29 points

27 comments1 min readLW link

[Question] Has Moore’s Law actually slowed down?

Matthew Barnett20 Aug 2019 19:18 UTC

14 points

7 comments1 min readLW link

Cerebras Systems unveils a record 1.2 trillion transistor chip for AI

avturchin20 Aug 2019 14:36 UTC

7 points

4 comments1 min readLW link

(venturebeat.com)

Lisbon SSC Meetup #1

tamkin&popkin20 Aug 2019 12:20 UTC

1 point

0 comments1 min readLW link

Self-supervised learning & manipulative predictions

Steven Byrnes20 Aug 2019 10:55 UTC

18 points

14 comments9 min readLW link

Negative “eeny meeny miny moe”

jefftk20 Aug 2019 2:48 UTC

25 points

6 comments1 min readLW link

Why I Am Not a Technocrat

Spugpow20 Aug 2019 2:06 UTC

−3 points

4 comments1 min readLW link

(radicalxchange.org)

A misconception about immigration

limerott19 Aug 2019 22:37 UTC

1 point

9 comments4 min readLW link

(limerott.com)

[Question] Do We Change Our Minds Less Often Than We Think?

Raemon19 Aug 2019 21:37 UTC

20 points

5 comments1 min readLW link

Classifying specification problems as variants of Goodhart’s Law

Vika19 Aug 2019 20:40 UTC

72 points

5 comments5 min readLW link 1 review

Unstriving

Jacob Falkovich19 Aug 2019 14:31 UTC

38 points

7 comments6 min readLW link

Goodhart’s Curse and Limitations on AI Alignment

Gordon Seidoh Worley19 Aug 2019 7:57 UTC

25 points

18 comments10 min readLW link

Raph Koster on Virtual Worlds vs Games (notes)

Raemon18 Aug 2019 19:01 UTC

26 points

8 comments2 min readLW link

“Can We Survive Technology” by von Neumann

Ben Pace18 Aug 2019 18:58 UTC

32 points

2 comments1 min readLW link

(geosci.uchicago.edu)

Prokaryote Multiverse. An argument that potential simulators do not have significantly more complex physics than ours

mako yass18 Aug 2019 4:22 UTC

0 points

5 comments2 min readLW link

Neural Nets in Python 1

lifelonglearner18 Aug 2019 2:48 UTC

10 points

3 comments8 min readLW link

Inspection Paradox as a Driver of Group Separation

Shmi17 Aug 2019 21:47 UTC

29 points

0 comments1 min readLW link

South Bay Meetup

David Friedman17 Aug 2019 19:56 UTC

1 point

0 comments1 min readLW link

Problems in AI Alignment that philosophers could potentially contribute to

Wei Dai17 Aug 2019 17:38 UTC

79 points

14 comments2 min readLW link

[Question] How can you use music to boost learning?

Matthew Barnett17 Aug 2019 6:59 UTC

11 points

1 comment1 min readLW link

A Primer on Matrix Calculus, Part 3: The Chain Rule

Matthew Barnett17 Aug 2019 1:50 UTC

12 points

4 comments6 min readLW link

Nashville SSC September Meetup

friedelcraftiness16 Aug 2019 15:16 UTC

1 point

0 comments1 min readLW link

Beliefs Are For True Things

Davis_Kingsley15 Aug 2019 23:23 UTC

8 points

5 comments3 min readLW link

[Question] What experiments would demonstrate “upper limits of augmented working memory?”

Raemon15 Aug 2019 22:09 UTC

33 points

6 comments2 min readLW link

Clarifying some key hypotheses in AI alignment

Ben Cottier and Rohin Shah

15 Aug 2019 21:29 UTC

79 points

12 comments9 min readLW link

Tessercube — OpenPGP Made Mobile

Suji Yan15 Aug 2019 9:34 UTC

4 points

0 comments1 min readLW link

A Primer on Matrix Calculus, Part 2: Jacobians and other fun

Matthew Barnett15 Aug 2019 1:13 UTC

22 points

7 comments7 min readLW link

Partial summary of debate with Benquo and Jessicata [pt 1]

Raemon14 Aug 2019 20:02 UTC

89 points

63 comments22 min readLW link 3 reviews

“Designing agent incentives to avoid reward tampering”, DeepMind

gwern14 Aug 2019 16:57 UTC

28 points

15 comments1 min readLW link

(medium.com)

Subagents, trauma and rationality

Kaj_Sotala14 Aug 2019 13:14 UTC

111 points

4 comments19 min readLW link

Predicted AI alignment event/meeting calendar

rmoehn14 Aug 2019 7:14 UTC

29 points

14 comments1 min readLW link

Natural laws should be explicit constraints on strategy space

ryan_b13 Aug 2019 20:22 UTC

8 points

6 comments1 min readLW link

Distance Functions are Hard

Grue_Slinky13 Aug 2019 17:33 UTC

31 points

19 comments6 min readLW link

Book Review: Secular Cycles

Scott Alexander13 Aug 2019 4:10 UTC

62 points

10 comments16 min readLW link 1 review

(slatestarcodex.com)

A Primer on Matrix Calculus, Part 1: Basic review

Matthew Barnett12 Aug 2019 23:44 UTC

25 points

4 comments7 min readLW link

[Question] What explanatory power does Kahneman’s System 2 possess?

Richard_Ngo12 Aug 2019 15:23 UTC

31 points

2 comments1 min readLW link

Mesa-Optimizers and Over-optimization Failure (Optimizing and Goodhart Effects, Clarifying Thoughts—Part 4)

Davidmanheim12 Aug 2019 8:07 UTC

15 points

3 comments4 min readLW link

Adjectives from the Future: The Dangers of Result-based Descriptions

Pradeep_Kumar11 Aug 2019 19:19 UTC

19 points

8 comments11 min readLW link

[Question] Could we solve this email mess if we all moved to paid emails?

jacobjacob11 Aug 2019 16:31 UTC

29 points

50 comments4 min readLW link

AI Safety Reading Group

Søren Elverlin11 Aug 2019 9:01 UTC

16 points

8 comments1 min readLW link

[Question] Does human choice have to be transitive in order to be rational/consistent?

jmh11 Aug 2019 1:49 UTC

9 points

6 comments1 min readLW link

Diana Fleischman and Geoffrey Miller—Audience Q&A

Jacob Falkovich10 Aug 2019 22:37 UTC

38 points

6 comments9 min readLW link

Intransitive Preferences You Can’t Pump

zulupineapple9 Aug 2019 23:10 UTC

0 points

2 comments1 min readLW link

Categorial preferences and utility functions

DavidHolmes9 Aug 2019 21:36 UTC

10 points

6 comments5 min readLW link

[Question] What is the state of the ego depletion field?

Eli Tyre9 Aug 2019 20:30 UTC

27 points

10 comments1 min readLW link

Why Gradients Vanish and Explode

Matthew Barnett9 Aug 2019 2:54 UTC

25 points

9 comments3 min readLW link

AI Forecasting Dictionary (Forecasting infrastructure, part 1)

jacobjacob and bgold

8 Aug 2019 16:10 UTC

50 points

0 comments5 min readLW link

[Question] Why do humans not have built-in neural i/o channels?

Richard_Ngo8 Aug 2019 13:09 UTC

25 points

23 comments1 min readLW link

Which of these five AI alignment research projects ideas are no good?

rmoehn8 Aug 2019 7:17 UTC

25 points

13 comments1 min readLW link

Calibrating With Cards

lifelonglearner8 Aug 2019 6:44 UTC

32 points

3 comments3 min readLW link