Mason Wang
18 followers · 3 following · 425 views
on the atlas — 35
- The Ultimate Guide for Becoming an Idea Machine - James Altucher2 savers
- 2025 letter | Dan Wang62 savers
- thalassophilia - by hannah - a lot about nothing3 savers
- The Best Essay19 savers
- Marley Xiong24 savers
- Asymmetry of verification and verifier’s law — Jason Wei15 savers
- A lighthouse of competence in a sea of mediocrity1 savers
- Joan Didion, The Art of Fiction39 savers
- Little Things4 savers
- How Attention Got So Efficient [GQA/MLA/DSA] - YouTube2 savers
- 21 observations from people watching - by shani10 savers
- Helen Mort: Keeping a Notebook – Young Poets Network1 savers
- Feynman’s Rigor « the jsomers.net blog3 savers
- Dan Hollick (@DanHollick)1 savers
- Some thoughts on writing6 savers
- Don't Make a Blog, Make a Brain Dump1 savers
- A parenthetical departure1 savers
- after my dad died, we found the love letters | Jenneral HQ8 savers
- We Induced Smells With Ultrasound11 savers
- home for the holidays - by rayne fisher-quann11 savers
- advice - nabeelqu43 savers
- Variational autoencoders.1 savers
- Growing Neural Cellular Automata10 savers
- Every productivity thought I've ever had, as concisely as possible - Alexey Guzey29 savers
- The Average Fourth Grader Is a Better Poet… | The Poetry Foundation2 savers
- Advice for a friend who wants to start a blog7 savers
- Andrej Karpathy — AGI is still a decade away7 savers
- Marble: A Multimodal World Model | World Labs2 savers
- Expertise makes people weirder, not more alike1 savers
- How To Ask For An Introduction - by Elad Gil - Elad Blog3 savers
- Curius / Onboarding2621 savers
- Defeating Nondeterminism in LLM Inference - Thinking Machines Lab40 savers
- Borges-The-Library-of-Babel.pdf8 savers
- Life lessons from reinforcement learning — Jason Wei7 savers
- PyTorch internals : ezyang’s blog7 savers
highlights — 31
I often think of the case of Xiaomi. In 2021, Lei Jun vowed that the company he founded would break into the EV business. Four years later, Xiaomi started shipping cars to customers. Not only that, a Xiaomi EV set a speed record at the Nürburgring racetrack in Germany. Compare Xiaomi to Apple, which spent 10 years and $10 billion studying whether to enter the EV market before it pulled the plug. The world’s most advanced consumer product company could not match Xiaomi’s feat. It’s cases like these that make me skeptical of reasoning about China’s tech successes through financial measures or pr…
2025 letter | Dan WangIt’s hard to know what to make of this document, because the authors keep tucking important context into footnotes, repeatedly saying they do not endorse a prediction. Six months after publication, they stated that their timelines were lengthening, but even at the start their median forecast for the arrival of superintelligence was later than 2027. Why they put that year in their title remains beyond me.
2025 letter | Dan WangBeijing has been preparing for Cold War without eagerness for waging it, while the US wants to wage a Cold War without preparing for it.
2025 letter | Dan WangBeijing has been preparing for Cold War without eagerness for waging it, while the US wants to wage a Cold War without preparing for it.
2025 letter | Dan WangOne thing I like in an initial question is outrageousness. I love questions that seem naughty in some way — for example, by seeming counterintuitive or overambitious or heterodox. Ideally all three.
The Best EssayWriting converts your ideas from vague to bad. But that's a step forward, because once you can see the brokenness, you can fix it.
The Best EssayIndeed, my belief is that any solvable problem that fits those five properties will be solved in the next few years.
Asymmetry of verification and verifier’s law — Jason WeiVerifier’s rule: The ease of training AI to solve a task is proportional to how verifiable the task is. All tasks that are possible to solve and easy to verify will be solved by AI.
Asymmetry of verification and verifier’s law — Jason Weihad a habit of studying how other people found success and trying to imitate them. Sometimes it worked, but eventually I realized that I would never surpass the full ability of someone else because they were playing to their strengths which I didn’t have
Life lessons from reinforcement learning — Jason Weionce you can take reasonable trajectories, we generally avoid imitation learning
Life lessons from reinforcement learning — Jason WeiThere isn’t much more to it—you just do kind things, and wait to earn their trust.
programming is terribleIf you don’t have fun sucking at it, you’ll never invest the time to be good at it.
programming is terribleA certain amount of resistance is good for anybody. It keeps you awake.
Joan Didion, The Art of FictionI dealt with it the same way I deal with everything. I just tended my own garden, didn't pay much attention, behaved—I suppose—deviously. I mean I didn't actually let too many people know what I was doing.
Joan Didion, The Art of FictionYou have said that writing is a hostile act; I have always wanted to ask you why. JOAN DIDION It's hostile in that you're trying to make somebody see something the way you see it, trying to impose your idea, your picture. It's hostile to try to wrench around someone else's mind that way. Quite often you want to tell somebody your dream, your nightmare. Well, nobody wants to hear about someone else's dream, good or bad; nobody wants to walk around with it. The writer is always tricking the reader into listening to the dream.
Joan Didion, The Art of FictionThe reason I posted it was because I was worried people would judge me when I made mediocre posts. This is pretty cringe.
A lighthouse of competence in a sea of mediocrityBut this standard is the only one that actually matters. The things you'd die for are the things you live for. Trying, not to enter the top 3.956%, or even 0.01%, but the literal peak of human achievement, with no one above you, is almost beyond cool. Not only for your own ego. But for the good of the human race.
A lighthouse of competence in a sea of mediocrityMany of the driven people I have met are closed fists: warm, charming, and ready to punch through a wall at any point in time. There is a rigidity, a tunnel vision that naturally follows grasping for certain outcomes.
21 observations from people watching - by shaniI never met a person who looked down on others and unconditionally accepted themselves.
21 observations from people watching - by shaniIt is easy to tell how happy someone is to see another person enter a conversation.
21 observations from people watching - by shaniHappy has a boundless quality: unpredictable, even when it is at a low level.
21 observations from people watching - by shanifull of irreplaceable (albeit often uninteresting) ideas
Helen Mort: Keeping a Notebook – Young Poets NetworkBut a better recommendation might be to start hacking.
Feynman’s Rigor « the jsomers.net blogI see a lot of essays framed as writing advice which are actually thinly veiled descriptions of how someone writes that basically say "you should write how I write"
Some thoughts on writing7. Write regularly, and learn to ‘think in writing’. This is true for literally everyone, regardless of whether you want to be a writer or not, whether you want to publish or not. Just have a Google Doc in which you add a page a day of whatever’s on your mind. This has a million benefits, but a simple one is just clearing your cache: if you don’t do this, your brain sort of gets clogged by all the things you have on your mind, whereas if you ‘empty’ your brain onto a page that creates room for new thoughts. If you really want to be a clear thinker, you need to learn to ‘think in writing’. I li…
advice - nabeelqu99% of the time procrastination is not about you being lazy or lacking work-ethic. Your body/brain is sending your some important information about the tasks at hand, and it’s important that you listen to those signals empathetically.
Every productivity thought I've ever had, as concisely as possible - Alexey GuzeyBy middle school/high school, the average student has learned how normal people talk. The resulting language is underwhelming and predictable—the safe regurgitations of a thoroughly socialized consciousness.
The Average Fourth Grader Is a Better Poet… | The Poetry FoundationIt is what is odd about you that is interesting.
Advice for a friend who wants to start a blogYou always have to seek entropy in your life. Talking to other people is a great source of entropy, and things like that. So maybe the brain has also built some internal mechanisms for increasing the amount of entropy in that process. That’s an interesting idea.
Andrej Karpathy — AGI is still a decade awayI also think humans collapse over time. These analogies are surprisingly good. Humans collapse during the course of their lives. This is why children, they haven’t overfit yet. They will say stuff that will shock you because you can see where they’re coming from, but it’s just not the thing people say, because they’re not yet collapsed. But we’re collapsed. We end up revisiting the same thoughts. We end up saying more and more of the same stuff, and the learning rates go down, and the collapse continues to get worse, and then everything deteriorates.
Andrej Karpathy — AGI is still a decade awayTaking advice is a fraught activity. As you become more skilled in a field, the advice you need to advance becomes more specific to you, and as such harder to find. At some point, depending on the field, it somehow becomes unwise to take general advice at all, since taking advice will at best do nothing, or at worst, be a reversion to the mean.
Expertise makes people weirder, not more alike