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)

Noblebridge’s Tyrone V. Ross: Crypto Wealth Management

Noblebridge’s Tyrone V. Ross: Crypto Wealth Management

Host Tom Shaughnessy of Delphi Digital (DelphiDigital.io) is joined by Tyrone V. Ross Jr, a managing partner at Noble Bridge Wealth Management.  The long story short is Tyrone is the crypto wealth manager of the future. Tyrone has decided to differentiate his practice by offering his wealth management services to the general public, unlike the majority of wealth managers and financial advisors who not only won’t help or touch crypto assets, but frankly most are not up to speed on the space. It’s great to hear Tyrone’s journey, his unreal energy, and how he helps clients across the spectrum with their crypto assets. While normal routines like crypto custody, buying/selling, and what to do in certain situations is easy for crypto natives to understand, Tyrone is serving the needs of the wider population outside of the crypto echo chamber. We need to educate and help the global population who are not crypto native get involved with crypto, and thats what Tyrone is doing every day. To access the insights package of Delphi's leading crypto research, visit DelphiDigital.io on your device and sign up using coupon code CHAINREACTION.   Follow Tom on Twitter @Shaughnessy119 Follow Tyrone on Twitter @TR401   Disclosure: Tom Shaughnessy owns tokens in BTC, ETH, DCR, MKR, XTZ and Loom. This podcast is NOT investment advice and is only informational. Do not make investment decisions based upon this podcast. Delphi Digital was not compensated by any party for this podcast episode other than Podbean's advertisers. This content is strictly informational and is not a solicitation to buy or sell any security or token.  - Advertisers: To advertise on this podcast, email Tom@DelphiDigital.io Potential Guests: If you're interested in appearing on the podcast, email Tom@DelphiDigital.io  -

3 Juni 201957min

Enigma’s Tor Bair: Disrupting Facebook and Solving Privacy for Web 3.0

Enigma’s Tor Bair: Disrupting Facebook and Solving Privacy for Web 3.0

Host Tom Shaughnessy of Delphi Digital (DelphiDigital.io) is joined by Tor Bair, Head Of Growth at Enigma. For those new to Enigma it is described as  a decentralized secure computation protocol, where “secret nodes” in the network perform computations over encrypted data. Enigma brings privacy to any kind of computation - not just transactions. Our conversation covers the desire for privacy despite consumers being ok giving away their data for free, if the incumbents (facebook/google can be displaced) and the entire mission of Enigma, an overview of how the protocol works and what Enigma offers. To access the insights package of Delphi's leading crypto research, visit DelphiDigital.io on your device and sign up using coupon code CHAINREACTION.   Follow Tom on Twitter @Shaughnessy119 Follow Joe on Twitter @TorBair    Disclosure: Tom Shaughnessy owns tokens in BTC, ETH, DCR, MKR, XTZ and Loom. This podcast is NOT investment advice and is only informational. Do not make investment decisions based upon this podcast. Delphi Digital was not compensated by any party for this podcast episode other than Podbean's advertisers. This content is strictly informational and is not a solicitation to buy or sell any security or token.  - Advertisers: To advertise on this podcast, email Tom@DelphiDigital.io Potential Guests: If you're interested in appearing on the podcast, email Tom@DelphiDigital.io  -

28 Maj 201956min

ConsenSys’ Joe Lubin: Ethereum’s Competition Isn’t Even Close

ConsenSys’ Joe Lubin: Ethereum’s Competition Isn’t Even Close

Host Tom Shaughnessy of Delphi Digital (DelphiDigital.io) is joined by Joe Lubin the co-founder of Ethereum and thee founder of ConsenSys.  For those new to the space, ConsenSys is a global blockchain technology company building the infrastructure, applications, and practices that enable a decentralized world. ConsenSys has over 50 spokes building on Ethereum ranging from infrastructure like Infura and Metamask to platforms and applications like OpenLaw and Airswap. After having numerous guests on the podcast from ConsenSys, from Gregory Rocco to Andrew Keys and others, it was excellent to have Joe on to explain his take on the space. We cover the goals of blockchain, Ethereum’s competition and its future under Serenity, the past, present and future of ConsenSys and so much more. Joe is by far one of the most plugged in people to the space, as such this was a very insightful conversation. To access the insights package of Delphi's leading crypto research, visit DelphiDigital.io on your device and sign up using coupon code CHAINREACTION.   Follow Tom on Twitter @Shaughnessy119 Follow Joe on Twitter @ethereumJoseph    Photo Credit Zaza Weissgerber Disclosure: Tom Shaughnessy owns tokens in BTC, ETH, DCR, MKR, XTZ and Loom. This podcast is NOT investment advice and is only informational. Do not make investment decisions based upon this podcast. Delphi Digital was not compensated by any party for this podcast episode other than Podbean's advertisers. This content is strictly informational and is not a solicitation to buy or sell any security or token.  - Advertisers: To advertise on this podcast, email Tom@DelphiDigital.io Potential Guests: If you're interested in appearing on the podcast, email Tom@DelphiDigital.io  -

6 Maj 201941min

Evan Feng: From Citadel and Point72 To Founding Tapestry Capital

Evan Feng: From Citadel and Point72 To Founding Tapestry Capital

Host Tom Shaughnessy of Delphi Digital (DelphiDigital.io) is joined by Even Feng, the Founder and CIO of Tapestry Capital. I wanted to have Evan on since he has a storied career in finance and incredible insight given his positions with Barclays’ investment banking department to Citadel and then to Point 72 before creating tapestry capital. Evan is able to link the legacy financial world, with Crypto, and he able to explain the differences between the two worlds. We close as Evan describes Tapestry capital, how he plans to differentiate, and so much more.   To access the insights package of Delphi's leading crypto research, visit DelphiDigital.io on your device and sign up using coupon code CHAINREACTION.   Follow Tom on Twitter @Shaughnessy119 Follow Evan on Twitter @EvanTheFeng  Visit Tapestry Capital    Disclosure: Tom Shaughnessy owns tokens in BTC, ETH, DCR, MKR, XTZ and Loom. This podcast is NOT investment advice and is only informational. Do not make investment decisions based upon this podcast. Delphi Digital was not compensated by any party for this podcast episode other than Podbean's advertisers. This content is strictly informational and is not a solicitation to buy or sell any security or token.  - Advertisers: To advertise on this podcast, email Tom@DelphiDigital.io Potential Guests: If you're interested in appearing on the podcast, email Tom@DelphiDigital.io  -

2 Maj 201944min

Vision Hill Group’s Scott Army: Digital Asset Management of the Future

Vision Hill Group’s Scott Army: Digital Asset Management of the Future

Host Tom Shaughnessy of Delphi Digital (DelphiDigital.io) is joined by Scott Army, the Founder and CEO of Vision Hill Group. While institutions are starting to warm up to  crypto today, Scott began work on his Crypto fund of funds years ago, and its first strategy successfully launched over 6 months ago. Since then, Scott is growing Vision Hill into a full service Crypto company focused on adding new strategies, providing research and advisory services as well. Scott has unparalleled insight into the crypto fund scene since Vision Hill is conducting extensive research into these managers, and the data they are compiling is also massive To access the insights package of Delphi's leading crypto research, visit DelphiDigital.io on your device and sign up using coupon code CHAINREACTION.   Follow Tom on Twitter @Shaughnessy119 Follow Scott on Twitter @scottarmy_   Disclosure: Tom Shaughnessy owns tokens in BTC, ETH, DCR, MKR, and XTZ. This podcast is NOT investment advice and is only informational. Do not make investment decisions based upon this podcast. Delphi Digital was not compensated by any party for this podcast episode other than Podbean's advertisers. This content is strictly informational and is not a solicitation to buy or sell any security or token.  - Advertisers: To advertise on this podcast, email Tom@DelphiDigital.io Potential Guests: If you're interested in appearing on the podcast, email Tom@DelphiDigital.io  -

29 Apr 201946min

CoinList’s Andy Bromberg: The Future of Funding and Community-Building In The Crypto Space

CoinList’s Andy Bromberg: The Future of Funding and Community-Building In The Crypto Space

Host Tom Shaughnessy of Delphi Digital (DelphiDigital.io) is joined by Andy Bromberg, the Co-Founder and President of CoinList. Andy is one of the most insightful guests I’ve had on, his ability to distill everything happening in crypto into digestable and understandable pieces is remarkable. On this episode we covered so much, including the future of funding, developer focused community building, a complete rundown on CoinList, and how CoinList only accepts 0.02% of the projects it is approached by. We also discussed governance, competition and regulations. To access the insights package of Delphi's leading crypto research, visit DelphiDigital.io on your device and sign up using coupon code ChainReaction.   Follow Tom on Twitter @Shaughnessy119 Follow Andy on Twitter @andy_bromberg   Disclosure: Tom Shaughnessy owns tokens in ETH, DCR, MKR, and BTC. This podcast is NOT investment advice and is only informational. Do not make investment decisions based upon this podcast. Delphi Digital was not compensated by any party for this podcast episode other than Podbean's advertisers. This content is strictly informational and is not a solicitation to buy or sell any security or token.  - Advertisers: To advertise on this podcast, email Tom@DelphiDigital.io Potential Guests: If you're interested in appearing on the podcast, email Tom@DelphiDigital.io  -

12 Apr 201952min

ConsenSys’ Ajit Tripathi - ICOs, STOs and Institutional Digital Assets: How Tokens are Digitizing Financial Markets

ConsenSys’ Ajit Tripathi - ICOs, STOs and Institutional Digital Assets: How Tokens are Digitizing Financial Markets

Host Tom Shaughnessy of Delphi Digital (DelphiDigital.io) is joined by Ajit Tripathi, a partner at ConsenSys focused on the solutions segment. We cover how tokens will invade enterprises in full force, building liquid and efficient private markets through tokenization, the benefits to various market participants and so much more.   To access all of Delphi's leading crypto research, visit DelphiDigital.io on your device and sign up now.   Follow Tom on Twitter @Shaughnessy119 Follow Ajit on Twitter @chainyoda    Disclosure: Tom Shaughnessy owns tokens in ETH, DCR, MKR, ZRX and HYDRO. This podcast is NOT investment advice and is only informational. Do not make investment decisions based upon this podcast. Delphi Digital was not compensated by any party for this podcast episode other than Podbean's advertisers. This content is strictly informational and is not a solicitation to buy or sell any security or token.  - Advertisers: To advertise on this podcast, email Tom@DelphiDigital.io Potential Guests: If you're interested in appearing on the podcast, email Tom@DelphiDigital.io  -

9 Apr 201946min

Cent’s Max Brody and Cameron Hejazi: The Social Income Network

Cent’s Max Brody and Cameron Hejazi: The Social Income Network

Host Tom Shaughnessy of Delphi Digital (DelphiDigital.io) is joined by CENT's co-founder's Cameron Hejazi to hear the story of two relentless co-founders and their vision to create a new social income network. The conversation covers the current CENT platform, how users are earning money seeding, why the two co-founders built on Ethereum, and so much more.   To access all of Delphi's leading crypto research, visit DelphiDigital.io on your device and sign up now.   Follow Tom on Twitter @Shaughnessy119 Follow Max on Twitter @maxbrody Follow Cameron on Twitter @chejazi    Disclosure: Tom Shaughnessy owns tokens in ETH, DCR, MKR, ZRX and HYDRO. This podcast is NOT investment advice and is only informational. Do not make investment decisions based upon this podcast. Delphi Digital was not compensated by any party for this podcast episode other than Podbean's advertisers. This content is strictly informational.  - Advertisers: To advertise on this podcast, email Tom@DelphiDigital.io Potential Guests: If you're interested in appearing on the podcast, email Tom@DelphiDigital.io  -

26 Mars 201950min

Populärt inom Teknik

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