Guillaume Verdon, a physicist and quantum computing researcher, discusses topics such as the Effective Acceleration movement, the removal of anonymity on the internet, the future of AI, the Kardashev scale, and the power of humor and memes. They also talk about Jeff Bezos, physics-based computing and AI algorithms, and effective altruism.
Read more
AI Summary
Highlights
AI Chapters
Episode notes
auto_awesome
Podcast summary created with Snipd AI
Quick takeaways
E/acc movement advocates for rapid technological progress and AI advancement as the ethically optimal course of action.
Maintaining variance in AI development prevents monopolies and centralized power, allowing for decentralized control of technology.
Anonymity and pseudonyms create a diverse marketplace of ideas, promoting growth and progress of civilization.
Complexity of predicting the future accurately highlights the need for freedom of speech and market-driven approach to AI development.
Combining physics-based AI with anthropomorphic AI can lead to more accurate and relatable artificially intelligent systems.
Deep dives
Multiple Identities: Guillaume Verdun and Bev Jazos on X
Guillaume Verdun and Bev Jazos on X are two identities that coexist in the mind of one individual. Guillaume is a physicist and quantum machine learning researcher, while Bev Jazos is the creator of the effective acceleration as a movement (EAC). They discuss the intellectual foundation of EAC and its focus on propelling rapid technological progress as the ethically optimal course of action for humanity. EAC followers counter the view that AI is unpredictable and potentially dangerous, advocating for its continued advancement.
Exploring AI and Quantum Computing
Guillaume's work is centered around quantum machine learning and exploring intelligence beyond human capabilities inspired by physics. He believes that by building physics-based computing hardware and developing AI representations that think like nature, we can achieve a better understanding and representation of the universe. Guillaume emphasizes the importance of maintaining variance in AI development, allowing for a decentralized control of technology to prevent monopolies and centralized power. He argues that the combination of human and AI intelligence can lead to advancements in various fields.
The Role of Anonymity in Idea Evaluation
Bev Jazos on X, an anonymous account, was created to explore and share philosophical thoughts and ideas without being restricted by job status or personal achievements. The account aimed to encourage free speech, free thought, and efficient idea evaluation. Bev Jazos emphasized the importance of anonymity and pseudonyms in creating a diverse marketplace of ideas and promoting the growth and progress of civilization.
Concerns and Perspectives on AI and the Future
The conversation delves into perspectives on AI development, existential risks, and possible future scenarios. Guillaume argues against deterministic predictions and calculations of the probability of doom, highlighting the vast complexity of predicting the future accurately. He advocates for freedom of speech, individual agency, and a market-driven approach to AI development as a means to achieve fault-tolerant progress and prevent the concentration of power in the hands of a few.
Physics-based Computing and AI
The goal of Extraptic is to develop a physics-based computing system and physics-based AI algorithms that leverage out-of-equilibrium thermodynamics. The company aims to create a physical substrate for AI by utilizing first principles of physics, information theory, computation, and thermodynamics. It seeks to build a thermodynamic computer that combines deep learning and differentiable programming with quantum computing.
TensorFlow Quantum and Differentiable Quantum Programming
TensorFlow Quantum was a successful attempt to combine deep learning and differentiable classical programming with quantum computing. The infrastructure allowed for the running and combination of differentiable quantum programs as part of broader deep learning graphs. The project demonstrated that differentiating through control parameters and optimizing the system based on a loss function can effectively program physics-based computers.
The EAC Movement
The EAC movement, spearheaded by the Extropic team, is a viral and optimistic movement that celebrates entrepreneurship, progress, and the harnessing of techno-capital. The movement promotes innovation, adaptation, and the building of technologies that have the potential to shape a better future. It advocates for decentralized and diverse ideas, encouraging individual agency and the pursuit of growth and human flourishing.
AI Safety and Centralization Concerns
While Extropic and the EAC movement emphasize the importance of safety in AI development, they are cautious of proposals that may lead to centralized control. They argue that open source and decentralized contributions to AI research allow for greater innovation and a broader exploration of ideas. The goal is to maintain ubiquity and diversity in AI development, enabling a competitive landscape and avoiding concentration of power.
Importance of interdisciplinary teams
Working across disciplinary boundaries is always a challenge, but when a small team of smart and passionate individuals who trust each other come together, they can create something special. By teaching one another and combining their different expertise, they can form a network of knowledge that leads to innovative ideas and products. Despite the initial friction, the end result makes working in an interdisciplinary team worth it.
Physics-based AI and anthropomorphic AI
The future of AI lies in combining physics-based AI and anthropomorphic AI. Physics-based AI provides accurate world models at different scales, allowing for a better representation of nature. Anthropomorphic AI, on the other hand, provides human-like interfaces for communication. By merging these two approaches, it is possible to create a fully general artificially intelligent system that is grounded in physics while still being relatable to humans. This combination can lead to more accurate and productive AI systems.
Guillaume Verdon (aka Beff Jezos on Twitter) is a physicist, quantum computing researcher, and founder of e/acc (effective accelerationism) movement. Please support this podcast by checking out our sponsors:
– LMNT: https://drinkLMNT.com/lex to get free sample pack
– Notion: https://notion.com/lex
– InsideTracker: https://insidetracker.com/lex to get 20% off
– AG1: https://drinkag1.com/lex to get 1 month supply of fish oil
OUTLINE:
Here’s the timestamps for the episode. On some podcast players you should be able to click the timestamp to jump to that time.
(00:00) – Introduction
(09:18) – Beff Jezos
(19:16) – Thermodynamics
(25:31) – Doxxing
(35:25) – Anonymous bots
(42:53) – Power
(45:24) – AI dangers
(48:56) – Building AGI
(57:09) – Merging with AI
(1:04:51) – p(doom)
(1:20:18) – Quantum machine learning
(1:33:36) – Quantum computer
(1:42:10) – Aliens
(1:46:59) – Quantum gravity
(1:52:20) – Kardashev scale
(1:54:12) – Effective accelerationism (e/acc)
(2:04:42) – Humor and memes
(2:07:48) – Jeff Bezos
(2:14:20) – Elon Musk
(2:20:50) – Extropic
(2:29:26) – Singularity and AGI
(2:33:24) – AI doomers
(2:34:49) – Effective altruism
(2:41:18) – Day in the life
(2:47:45) – Identity
(2:50:35) – Advice for young people
(2:52:37) – Mortality
(2:56:20) – Meaning of life
Get the Snipd podcast app
Unlock the knowledge in podcasts with the podcast player of the future.
AI-powered podcast player
Listen to all your favourite podcasts with AI-powered features
Discover highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
Save any moment
Hear something you like? Tap your headphones to save it with AI-generated key takeaways
Share & Export
Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more
AI-powered podcast player
Listen to all your favourite podcasts with AI-powered features
Discover highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
