The Hope Axis by Anna Gát Emmett Shear - Explaining AI to the Humanities
This week on The Hope Axis, I’m joined by Emmett Shear to talk about AI, hope, and how we should actually think about the future. Emmett is the co-founder and former CEO of Twitch, former interim CEO of OpenAI, and now the co-founder and CEO of Softmax. We discuss why AI inspires so much fear, what it really is capable of, and where hope still shows up. Hope you enjoy.
For the full transcript of our conversation, click here: https://drive.google.com/file/d/10qooWrDB9Nr_0MgKy_kkpd1XMo_2gMN0/view?usp=drive_link
Important Links:
X (Twitter) – https://x.com/eshear?lang=en
Wikipedia – https://en.wikipedia.org/wiki/Emmett_Shear
Softmax – https://softmax.com/about
LinkedIn – https://www.linkedin.com/in/emmettshear
Show Notes:
(0:00:00) Introduction
(0:03:00) AI and The Humanities
(0:06:29) Reframing "The Essence", or "The Core" of Things
(0:13:01) Why People Want to Work on AI
(0:20:13) Kant As The Grandfather of AI
(0:23:57) Thomas Aquinas, Informing Onto The Boundary
(0:29:41) Back To Teleological Thinking
(0:31:44) Art As Imprinting One's Soul On An Object
(0:35:00) "Real" and "Fake" Interiority, AI's Nature Without Nurture
(0:41:46) AI & Art
(0:54:40) A Village of AI
(1:01:58) The Risk of Human Extinction
(1:08:47) AI Slop, Human Experience & Taste
(1:12:22) The Beauty, The Good, and The True as Temporal Coarse-Grainings
(1:18:08) Collective Aliveness of The Biosphere
(1:24:05) Problems Too Big For Individuals, But Not Too Big For Humanity
Books, Essays & References Mentioned:
If Anyone Builds It, Everyone Will Die — Eliezer Yudkowsky
Superintelligence — Nick Bostrom
Thomas Aquinas (informatio; origin of “information”)
Gottfried Wilhelm Leibniz
Immanuel Kant
Takens’ Embedding Theorem
Homotopy / homotopic stability
Groupoids (one-, two-, three-, infinity-groupoid)
AGI (Artificial General Intelligence)
Optimizers & goal-binding
Telos vs goals
Taste vs skill
Beauty, Truth, and the Good
Learning systems & dynamic systems
The biosphere / Gaia hypothesis
Collective intelligence
Speed as risk factor in AI systems