   - Ledger’s institutional custodianship offering is competitively priced, especially in comparison to gold.   - Details on Komainu the first full custodian ready for institutional money. With Nomura. Komainu will hold the private keys.   - Ledger can easily implement updates faster than exchanges to support new cryptocurrencies   - Ledger’s future plans to support security tokens.     Add your email on 51pct.io for our extensive research reports. 51percent's Institutional Crypto Podcasts are to the point discussions with crypto leaders for analysts, funds and institutions. Make sure to add your email on 51pct.io Disclosure: Tom Shaughnessy owns tokens in ETH. This podcast is NOT investment advice and is only informational. Do not make investment decisions based upon this podcast.

15 Okt 201831min

Myles Snider of Aurora EOS: A Response To The Massive Controversy On The 5th Largest Blockchain

Myles Snider of Aurora EOS: A Response To The Massive Controversy On The 5th Largest Blockchain

Today on 51percent's Institutional Podcast we have Myles Snider, the CEO of Aurora EOS a candidate block producer for EOS. Myles was also previously head of research at Multicoin Capital.  EOS has faced intense scrutiny over the past few weeks, and we take a detailed and unbiased look under the hood.   Unbiased deep dive covering everything EOS   Myles’ response to the massive controversy regarding EOS (limited nodes, vote buying, cartels, the future, renting out resources)   Will EOS solve its issues before Ethereum is able to scale with Level 2?   Can BlockOne actually deploy its $4B?   Info on EOS' 21 validators, how they can change and how 70 are compensated on standby.    Details on the resource exchange to lend out EOS for a return   How the EOS constitution cannot be enforced by code.   The beauty of on chain governance and so much more. Add your email on 51pct.io for our extensive research reports. 51percent's Institutional Crypto Podcasts are to the point discussions with crypto leaders for analysts, funds and institutions. Make sure to add your email on 51pct.io Disclosure: Tom Shaughnessy owns tokens in ETH and has no position in EOS. This podcast is NOT investment advice and is only informational. Do not make investment decisions based upon this podcast.

11 Okt 20181h 9min

ConsenSys Capital Co-Founder Andrew Keys: The Future of Ethereum and ConsenSys

ConsenSys Capital Co-Founder Andrew Keys: The Future of Ethereum and ConsenSys

Andrew Keys, the Co-Founder of ConsenSys Capital discusses all things blockchain and the future of both ConsenSys and Crypto - a must listen to episode. - ConsenSys Capital Overview (token Foundry, consensus digital securities, consensus ventures, Trustology, Balan3e)   -Custodianship is the main factor for institutional adoption, through Trustology.   - What types of investments ConsenSys Ventures is making (Rocket Pool, Exchanges)   - Brooklyn Project (regulatory frameworks for tokens), Civil for Journalism, focus on consumer utility tokens as software licenses.   Add your email on 51pct.io for our extensive research reports. 51percent's Institutional Crypto Podcasts are to the point discussions with crypto leaders for analysts, funds and institutions. Make sure to add your email on 51pct.io Disclosure: Tom Shaughnessy owns tokens in ETH. This podcast is NOT investment advice and is only informational. Do not make investment decisions based upon this podcast.

10 Okt 201857min

Polymath Co-Founder and CEO Trevor Koverko: The Security Token War

Polymath Co-Founder and CEO Trevor Koverko: The Security Token War

In the first episode of 51percent's Institutional Crypto Podcast, we have the co-founder and CEO of Polymath, Trevor Koverko. Polymath is one of the most well know and largest security token platforms, and is leading the charge to securitize potentially trillions in assets.  In this episode we dive into - The creation of Polymath State of security token platforms Size of the security token market How Trevor's mentality has changed over the years since launching Polymath, and how Polymath wants to work with institutions. Will Polymath stay linked to Ethereum? 40+ tokens have been launched on the platform. "The security token market will be bigger than the market for utility tokens" There are no demand problems on attracting entities interested in doing security token offerings. Existing banks are going to have their taxi cab moment and if they don't take us seriously, they're going to get uberized"  Add your email on 51pct.io for our extensive research reports. 51percent's Institutional Crypto Podcasts are to the point discussions with crypto leaders for analysts, funds and institutions. Make sure to add your email on 51pct.io Disclosure: Tom Shaughnessy owns tokens in ETH and POLY. This podcast is NOT investment advice and is only informational. Do not make investment decisions based upon this podcast.

4 Okt 201833min

Populärt inom Teknik

uppgang-och-fall
elbilsveckan
market-makers
rss-racevecka
rss-elektrikerpodden
rss-uppgang-och-fall
bli-saker-podden
skogsforum-podcast
natets-morka-sida
rss-technokratin
developers-mer-an-bara-kod
rss-veckans-ai
har-vi-akt-till-mars-an
solcellskollens-podcast
mediepodden
bilar-med-sladd
hej-bruksbil
rss-it-sakerhetspodden
rss-fabriken-2
rss-laddstationen-med-elbilen-i-sverige