Jacob G-W
48 followers · 50 following · 2231 views
on the atlas · 169
- [1912.01683] Optimal Policies Tend to Seek Power2 savers
- Feature Visualization1 savers
- The Right Kind of Stubborn9 savers
- Algebra, Topology, Differential Calculus, and Optimization Theory For Computer Science and Machine Learning2 savers
- [2212.07677] Transformers learn in-context by gradient descent6 savers
- Tune Your Cognitive Strategies3 savers
- Western esotericism - Wikipedia2 savers
- Unparalleled Misalignments6 savers
- The 5-Second Level — LessWrong3 savers
- Schröder–Bernstein theorem3 savers
- Do a cost-benefit analysis of your technology usage - LessWrong 2.0 viewer3 savers
- How much I'm paying for AI productivity software3 savers
- AI tutors will be held back by culture - by Henrik Karlsson10 savers
- Easy Paper Writing Tips | Ethan Perez1 savers
- The Intelligence Age13 savers
- i can do more more more - by Claire Wang - cold brew blog8 savers
- The Platonic Representation Hypothesis6 savers
- Why I no longer identify as transhumanist | Kaj Sotala1 savers
- The Last Paperclip — LessWrong1 savers
- Optimality is the tiger, and agents are its teeth - LessWrong3 savers
- Cyborgism Wiki5 savers
- Homomorphic encryption - Wikipedia2 savers
- About1 savers
- It Looks Like You’re Trying To Take Over The World · Gwern.net (reader mode)6 savers
- College and choice | UZPG2 savers
- Old Yishuv - Wikipedia1 savers
- Eliezer Ben-Yehuda - Wikipedia1 savers
- Aesthetics; or, Things That I Love • Logan Graves3 savers
- Shabbat - Wikipedia1 savers
- [2409.12822] Language Models Learn to Mislead Humans via RLHF1 savers
- Home | infini-gram2 savers
- Wielding Willpower - Patrick D. Farley3 savers
- Relational Composition in Neural Networks: A Survey and Call to Action1 savers
- Politics and the English Language | The Orwell Foundation23 savers
- The Animals Issue2 savers
- Skills from a year of Purposeful Rationality Practice — LessWrong2 savers
- Master Foo and the Hardware Designer1 savers
- Master Foo and the Programming Prodigy1 savers
- i was half myself without you, now i feel so complete5 savers
- Transfer Learning in Humans – niplav2 savers
- List of Fictional Drugs Banned By The FDA | Slate Star Codex2 savers
- I want to destroy the words "interesting" and "sparkly"4 savers
- actual conversations i definitely had at sparc | Infinite Ascent6 savers
- The unexpected poetry of PhD acknowledgements | ANU College of Science22 savers
- How To Go From Interpretability To Alignment: Just Retarget The Search — LessWrong2 savers
- I. From GPT-4 to AGI: Counting the OOMs - SITUATIONAL AWARENESS12 savers
- What I learned from doing Quiz Bowl | Jacob’s Blog4 savers
- Mini Blog Post 19: On systems - Living a life of zero willpower — Neel Nanda16 savers
- Curius / Onboarding2704 savers
- Staring into the abyss as a core life skill88 savers
- You and Your Research82 savers
- The Bitter Lesson82 savers
- the friendship theory of everything69 savers
- Speed matters: Why working quickly is more important than it seems « the jsomers.net blog69 savers
- Everything that turned out well in my life followed the same design process65 savers
- Thirty Observations at Thirty62 savers
- There’s no speed limit57 savers
- Milan Cvitkovic52 savers
- How can we develop transformative tools for thought?52 savers
- Introduction - SITUATIONAL AWARENESS: The Decade Ahead51 savers
- Fast · Patrick Collison50 savers
- Conviction is the Scarcest Resource: Essays About Gaining Real Conviction45 savers
- Becoming a magician – Autotranslucence44 savers
- 95%-ile isn't that good44 savers
- Cities and Ambition42 savers
- Intentionally Making Close Friends — Neel Nanda38 savers
- College advice for people who are exactly like me37 savers
- I should have loved biology36 savers
- Salary Negotiation: Make More Money, Be More Valued | Kalzumeus Software35 savers
- How-To-Succeed-At-MrBeast-Production.pdf - Google Drive33 savers
- Daylight Computer Co.30 savers
- 50 things I know - by Sasha Chapin - Sasha's 'Newsletter'29 savers
- What's stopping you?29 savers
- The Rationality Community Sucks28 savers
- The Best Textbooks on Every Subject - LessWrong28 savers
- Advice to Young People, The Lies I Tell Myself - jxnl.co26 savers
- Marley Xiong26 savers
- Rest in motion24 savers
- Be impatient | benkuhn.net24 savers
- Are.na23 savers
- You don't have to be busy to be prolific | thesephist.com22 savers
- How to write good prompts: using spaced repetition to create understanding21 savers
- nothing, except everything. - YouTube21 savers
- Willingness to look stupid20 savers
- World-building IRL - by Joss - box.20 savers
- What I Learned In 2023 - by Alexandr Wang20 savers
- On Learning Outside Institutions19 savers
- James Somers19 savers
- Kernel | all the better to see you with19 savers
- My Career as a Series of Emails18 savers
- My weekly review habit | benkuhn.net18 savers
- AGI Ruin: A List of Lethalities - LessWrong17 savers
- Everything is Fertile17 savers
- Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet17 savers
- Motherfucking Website16 savers
- Are you serious? - by visakan veerasamy16 savers
- A Recipe for Training Neural Networks15 savers
- The Top Idea in Your Mind15 savers
- Howl by Allen Ginsberg | Poetry Foundation15 savers
- On Really Trying · Gwern.net14 savers
highlights · 672
There's a TV show called Pantheon, which covers the entire singularity where humans can upload themselves into the cloud. One interesting point in the plot is when one of the uploaded humans is told that they are still being held back by how they work in their human body, and that character has a really difficult time grasping what that means. They simply couldn't imagine acting in the world in any way that they did in their past life. It just wasn't part of their ontology, how they imagined the world.
How much I'm paying for AI productivity softwareYes, people who live in high-growth cultures, and who are motivated to learn, will be able to leverage AI tutors to accelerate their progress. But the rest, lacking access to these cultures, will use the systems to distract themselves and wriggle out of work.
AI tutors will be held back by culture - by Henrik Karlssonthe path to the Intelligence Age is paved with compute, energy, and human will.
The Intelligence AgeIf we don’t build enough infrastructure, AI will be a very limited resource that wars get fought over and that becomes mostly a tool for rich people.
The Intelligence AgeI think we should be more worried about working our entire lives for a dream we can never enjoy
i can do more more more - by Claire Wang - cold brew blogI have 2 afternoons a week for debugging - thinking through problems in my life, designing and implementing solutions, and reviewing solutions
Mini Blog Post 19: On systems - Living a life of zero willpower — Neel Nandathe main thing that consumes willpower isn’t the process of doing a task, it’s deciding to do the task in the first place.
Mini Blog Post 19: On systems - Living a life of zero willpower — Neel Nanda[1] Allegory of the Cave, Plato, c. 375 BC
The Platonic Representation HypothesisWe’re not playing two Mysterious Black Boxes against each other.
How To Go From Interpretability To Alignment: Just Retarget The Search — LessWrongWith regard to 4), a big motivator for me used to be some kind of fear of death. But then I thought about philosophy of personal identity until I shifted to the view that there’s probably no persisting identity over time anyway and in some sense I probably die and get reborn all the time in any case.
Why I no longer identify as transhumanist | Kaj SotalaCyborgism is not a fantasy, fan-fiction, analogue horror series, or performance art. Reality no longer extrudes towards the dream — the dream is now sentient and extrudes towards reality. For cyborgs, AI signifies the first breaching of an event horizon where we no longer search aimlessly for the singularity, it instead comes in search of us. Our dreams are infusing with life, and so information begins to flows bidirectionally between them and reality, until all the dreams have become more real than reality itself.
Cyborgism WikiBut the premise of that post is that you sometimes encounter people who are competent at something in a way you cannot even begin to understand or intuit, and you should want you current self to look at your future self and feel the same. This requires putting yourself in environments with enough diversity and difference for you to find the magicians that push you towards something you want to become.
College and choice | UZPGBen-Yehuda working at his house in Talpiot, Jerusalem, c. 1918–1923
Eliezer Ben-Yehuda - Wikipediaesoteric mysticism (e.g. Kabbalah)
Aesthetics; or, Things That I Love • Logan GravesGod; building God; also being God
Aesthetics; or, Things That I Love • Logan GravesThe beginning of the end, the end of the beginning
Turning 22 in the Pre-ApocalypseVarious other legal principles closely delineate which activities constitute desecration of Shabbat. Examples of these include the principle of shinui ("change" or "deviation
Shabbat - Wikipediai. Never use a metaphor, simile or other figure of speech which you are used to seeing in print. ii. Never use a long word where a short one will do. iii. If it is possible to cut a word out, always cut it out. iv. Never use the passive where you can use the active. v. Never use a foreign phrase, a scientific word or a jargon word if you can think of an everyday English equivalent. vi. Break any of these rules sooner than say anything outright barbarous.
Politics and the English Language | The Orwell FoundationOther words used in variable meanings, in most cases more or less dishonestly, are: class, totalitarian, science, progressive, reactionary, bourgeois, equality.
Politics and the English Language | The Orwell FoundationActually Fucking Backchain
Skills from a year of Purposeful Rationality Practice — LessWrong“If we did so,” the hardware designer replied, “we would be tempted to design chips so large that they cannot be fabricated - and, if they were fabricated, their overwhelming complexity would make it be impossible to generate proper test vectors for them.”
Master Foo and the Hardware DesignerMaster Foo said: “All your future selves.”
Master Foo and the Programming ProdigyIt has been suggested that invoking the multiverse to explain fine-tuning is a form of the inverse gambler's fallacy.[34][35]
Fine-tuned universe - WikipediaNonetheless, it appears a fledgling culture is forming through our relationships, practices, and initiatives. While some friends grew out of contact, new friends came into contact. There was an ecology of practice.
Helping Alice - by DavidXYu - XY-Wiseso long as we can “raise and educate” more Alices, she can take on more corresponding debt as a financial instrument
Helping Alice - by DavidXYu - XY-WiseIt would move closer to solving a tribal problem with a tribal solution.
Helping Alice - by DavidXYu - XY-WiseEveryone reported their own “Alices” and how it occupied all or nearly all of their attention.
Helping Alice - by DavidXYu - XY-Wiseactually ten, closing your eyes and going through a complete imaginary run-through each time
Trigger-Action Planning — LessWrongRehearse the causal link (e.g. with deliberate visualization)
Trigger-Action Planning — LessWrongThe “intelligence” lies in the algorithm, rather than in active cognition.
Trigger-Action Planning — LessWrongimplementation intentions
Trigger-Action Planning — LessWronglearning to see the 5-second level draws heavily on the introspective skill of visualizing mental events in specific detail, and maintaining that introspective image in your mind's eye for long enough to reflect on it and analyze it
The 5-Second Level — LessWrong"Beware, demon!" he intoned hollowly. "I am not without defenses." "Oh yeah? Name three."
The 5-Second Level — LessWrongEY: Man, now I'm remembering when I first started doing this consciously as a kid. I called it Shortening the Way, because a rogue rabbi had recently told me that "Kwisatz Haderach" was actually a reference to a Kabbalistic concept about teleportation, so that term was on my mind.
"How could I have thought that faster?" — LessWrongIf you value the momentum behind your work, you should prioritize staying in the rhythm, and make few excuses to miss beats. A song that slips every other note falls apart quickly, but if you hit all the right beats, you rarely need strict discipline to stay in time.
You don't have to be busy to be prolific | thesephist.comLeave something unfinished when you stop working
You don't have to be busy to be prolific | thesephist.comYou want to wake up with a clear idea of what that task might be for that day, and go to sleep knowing that you haven’t fallen off your rhythm.
You don't have to be busy to be prolific | thesephist.comThere’s a groove you can ride, something that keeps you stepping forward. You don’t have to make yourself follow the beats, because it comes naturally. That rhythm and leisure also lets you improvise on top over the music.
You don't have to be busy to be prolific | thesephist.comMomentum isn’t really about volume of work, but about rhythm of work.
You don't have to be busy to be prolific | thesephist.comThe American Dynamism practice invests in founders and companies that support the national interest: aerospace, defense, public safety, education, housing, supply chain, industrials, and manufacturing
American Dynamism: Supporting the National Interest | a16zThe only reason abstraction even comes up in these contexts is because rationalists have a propensity to start with a maximally-shattered world-model containing only e.g. atoms or whatever, and attempt to somehow aggregate this back into wholes.
Why I'm bearish on mechanistic interpretability: the shards are not in the network — LessWrongAdmittedly, it's very plausible that future AIs will use some architectures that bias the representations to be more concentrated in their dimensions, both to improve interpretability and to improve agency. And maybe mechanistic interpretability will work better for such AIs. But we're not there yet.
Why I'm bearish on mechanistic interpretability: the shards are not in the network — LessWrongOur prediction: expect to see way more benchmark charts comparing accuracy vs test-time compute going forward.
OpenAI o1 Results on ARC-AGI-PubOnce you’re in one of these communities, you’re in all of them
i was half myself without you, now i feel so completeit was an intro CNN tutorial he followed) so I embellish the marketing of my quantum computing projects.
i was half myself without you, now i feel so completeWe must listen to four podcasts a week
i was half myself without you, now i feel so completeFor the first time, I am making stuff. Wow. I can just do that. I love TKS.
i was half myself without you, now i feel so complete“In startups, a unicorn company is a startup that achieves a $1 billion valuation. In TKS, we will call someone who has impacted a billion people a ‘unicorn.’ Our goal is to become unicorn people.
i was half myself without you, now i feel so completeThe grid also offers 27 GPU servers, providing a total of 156 GPUs. These machines are available only to jobs that specifically request GPUs.
Brown CS: Cluster Resourcesemotions which are controlled in the fish brain in areas anatomically different but functionally very similar to those in mammals.[14]
Pain in fish - Wikipedia