Jesse Clifton's guest appearances

Jesse Clifton

EA Talks

AI Safety Fundamentals: Alignment

Tim Urban

Peter Thiel

Ezra Klein

Erik Torenberg

Irene Pepperberg

Paul Shapiro

Allison Smith

Dan Brown

Nick Bostrom

Robin Hanson

Stuart Russell

Ben Hoffman

Daron Acemoglu

Brian Nosek

Rob Reich

OwlTail

Transformative artificial intelligence (TAI) may be a key factor in the long-run trajectory of civilization. A growing interdisciplinary community has begun to study how the development of TAI can be made safe and beneficial to sentient life (Bostrom 2014; Russell et al., 2015; OpenAI, 2018; Ortega and Maini, 2018; Dafoe, 2018). We present a research agenda for advancing a critical component of this effort: preventing catastrophic failures of cooperation among TAI systems. By cooperation failures we refer to a broad class of potentially-catastrophic inefficiencies in interactions among TAI-enabled actors. These include destructive conflict; coercion; and social dilemmas (Kollock, 1998; Macy and Flache, 2002) which destroy value over extended periods of time. We introduce cooperation failures at greater length in Section 1.1. Karnofsky (2016) defines TAI as ''AI that precipitates a transition comparable to (or more significant than) the agricultural or industrial revolution''. Such systems range from the unified, agent-like systems which are the focus of, e.g., Yudkowsky (2013) and Bostrom (2014), to the "comprehensive AI services’’ envisioned by Drexler (2019), in which humans are assisted by an array of powerful domain-specific AI tools. In our view, the potential consequences of such technology are enough to motivate research into mitigating risks today, despite considerable uncertainty about the timeline to TAI (Grace et al., 2018) and nature of TAI development.Original text:https://www.alignmentforum.org/s/p947tK8CoBbdpPtyK/p/KMocAf9jnAKc2jXriNarrated for AGI Safety Fundamentals by Perrin Walker of TYPE III AUDIO.Share feedback on this narration.

Transformative artificial intelligence (TAI) may be a key factor in the long-run trajectory of civilization. A growing interdisciplinary community has begun to study how the development of TAI can be made safe and beneficial to sentient life (Bostrom 2014; Russell et al., 2015; OpenAI, 2018; Ortega and Maini, 2018; Dafoe, 2018). We present a research agenda for advancing a critical component of this effort: preventing catastrophic failures of cooperation among TAI systems. By cooperation failures we refer to a broad class of potentially-catastrophic inefficiencies in interactions among TAI-enabled actors. These include destructive conflict; coercion; and social dilemmas (Kollock, 1998; Macy and Flache, 2002) which destroy value over extended periods of time. We introduce cooperation failures at greater length in Section 1.1. Karnofsky (2016) defines TAI as ''AI that precipitates a transition comparable to (or more significant than) the agricultural or industrial revolution''. Such systems range from the unified, agent-like systems which are the focus of, e.g., Yudkowsky (2013) and Bostrom (2014), to the "comprehensive AI services’’ envisioned by Drexler (2019), in which humans are assisted by an array of powerful domain-specific AI tools. In our view, the potential consequences of such technology are enough to motivate research into mitigating risks today, despite considerable uncertainty about the timeline to TAI (Grace et al., 2018) and nature of TAI development.Original text:<a href="https://www.alignmentforum.org/s/p947tK8CoBbdpPtyK/p/KMocAf9jnAKc2jXri" rel="noopener noreferrer nofollow" target="_blank">https://www.alignmentforum.org/s/p947tK8CoBbdpPtyK/p/KMocAf9jnAKc2jXri</a>Narrated for <a href="https://www.agisafetyfundamentals.com/" rel="noopener noreferrer nofollow" target="_blank">AGI Safety Fundamentals</a> by <a href="https://twitter.com/perrinjwalker” target=" rel='“noreferrer" nofollow' target="_blank">Perrin Walker</a> of <a href="https://type3.audio/" rel="noopener noreferrer nofollow" target="_blank">TYPE III AUDIO</a>.<a href="https://docs.google.com/forms/d/e/1FAIpQLScC_z9tBDWzxoS3D0EfNT0Kcjs_T4rNN3FyVwElSdyanel0rA/viewform?usp=pp_url&amp;entry.584848066=https://www.alignmentforum.org/s/p947tK8CoBbdpPtyK/p/KMocAf9jnAKc2jXri" rel="noopener noreferrer nofollow" target="_blank">Share feedback on this narration</a>.

[Week 7] “Cooperation, Conflict, and Transformative Artificial Intelligence: Sections 1 & 2 — Introduction, Strategy and Governance” by Jesse Clifton

AI systems are much more transparent to one another than humans are to other humans; they can share information about the reward functions that drive their actions. Jesse Clifton, a PhD candidate in statistics at North Carolina State University, presents a framework for using “open-source learning algorithms” — which, thanks to this transparency, can make bargains to achieve mutually beneficial outcomes.This talk was filmed at EA Global 2019: London. Find out how to attend a future EA Global conference: https://www.eaglobal.org/.Learn more about effective altruism: https://www.effectivealtruism.org/.Original VideoEffective Altruism is a social movement dedicated to finding ways to do the most good possible, whether through charitable donations, career choices, or volunteer projects. EA Global conferences are gatherings for EAs to meet. You can also listen to this talk along with its accompanying video on YouTube. 

AI systems are much more transparent to one another than humans are to other humans; they can share information about the reward functions that drive their actions. Jesse Clifton, a PhD candidate in statistics at North Carolina State University, presents a framework for using “open-source learning algorithms” — which, thanks to this transparency, can make bargains to achieve mutually beneficial outcomes.This talk was filmed at EA Global 2019: London. Find out how to attend a future EA Global conference: <a href="https://www.eaglobal.org/" rel="nofollow" target="_blank">https://www.eaglobal.org/</a>.Learn more about effective altruism: <a href="https://www.effectivealtruism.org/" rel="nofollow" target="_blank">https://www.effectivealtruism.org/</a>.<a href="https://www.youtube.com/watch?v=QKRDp8nUMtk" rel="nofollow" target="_blank">Original Video</a><a href="https://www.effectivealtruism.org/" rel="nofollow" target="_blank">Effective Altruism</a> is a social movement dedicated to finding ways to do the most good possible, whether through charitable donations, career choices, or volunteer projects. <a href="https://www.eaglobal.com/" rel="nofollow" target="_blank">EA Global</a> conferences are gatherings for EAs to meet. You can also listen to this talk along with its accompanying video on <a href="https://www.youtube.com/c/EffectiveAltruismVideos" rel="nofollow" target="_blank">YouTube</a>.

EAG London 2019: Open-source learning—A bargaining approach (Jesse Clifton)