Oldest pages
Showing below up to 329 results in range #1 to #329.
View (previous 500 | next 500) (20 | 50 | 100 | 250 | 500)
- MIRI vs Paul research agenda hypotheses (03:57, 26 April 2020)
- Iterated amplification (03:58, 26 April 2020)
- Hardware-driven vs software-driven progress (03:59, 26 April 2020)
- The Precipice notes (00:23, 27 April 2020)
- Different senses of claims about AGI (22:14, 28 April 2020)
- Discontinuities in usefulness of whole brain emulation technology (09:49, 6 May 2020)
- AI safety field consensus (01:33, 13 May 2020)
- Managing micro-movements in learning (00:07, 16 May 2020)
- Intelligence amplification (01:10, 18 May 2020)
- Mixed messaging regarding independent thinking (20:38, 18 May 2020)
- Timeline of my involvement in AI safety (21:27, 18 May 2020)
- My take on RAISE (21:33, 18 May 2020)
- My current thoughts on the technical AI safety pipeline (outside academia) (01:30, 20 May 2020)
- List of thought experiments in AI safety (07:27, 20 May 2020)
- Content sharing between AIs (07:59, 20 May 2020)
- People are bad (00:31, 21 May 2020)
- Intra-personal comparison test (00:41, 21 May 2020)
- The Hour I First Believed (06:18, 21 May 2020)
- Choosing problems for spaced proof review (23:46, 22 May 2020)
- Highly reliable agent designs (06:21, 27 May 2020)
- Competence gap (00:01, 30 May 2020)
- Lumpiness (06:53, 3 June 2020)
- One wrong number problem (06:54, 3 June 2020)
- Why ain'tcha better at math (08:51, 9 June 2020)
- Missing gear vs secret sauce (21:16, 9 June 2020)
- Personhood API vs therapy axis of interpersonal interactions (22:46, 9 June 2020)
- List of breakthroughs plausibly needed for AGI (07:11, 17 June 2020)
- Architecture (23:18, 23 June 2020)
- Narrow window argument against continuous takeoff (16:56, 24 June 2020)
- Progress in self-improvement (17:15, 24 June 2020)
- Goalpost for usefulness of HRAD work (20:17, 26 June 2020)
- List of success criteria for HRAD work (20:18, 26 June 2020)
- Something like realism about rationality (20:24, 26 June 2020)
- Website to aggregate solutions to textbook exercises (01:19, 30 June 2020)
- Kanzi (21:30, 30 June 2020)
- Missing gear for intelligence (21:42, 30 June 2020)
- Secret sauce for intelligence vs specialization in intelligence (23:01, 6 July 2020)
- Hardware overhang (20:58, 27 July 2020)
- Paperclip maximizer (23:29, 27 July 2020)
- Spoiler test of depth (22:18, 3 August 2020)
- Short-term preferences-on-reflection (23:00, 26 August 2020)
- Comparison of sexually transmitted diseases (19:33, 30 August 2020)
- Future planning (18:53, 7 September 2020)
- Meta-execution (19:02, 23 September 2020)
- My understanding of how IDA works (00:27, 6 October 2020)
- List of teams at OpenAI (05:20, 7 October 2020)
- Second species argument (04:22, 22 October 2020)
- Text to speech software (20:15, 9 November 2020)
- Summary of my beliefs (20:21, 11 November 2020)
- Pascal's mugging and AI safety (22:16, 17 November 2020)
- Popularity symbiosis (23:44, 25 November 2020)
- Carl Shulman (21:52, 28 November 2020)
- Existential win (01:02, 1 December 2020)
- Tao Analysis Solutions (01:39, 1 December 2020)
- Quotability vs ankifiability (21:10, 13 December 2020)
- Using spaced repetition to improve public discourse (02:30, 16 December 2020)
- Kasparov window (22:02, 4 January 2021)
- Analyzing disagreements (22:52, 8 February 2021)
- Resource overhang (03:19, 24 February 2021)
- Paul Christiano (23:00, 25 February 2021)
- HCH (23:03, 25 February 2021)
- Christiano's operationalization of slow takeoff (23:48, 25 February 2021)
- Agent foundations (19:08, 27 February 2021)
- Coherence and goal-directed agency discussion (19:09, 27 February 2021)
- Comparison of terms related to agency (19:09, 27 February 2021)
- Jessica Taylor (19:10, 27 February 2021)
- List of big discussions in AI alignment (19:10, 27 February 2021)
- Minimal AGI vs task AGI (19:10, 27 February 2021)
- Prosaic AI (19:10, 27 February 2021)
- Richard Ngo (19:11, 27 February 2021)
- Simple core of consequentialist reasoning (19:11, 27 February 2021)
- The Uncertain Future (19:12, 27 February 2021)
- Test (19:15, 27 February 2021)
- Whole brain emulation (19:16, 27 February 2021)
- Rapid capability gain vs AGI progress (19:17, 27 February 2021)
- Selection effect for who builds AGI (19:18, 27 February 2021)
- Deconfusion (19:18, 27 February 2021)
- Continuous takeoff (22:16, 1 March 2021)
- Hyperbolic growth (00:07, 2 March 2021)
- Soft-hard takeoff (01:43, 2 March 2021)
- Comparison of AI takeoff scenarios (00:50, 5 March 2021)
- AI takeoff (01:01, 5 March 2021)
- Importance of knowing about AI takeoff (02:13, 5 March 2021)
- Scaling hypothesis (00:45, 12 March 2021)
- Asymmetric institution (21:51, 12 March 2021)
- Counterfactual of dropping a seed AI into a world without other capable AI (20:51, 15 March 2021)
- Main Page (21:22, 19 March 2021)
- OpenAI (19:54, 22 March 2021)
- One-sentence summary card (21:01, 23 March 2021)
- Central node trick for remembering equivalent properties (21:04, 23 March 2021)
- Steam game buying algorithm (23:07, 25 March 2021)
- List of timelines for futuristic technologies (01:01, 26 March 2021)
- List of terms used to describe the intelligence of an agent (20:56, 26 March 2021)
- Stupid questions (20:58, 26 March 2021)
- Spaced proof review as a way to understand key insights in a proof (23:54, 26 March 2021)
- Different mental representations of mathematical objects is a blocker for an exploratory medium of math (02:27, 28 March 2021)
- AI safety is harder than most things (02:28, 28 March 2021)
- AI safety is not a community (02:28, 28 March 2021)
- AI safety lacks a space to ask stupid or ballsy questions (02:28, 28 March 2021)
- AI safety technical pipeline does not teach how to start having novel thoughts (02:28, 28 March 2021)
- Add the complete proof on proof cards to reduce friction when reviewing (02:29, 28 March 2021)
- Corrigibility may be undesirable (02:30, 28 March 2021)
- Debates shift bystanders' beliefs (02:30, 28 March 2021)
- Depictions of learning in The Blue Lagoon are awful (02:30, 28 March 2021)
- Discursive texts are difficult to ankify (02:31, 28 March 2021)
- Flag things to fix during review (02:32, 28 March 2021)
- Giving advice in response to generic questions is difficult but important (02:32, 28 March 2021)
- How doomed are ML safety approaches? (02:32, 28 March 2021)
- How meta should AI safety be? (02:33, 28 March 2021)
- Ignore Anki add-ons to focus on fundamentals (02:33, 28 March 2021)
- Is AI safety no longer a scenius? (02:34, 28 March 2021)
- It is difficult to find people to bounce ideas off of (02:34, 28 March 2021)
- It is difficult to get feedback on published work (02:34, 28 March 2021)
- Make Anki cards based on feedback you receive (02:34, 28 March 2021)
- Mass shift to technical AI safety research is suspicious (02:35, 28 March 2021)
- Newcomers in AI safety are silent about their struggles (02:35, 28 March 2021)
- Nobody understands what makes people snap into AI safety (02:35, 28 March 2021)
- Ongoing friendship and collaboration is important (02:35, 28 March 2021)
- Online question-answering services are unreliable (02:36, 28 March 2021)
- Spaced repetition prevents unrecalled unrecallables (02:37, 28 March 2021)
- Stream of low effort questions helps with popularity (02:38, 28 March 2021)
- There is pressure to rush into a technical agenda (02:38, 28 March 2021)
- Unreliability of online question-answering services makes it emotionally taxing to write up questions (02:39, 28 March 2021)
- Use paper during spaced repetition reviews (02:39, 28 March 2021)
- Use temporary separate Anki decks to learn new cards based on priority (02:39, 28 March 2021)
- Will it be possible for humans to detect an existential win? (02:40, 28 March 2021)
- Will there be significant changes to the world prior to some critical AI capability threshold being reached? (02:40, 28 March 2021)
- Existing implementations of card sharing have nontrivial overhead (17:21, 29 March 2021)
- Value learning (04:53, 30 March 2021)
- Combinatorial explosion in math (20:31, 30 March 2021)
- List of technical AI alignment agendas (21:29, 2 April 2021)
- Simple core (21:32, 2 April 2021)
- Late singularity (21:33, 4 April 2021)
- AI timelines (01:36, 5 April 2021)
- Laplace's rule of succession argument for AI timelines (02:04, 5 April 2021)
- Statistical analysis of expert timelines argument for AI timelines (05:10, 9 April 2021)
- Convergent evolution of values (17:50, 9 April 2021)
- Interacting with copies of myself (20:40, 12 April 2021)
- Selection effect for successful formalizations (20:41, 12 April 2021)
- Setting up Windows (06:34, 20 April 2021)
- The Secret of Psalm 46 outline (21:01, 23 April 2021)
- If you want to succeed in the video games industry (08:37, 29 April 2021)
- Spaced repetition world (03:07, 3 May 2021)
- Potpourri hypothesis for math education (21:59, 6 May 2021)
- Switching costs of various kinds of software (20:37, 8 May 2021)
- Anki (20:00, 10 May 2021)
- Spaced repetition as generator of questions (20:02, 10 May 2021)
- Add easy problems as cards with large graduating interval (19:33, 11 May 2021)
- Using Anki for math (19:34, 11 May 2021)
- Cognitive biases that are opposites of each other (01:14, 12 May 2021)
- SuperMemo (23:48, 12 May 2021)
- Are due counts harmful? (06:31, 13 May 2021)
- Encoding dependence problem (20:02, 14 May 2021)
- Iteration cadence for spaced repetition experiments (07:49, 19 May 2021)
- Spaced repetition allows graceful deprecation of experiments (07:51, 19 May 2021)
- Secret sauce for intelligence (23:45, 19 May 2021)
- Evolution (23:46, 19 May 2021)
- Incremental reading (19:51, 22 May 2021)
- Explosive aftermath (23:22, 25 May 2021)
- Sudden emergence (23:23, 25 May 2021)
- List of critiques of iterated amplification (19:55, 31 May 2021)
- Science argument (07:09, 15 June 2021)
- Textbook test for AI theory (23:32, 18 June 2021)
- Robin Hanson (23:33, 18 June 2021)
- My beginner incremental reading questions (01:15, 21 June 2021)
- AI will solve everything argument against AI safety (20:02, 23 June 2021)
- Politicization of AI (22:06, 11 July 2021)
- Reference class forecasting on human achievements argument for AI timelines (00:14, 12 July 2021)
- Fractional progress argument for AI timelines (00:19, 12 July 2021)
- Hardware argument for AI timelines (00:24, 12 July 2021)
- SuperMemo shortcuts (01:45, 12 July 2021)
- Word explanations are already great (00:49, 16 July 2021)
- We still don't know how to systematically write great word explanations (00:53, 16 July 2021)
- Distillation is not enough (01:27, 16 July 2021)
- List of experiments with Anki (01:46, 16 July 2021)
- Medium that reveals flaws (04:09, 16 July 2021)
- Exhaustive quizzing allows impatient learners to skip the reading (04:20, 16 July 2021)
- Learning-complete (05:02, 16 July 2021)
- Unbounded working memory assumption in explanations (00:40, 17 July 2021)
- List of techniques for managing working memory in explanations (00:40, 17 July 2021)
- Tinkering in math requires loading the situation into working memory (00:40, 17 July 2021)
- Video games comparison to math (00:42, 17 July 2021)
- Braid for math (00:43, 17 July 2021)
- Live math video (00:46, 17 July 2021)
- Finiteness assumption in explorable media (00:46, 17 July 2021)
- Representing impossibilities (00:51, 17 July 2021)
- Probability and statistics as fields with an exploratory medium (00:51, 17 July 2021)
- The Witness (00:51, 17 July 2021)
- Jonathan Blow (00:51, 17 July 2021)
- Improvement curve for good people (00:51, 17 July 2021)
- People watching (00:52, 17 July 2021)
- Inter-personal comparison test (00:52, 17 July 2021)
- Spaced repetition (00:52, 17 July 2021)
- Expert response heuristic for prompt writing (00:52, 17 July 2021)
- List of techniques for making small cards (00:53, 17 July 2021)
- Self-graded prompts made for others must provide guidance for grading (00:54, 17 July 2021)
- Card sharing (00:54, 17 July 2021)
- Booster card (00:57, 17 July 2021)
- Optimal unlocking mechanism for booster cards is unclear (00:57, 17 July 2021)
- Should booster cards be marked as new? (00:57, 17 July 2021)
- Spaced repetition is useful because most knowledge is sparsely applicable (00:59, 17 July 2021)
- What would a vow of silence look like for math? (01:00, 17 July 2021)
- Vow of silence (01:01, 17 July 2021)
- Jelly no Puzzle (01:01, 17 July 2021)
- Video games allow immediate exploration (01:02, 17 July 2021)
- The mathematics community has no clear standards for what a mathematician should know (01:03, 17 July 2021)
- Spaced proof review is not about memorizing proofs (01:04, 17 July 2021)
- Linked list proof card (01:06, 17 July 2021)
- Spaced proof review routine (01:08, 17 July 2021)
- Deck options for proof cards (01:09, 17 July 2021)
- Anki deck options (01:09, 17 July 2021)
- Incremental reading in Anki (01:10, 17 July 2021)
- Large graduating interval as substitute for putting effort into making atomic cards (01:10, 17 July 2021)
- Large graduating interval as a way to prevent pattern-matching (01:10, 17 July 2021)
- Spaced repetition as soft alarm clock (01:11, 17 July 2021)
- Spaced writing inbox (01:13, 17 July 2021)
- Spaced inbox ideas (01:17, 17 July 2021)
- Dual ratings for spaced inbox (01:17, 17 July 2021)
- Spaced everything (01:18, 17 July 2021)
- Snoozing epicycle (18:19, 18 July 2021)
- Busy life periods and spaced inbox (18:19, 18 July 2021)
- Mapping mental motions to parts of a spaced repetition algorithm (18:19, 18 July 2021)
- Soren Bjornstad (18:21, 18 July 2021)
- Piotr Wozniak (18:21, 18 July 2021)
- Emotional difficulties of spaced repetition (18:21, 18 July 2021)
- Feeling like a perpetual student in a subject due to spaced repetition (18:22, 18 July 2021)
- Spaced repetition constantly reminds one of inadequacies (18:22, 18 July 2021)
- Duolingo does repetition at the lesson level (18:22, 18 July 2021)
- Repetition granularity (18:22, 18 July 2021)
- Small card (18:22, 18 July 2021)
- Do an empty review of proof cards immediately after adding to prevent backlog (18:23, 18 July 2021)
- Empty review (18:23, 18 July 2021)
- Maintaining habits is hard, and spaced repetition is a habit (18:25, 18 July 2021)
- Emotional difficulties of AI safety research (18:25, 18 July 2021)
- The Sequences vs evergreen notes (18:25, 18 July 2021)
- Uninsightful articles can seem insightful due to unintentional spaced repetition (18:26, 18 July 2021)
- Open-ended questions are common in real life (18:28, 18 July 2021)
- Integration card (18:28, 18 July 2021)
- Tutoring heuristic for prompt writing (18:29, 18 July 2021)
- Proof card (18:29, 18 July 2021)
- Physical vs digital clutter (18:29, 18 July 2021)
- Spaced repetition and cleaning one's room (18:30, 18 July 2021)
- There is room for something like RAISE (18:34, 18 July 2021)
- List of people who have thought a lot about spaced repetition (18:35, 18 July 2021)
- Hnous927 (18:35, 18 July 2021)
- Big card (01:39, 23 July 2021)
- Spaced proof review as a way to invent novel proofs (01:44, 23 July 2021)
- Big cards can be good for mathematical discovery (02:44, 23 July 2021)
- Page template (20:15, 24 July 2021)
- Desiderata for dissolving the question (21:06, 30 July 2021)
- Spaced proof review (00:01, 2 August 2021)
- List of men by number of sons, daughters, and wives (17:21, 2 August 2021)
- Anki reviews are more fun on mobile (18:43, 11 August 2021)
- Single-architecture generality (20:18, 11 August 2021)
- Single-model generality (20:19, 11 August 2021)
- AlphaGo (20:19, 11 August 2021)
- Bury effortful cards to speed up review (20:22, 11 August 2021)
- Wei Dai (20:47, 11 August 2021)
- Tao Analysis I exercise count (21:59, 11 August 2021)
- Creative forgetting (00:15, 12 August 2021)
- Duolingo (23:48, 13 August 2021)
- Definitions last (21:20, 15 August 2021)
- Fake motivation (21:28, 15 August 2021)
- You don't need to eat your own dogfood in explanation science (00:24, 20 August 2021)
- Changing selection pressures argument (04:17, 30 August 2021)
- Task-dependent diversity (20:52, 15 September 2021)
- List of reasons something isn't popular or successful (08:03, 24 September 2021)
- Fractally misfit (08:04, 24 September 2021)
- List of interesting search engines (05:58, 27 September 2021)
- Tao Analysis Flashcards (20:00, 29 September 2021)
- Asynchronous support (19:28, 3 October 2021)
- The Secret of Psalm 46 (05:38, 6 October 2021)
- Can the behavior of approval-direction be undefined or random? (17:24, 15 October 2021)
- Credit card research 2021 (21:11, 23 October 2021)
- Feynman technique fails when existing explanations are bad (20:27, 24 October 2021)
- Deliberate practice for learning proof-based math (18:37, 25 October 2021)
- Narrow vs broad cognitive augmentation (19:56, 5 November 2021)
- Tricky examples in basic probability (20:05, 5 November 2021)
- Unintended consequences of AI safety advocacy argument against AI safety (21:23, 6 November 2021)
- Duolingo for math (04:11, 8 November 2021)
- Thinking Mathematics (04:13, 8 November 2021)
- Corrigibility (23:29, 8 November 2021)
- Number of relevant actors around the time of creation of AGI (23:37, 8 November 2021)
- Weird recursion (23:38, 8 November 2021)
- Braid (04:06, 9 November 2021)
- Human safety problem (20:36, 11 November 2021)
- List of AI safety projects I could work on (14:25, 4 February 2022)
- Instruction manuals vs giving the answers (11:21, 6 February 2022)
- 3Blue1Brown (12:43, 7 February 2022)
- Scenius (12:44, 7 February 2022)
- Explanation science (11:06, 8 February 2022)
- Philosophical difficulty (10:59, 26 February 2022)
- List of disagreements in AI safety (11:00, 26 February 2022)
- Difficulty of AI alignment (11:10, 26 February 2022)
- Aligning smart AI using slightly less smart AI (11:21, 26 February 2022)
- Minimal AGI (14:42, 26 February 2022)
- Pivotal act (15:15, 26 February 2022)
- Doomer argument against AI safety (20:55, 18 March 2022)
- Application of functional updateless timeless decision theory to everyday life (19:53, 1 April 2022)
- How similar are human brains to chimpanzee brains? (22:25, 12 April 2022)
- Discovery fiction (07:05, 23 June 2022)
- Explorable explanation (19:00, 3 July 2022)
- Comparison of pedagogical scenes (19:28, 4 July 2022)
- List of arguments against working on AI safety (00:29, 24 July 2022)
- Late 2021 MIRI conversations (21:27, 1 August 2022)
- AlphaGo as evidence of discontinuous takeoff (23:35, 2 August 2022)
- AI prepping (04:20, 26 November 2022)
- UDASSA (09:13, 8 January 2023)
- What counts as good motivation? (19:36, 20 January 2023)
- Interaction reversal between knowledge-to-be-memorized and ideas-to-be-developed (23:04, 3 February 2023)
- Finding the right primitives for spaced repetition responses (23:23, 3 February 2023)
- Spaced repetition response as chat message or chat reaction (23:26, 3 February 2023)
- Spaced inbox review should not be completionist or obligatory (08:53, 14 February 2023)
- What makes a word explanation good? (20:46, 9 May 2023)
- Anki deck philosophy (18:19, 8 June 2023)
- Cards created by oneself can be scheduled more aggressively (22:45, 16 July 2023)
- Dealing with bad problems in spaced proof review (18:26, 2 August 2023)
- Continually make new cards (18:26, 2 August 2023)
- Make new cards when you get stuck (18:27, 2 August 2023)
- Andy Matuschak (18:28, 2 August 2023)
- Michael Nielsen (18:28, 2 August 2023)
- Tips for reviving a spaced repetition practice (20:01, 2 August 2023)
- Can spaced repetition interfere with internal sense of relevance? (01:54, 16 August 2023)
- Add all permutations of a card to prevent pattern-matching (07:26, 18 August 2023)
- Equivalence classes of prompts (00:29, 19 August 2023)
- Card sharing allows less valuable cards to be created (00:29, 19 August 2023)
- Reverse side card for everything (00:41, 19 August 2023)
- Spaced repetition is not about memorization (01:24, 21 August 2023)
- Deck options for small cards (20:27, 25 November 2023)