Pages with the most categories
Showing below up to 88 results in range #1 to #88.
View (previous 500 | next 500) (20 | 50 | 100 | 250 | 500)
- Distillation is not enough (3 categories)
- Ignore Anki add-ons to focus on fundamentals (3 categories)
- There is room for something like RAISE (3 categories)
- Debates shift bystanders' beliefs (3 categories)
- List of teams at OpenAI (2 categories)
- Will it be possible for humans to detect an existential win? (2 categories)
- There is pressure to rush into a technical agenda (2 categories)
- Switching costs of various kinds of software (2 categories)
- Mass shift to technical AI safety research is suspicious (2 categories)
- Spaced repetition prevents unrecalled unrecallables (2 categories)
- Anki deck philosophy (2 categories)
- Flag things to fix during review (2 categories)
- Exhaustive quizzing allows impatient learners to skip the reading (2 categories)
- AI safety lacks a space to ask stupid or ballsy questions (2 categories)
- Duolingo does repetition at the lesson level (2 categories)
- Online question-answering services are unreliable (2 categories)
- Stream of low effort questions helps with popularity (2 categories)
- List of critiques of iterated amplification (2 categories)
- Make Anki cards based on feedback you receive (2 categories)
- Feynman technique fails when existing explanations are bad (2 categories)
- Iterated amplification (2 categories)
- Unreliability of online question-answering services makes it emotionally taxing to write up questions (2 categories)
- The mathematics community has no clear standards for what a mathematician should know (2 categories)
- Deck options for small cards (2 categories)
- Make new cards when you get stuck (2 categories)
- My take on RAISE (2 categories)
- Maintaining habits is hard, and spaced repetition is a habit (2 categories)
- My understanding of how IDA works (2 categories)
- Will there be significant changes to the world prior to some critical AI capability threshold being reached? (2 categories)
- How doomed are ML safety approaches? (2 categories)
- How meta should AI safety be? (2 categories)
- Uninsightful articles can seem insightful due to unintentional spaced repetition (2 categories)
- Are due counts harmful? (2 categories)
- Use paper during spaced repetition reviews (2 categories)
- Reference class forecasting on human achievements argument for AI timelines (2 categories)
- AI safety is harder than most things (2 categories)
- Open-ended questions are common in real life (2 categories)
- Video games allow immediate exploration (2 categories)
- Tinkering in math requires loading the situation into working memory (2 categories)
- Use temporary separate Anki decks to learn new cards based on priority (2 categories)
- Fractional progress argument for AI timelines (2 categories)
- Nobody understands what makes people snap into AI safety (2 categories)
- Self-graded prompts made for others must provide guidance for grading (2 categories)
- Scenius (2 categories)
- Optimal unlocking mechanism for booster cards is unclear (2 categories)
- Different mental representations of mathematical objects is a blocker for an exploratory medium of math (2 categories)
- Continually make new cards (2 categories)
- Laplace's rule of succession argument for AI timelines (2 categories)
- Giving advice in response to generic questions is difficult but important (2 categories)
- Spaced proof review is not about memorizing proofs (2 categories)
- Is AI safety no longer a scenius? (2 categories)
- Cards created by oneself can be scheduled more aggressively (2 categories)
- Anki reviews are more fun on mobile (2 categories)
- Tips for reviving a spaced repetition practice (2 categories)
- Hardware argument for AI timelines (2 categories)
- HCH (2 categories)
- Newcomers in AI safety are silent about their struggles (2 categories)
- Spaced repetition allows graceful deprecation of experiments (2 categories)
- Using spaced repetition to improve public discourse (2 categories)
- SuperMemo shortcuts (2 categories)
- Existing implementations of card sharing have nontrivial overhead (2 categories)
- OpenAI (2 categories)
- Corrigibility may be undesirable (2 categories)
- AI safety technical pipeline does not teach how to start having novel thoughts (2 categories)
- Card sharing allows less valuable cards to be created (2 categories)
- Statistical analysis of expert timelines argument for AI timelines (2 categories)
- It is difficult to find people to bounce ideas off of (2 categories)
- Spaced repetition constantly reminds one of inadequacies (2 categories)
- Can spaced repetition interfere with internal sense of relevance? (2 categories)
- List of experiments with Anki (2 categories)
- Add all permutations of a card to prevent pattern-matching (2 categories)
- What would a vow of silence look like for math? (2 categories)
- AI safety is not a community (2 categories)
- Depictions of learning in The Blue Lagoon are awful (2 categories)
- Tao Analysis Solutions (2 categories)
- It is difficult to get feedback on published work (2 categories)
- Spaced repetition is not about memorization (2 categories)
- Should booster cards be marked as new? (2 categories)
- Linked list proof card (2 categories)
- Add the complete proof on proof cards to reduce friction when reviewing (2 categories)
- Discursive texts are difficult to ankify (2 categories)
- If you want to succeed in the video games industry (2 categories)
- Popularity symbiosis (2 categories)
- Ongoing friendship and collaboration is important (2 categories)
- Spaced repetition is useful because most knowledge is sparsely applicable (2 categories)
- Meta-execution (2 categories)
- Incremental reading in Anki (2 categories)
- Do an empty review of proof cards immediately after adding to prevent backlog (2 categories)
