Published: August 4, 2023
4
7
23

I found this slide by @ai_risks' @DanHendrycks and thought: "Everyone should enjoy it!" Here's some background: đź§µ

Image in tweet by Nirit Weiss-Blatt, PhD

May 30: Dan Hendrycks promoted the second X-Risk Open Letter (initiated by his new "Center for AI Safety") https://x.com/DanHendrycks/sta...

May 31: Dan Hendrycks in @TIME magazine: "Evolution tends to produce selfish behavior. Evolutionary pressure will likely ingrain AIs with behaviors that promote self-preservation. …a pathway toward being supplanted as the Earth's dominant species." https://x.com/TIME/status/1664...

July 7: The @BostonGlobe published this profile: "Dan Hendrycks wants to save us from an AI catastrophe. He's not sure he'll succeed." Apparently, his evangelical beliefs have been replaced by a new religion: AI Doomerism. https://archive.vn/Sn1v0

Image in tweet by Nirit Weiss-Blatt, PhD

This profile was praised by the "AI Safety" community as being "great and moving." It only "moved" me toward Hendrycks' ML Safety online course – in order to "understand the danger as he does"…

Image in tweet by Nirit Weiss-Blatt, PhD

The "Introduction to ML Safety" course (https://course.mlsafety.org/ab... has a comprehensive syllabus. Among the lessons, the "X-Risk Overview" was "Based on slides from Cullen O'Keefe" – Policy & Governance at @OpenAI (https://cullenokeefe.com/)

Image in tweet by Nirit Weiss-Blatt, PhD

The most striking presentation was: "Natural Selection Favors AIs over Humans" It is based on Dan Hendrycks' paper (with the same title): https://arxiv.org/abs/2303.162...

Image in tweet by Nirit Weiss-Blatt, PhD

This is how Dan Hendrycks framed his paper https://x.com/DanHendrycks/sta...

Image in tweet by Nirit Weiss-Blatt, PhD

The "Natural Selection Favors AIs over Humans" paper was criticized, for example: https://x.com/jeffjarvis/statu...

Image in tweet by Nirit Weiss-Blatt, PhD

It's pretty entertaining to go down the Darwinian rabbit hole. But could we please avoid letting these Safety guys determine how #AI should be regulated? Thanks.

Image in tweet by Nirit Weiss-Blatt, PhD

@DrTechlash @BostonGlobe That article also states that Hendrycks “was never an EA adherent, even if he brushed up against the movement,” though a post coauthored by Hendrycks himself on the EA Forum says “he was advised by Bastian Stern (now at Open Philanthropy) to get into AI to reduce x-risk, and

@DrTechlash @BostonGlobe This title implies an insult. As if you can only be either an evangelical or a computer scientist. Of course, there are many who are both.

@DrTechlash @BostonGlobe it's fun how these guys have so many important things to do and yet can appear on so much media. here in german sunday-tv for grandmas. I guess their real business is fear and their investments....

Image in tweet by Nirit Weiss-Blatt, PhD

@DrTechlash @BostonGlobe These fantasies are the strength of our species, combined with the fierce debates around them. Some fantasies come true but luckily most doesn’t.

Share this thread

Read on Twitter

View original thread

Navigate thread

1/15