Azuremis @azuremis

thinking of some of my bag of tricks if use in my "Feynman Method": - where is the entropy being created -- drive that towards zero. everywhere there is entropy being created, e.g., "energy" diffusing across effective temperature differences as "heat", it could be reversibly converted into work, useful "energy" via a 100% efficient, reversible "Carnot engine." * this is true up to a model that conserves something like "energy for any kind of diffusion and any kind of conserved "energy." even psychological energy, LLM energy. - what is the strongest scaling? T^4? exponential? how can you drive that to saturation. - explore the tradeoffs from a new extremum with the new tools available in technology since people built or rebuilt the industry. what about with the physics that nature has given us. push all the way to the limit, first in theory, and then toward practice. - what is the most severe problem? * is there a way to invert it? * if you relieve this problem, for what and who are you enabling the most. the max-flow min-cut theorem states that in a flow network, the maximum amount of flow passing from the source to the sink is equal to the total weight of the edges in a minimum cut. this is deeply true, not just about graphs in computer science, but about energy, about economies, about negotiating positions in games and business, and in the circulatory systems of life itself in all its forms, in anything that is conserved. it is a deeply spiritual, taoist, physical, visual, and mathematical lens to view the world, between the flow of conversed quantities, and the structure of chokepoints. * in the case of business, for example, min-cut max-flow is related to the maximum power that can be brought into a location. it's now manifesting as time-to-power at scale for hyperscalers. and we are happily serving it. * in the case of Taoist energies, if you notice an excess of Yang, everyone is Yanging, the pivot point that has the maximum power is Yin energy. and vice verse. If everyone is showing off, be mysterious. If everyone is confident, raise a question. If people are lacking direction, pick a direction. If people are being mysterious, shine a light. If volatility is melting down, look for out-of the money optionality. If it looks like the end of the world, bet on the sun rising. If it feels like you can't breathe, remind everyone, including yourself, to breathe. - what are the natural scales in which to solve a problem, and which are the problems that people fail to solve systematically, because people cannot solve them, or shy away from solving them, at adequate scale? - this strikes in both directions * at indie or @lightcellenergy scale, or Wright Brothers scale, where really the limiting reagents are a combination of inventor time spent obsessed with a problem, which doesn't scale well with organization size, and a combination of the necessary physical courage (either flying the thing or doing T > 2000+ C experiments in front of your face * at @elonmusk or @sama scale, where they use the fact that nobody else actually marshals the resources in spaceflight, or AI, to solve the problems at natural scale (in size dimensions, or at vertical scale) like they do. So they have this neat trick that there's decades of good ideas that are free alpha you can pick up, that are good ideas but required adequate scale in the effort to apply, and if you're among the most ambitious person in the area you'll probably want to go work for their efforts, or are at least are sorely tempted, because getting the resources to pursue it is such a severe requirement in such shortage that it's among your best hopes. and nobody else dares compete with you on scale! - think in public. if you're doing something complex enough, it's too challenging to rip you off completely. you're better off drawing from an open door. the scientists at Bell Labs looked at the meta question of what distinguished their best researchers. they found that most predictive factor was that they all had lunch with Henry Nyquist. So, i've always tried to be Henry Nyquist, and what's more, in a positive sum way, harnessed the output by investing in the output of particular fertile luncheons of the mind. There's 1000x more good ideas that a single person can execute on directly. You can share a lot of what you're doing safe in the knowledge that even if you give away as much as you can, the generator of knowledge and progress comes from my own heart, perspective, background, and obsession. it gives you the psychological safety to have the generosity of spirit to have a certain amount of magnanimity. - and though i have not found the karmic calculus of the universe to be exactly same in dimension and particular, i have found that over the long haul, you get what you give, more often than not. and spiritual energy defies the conservation of energy in physical law, inside the vortex of the human heart. you can feel yourself generate positive energy as your heart pumps and you inhale and exchange. you can transform the negative to positive with attention. you can extend that field of awareness around the bubble over your self and extend it as you will to the farthest reaches. the ancients may have called this attention practice meditation, but it is also useful in a totally secular sense, to physicalize awareness into a "context window" a mind palace for your thoughts. - if you extend your context into a well designed workshop, your physical awareness can bootstrap into the memory systems of your mind. this is how i "live in my workshop" as many great craftsmen and inventors of the past have utilized, and how those creators in possession of a workshop or studio are so much more creatively powerful relative to their financial Superiors -- even when financially broke they are rich in tools, space, context window, and time available to obsess and expend on the problem. and they have less to lose, which might help with the physical courage part. this is enhanced if you are "neurodiverse" / ADHD for some stripes. - nature gives you much: every common or uncommon atom, physical laws that act at a massive range of scales multiplied against time and space, far more than our computational analogs. what geometry does physical nature suggest? - biological nature has often solved its problems bottom up using geometry to solve flow problems. all natural systems must flow energy, material input and product output, and waste (among other conservations, like cells and immune systems). it always solves problems by folding new dimensions into the system, either in the physical layout of arteries, veins, cell or cellulose walls stacked folded capillaries, or extended all the way between the folded proteins necessary to structure themselves at low energy per base pair, the physical antenna to couple to the visible light (eV per mole), the chemical energy required to store the energy in ATP at high availability (0.3 eVs per mole), and the low physical energy (micro eVs per mole) required to transport circulation through a living body. by which means has nature discovered to arbitrage entropy through her natural time scale, and energy natural scales? true universal thermal equilibrium is a dead universe, and we are in a live one -- how has nature bridged between dynamic quasi-equilibria, in life and nature's other storms. - how can I apply these ideas given the toolkit human hands over mastered? what new tools and possibilities have I, permutationally, hyperdimensionally expanded at the frontier of the possible, now that i have, with the benefit of human technological civilization, unlocked the machining and patterning of metals, of semiconductors, of ceramic? invented macroscopic rotating machinery -- assembled machines that work at far greater temperatures and in shapes that, biology could not perfuses as a living object, and those could not optimize in her billions of years? in which ways am I, the human investor, have, in some dimensions, more to be grateful for, in the creative tools i have at my disposal, and my few years at a pivotal time in history, than nature has, for her planet full of live and her billions of years of evolution. if i make them count!! - most of the hyperdimensional orange is peel - as an example: they highest temp heat exchangers have been metal because they need to tolerate tension under high temperature -- well, what if you can extract energy at high temperature for ultrahigh carnot efficiency, ultrahigh power density like @lightcellenergy, at minimal pressure over ambient? then you can use an oxide ceramic, which unlike steel won't further oxidize. but now you have something that's much stronger in compression, and you have to design something that will balance and distribute all tensions with compressions. ok, how does nature solve problems like this -- with an extra dimension of curvature, as meta materials and cellular geometry. - most of a hyperdimensional market arbitrage, which could be literally any trade between any set of parties with any set of comparative advantage or difference in their value of the object, is untapped. the efficient market hypothesis is a lazy lie. it is only directionally true if you ignore the expansion of the frontier of the possible. - there are weirdly only a small high school's worth of people that you find out are behind new things in the domains that you know about. maybe like 500 people or something. your job is to surround yourself with, and be one of those people. live in the future, and build what's missing. it's a literal scene. golden age that you get once in a while. don't forget it. - find a way to work like hell, over the long haul. it helps to be obsessed. that's why you gotta work on what you're obsessed with. - let the work pull you, as much as you can rather than push. that makes sure you have the right energy and direction. - nevertheless, some things just work nonlinearly better annealed or forged under high temperature and pressure, cyclically. - if you're trying to make a large difference as a small player out of a system, look for the resonances. what is conserved, or what can be. what are the capacitances, inductances (like pendulum swings, resisting the force but displaced in time), what are the dissipations or resistances. many things in economies, or ecosystems, or other dynamic systems, can be modeled to first order as simple RLC circuits, and it's a powerful model even designing systems to put things on or off resonance, stack them, or add dissipation or detuning. this goes across all systems, even psychological systems! - what haven't been explored because people haven't been brave enough? clever enough? haven't had access to a tool you do? look especially hard under that rock! - try to prove, by counterexample, what a physical or computational limit is, for the performance of an idea. now, depending on how hard or soft the limit is, can you flip it, and prove by construction, a physical example that reaches or approaches this limit? how far can you get? come at it from every angle. how close did anybody come, in some dimension, from every field of technology. what can you apply from those fields of technology to your idea, in a different combination than they did. - there are most geniuses dead than alive. you must learn from their lives and efforts. befriend them. don't just stand on their shoulders. learn from their methods. ask the questions that they did. ask the questions that they would if they were sitting on your shoulders. use what they would use, if they had the benefit of all the inventions since they past. they should be some of your closest friends, even if they're dead. they can live through your mind, hands, and heart. - think from the perspective of an electron. get really familiar with what makes electrons happy. that is everything from atomic happy electron theory (we love a full shell) to inference, which gives you everything from inferometry and subwavelength electron lithography to the formation of bandgaps itself. - if you know how waves work in D dimensions in one material at time/space scale, you can map it to any other scale. - the first order response to a problem, as in the first order change of something measurable (eg a force) in response to an input, is always missing something in nature. even though it's usually what pundits and statisticians point out. why? it's literally impossible to make the system do anything other than exponential growth (explosion) exponential decay, or stasis! mathematically. but as soon as you get to second other systems, like simple harmonic oscillators, like RLC circuits, like atoms, you can admit not just cyclic solutions, like simple Sine waves, but literally any superposition of them! so nature often stacks couplings in resonances between systems that have toy models that share them! and this presents why physics give so many useful toy models that are simple harmonic oscillators working well in one domain or another, and also why doing this, a highly trained part of the physics toolkit, has has such good yield in so many fields. not that it's perfect! this also explains for so many things, like business cycles, or emotions, have cycles. cybernetics. and why so many things respond to being driven at a resonant frequency. - if there's something standing between you and a solution to a long standing problem it's probably embarrassment. there's glory in cringing through it. pick up the phone. write the email. make the video. write the application. fill out the contact us form. give the talk. get on the plane. slide in the DMs. the universe is literally waiting for your courage. - that's why it keeps trying to teach you the same lesson until you pass the class - a lot of this is in Thinking in Systems. Read thinking in Systems. - a lot of this is in Taoist, Buddhist, even early Christian thought. Wisdom will make more sense as you get older. - try to keep up with the prodigies. they're coming up from behind. you don't want to be "i'm losing my edge" "the kids from London are coming up from behind." - music is memory, food is memory, smell is memory, moving (on foot) is memory, smell is memory, dance is memory. if you're having trouble thinking, go for a walk. get moving. get dancing. at least stretch. put cues in your environment. it's good for a 10x context window, seriously. - "perspective is worth 80 IQ" -- Alan Kay - if you can't solve a problem, what truths grow outward from it. embrace those. for me, that includes being a girl (transition, move tho SF, duh), being essentially unemployable (be an entrepreneur) or having liquid molten (yet transparent) sodium chloride coating our entire combustion system (make a continuous surface to a reservoir and ensure that NaCl is molten and thus transparent. now the problem is a solution, not only is the sodium transparent and wicks itself by capillary action, but it moderates the temperature gradients of sapphire and alumina ceramic by itself being a vapor chamber, clamped to the vapor temperature of the NaCl). also that only curved surfaces will print and sinter successfully -- ok -- make all forms out of compounding curved surfaces! etc. - one of the reasons is that invention is at the intersection of art and technology is that art is at the intersection of creation and imagination, and imagination is amplified by perception. and which the artist uses every tool at their disposal to amplify their powers of perception, and of creation, the technologist has the power to create their own tools to amplify their perception and powers of creation, using science and engineering, and *aided* by the art of invention and the unlocked imagination. it's a self-reinforcing loop! this explains why single genius and scenius can plow decades, even centuries after of other civilizations if they get the self-reinforcing reaction together, like light from an enhanced flame. - there's always another move. think! - if you can't think, play around. walk. look from another perspective. let your feet think. let your hands thing. lift heavy rock. make your muscles think. post. let your exobrain think. - honor thy mistake as intention. - invent a smarter brick, then invent a better meta brick (like an arch), then invent a way to compound an arch, before you build a cathedral. - leave yourself a juicy, simple task to get started with in the morning. - if you haven't, open a file. - you can always have second breakfast. - a latte is breakfast - there's free alpha in shilling good stuff that you know about, especially from friends. - if you're suffering, you probably aren't even at 40% of what you could try yet. have you tried really focusing for three more hours? have you tried long walking about it? have you tried calling up or cold emailing the world experts? have you tried begging? have you tried paying them money? have to tried promising shilling for them? probably not!! - be deeply comfortable with unresolved mysteries, questions, and problems, but never forget about them. remember the state of knowledge around them, and keep it in your back pocket. remember why and where you learned something. then, if something comes along, and sheds light on an old question or unresolved mystery, you put it together, and then people say "you're a genius" or possibly, a prophet. But you just are ahead of the average -- you're just relatively in the future relative to people referented to the past. - remember where and why you know something. that way you have more robust networks of knowledge, and can put new stuff together more easily, deriving from a deeper web of what you know, and if you get a big update on what you know or believe, you can update the entire network. this lets you be right an annoyingly large fraction of the time, not just because you're right on the object level more often, but because you're more often right about how right about you are. this lets your esteem and exobrain (what you tell other people in your company in your network) be much smarter and more useful. @paulg refers to this kind of thing as a "Morris Method" after Robert Morris, who is often right about how right he is, although I am sure I am on another side of the tradeoff between that and speculating! - the value of most of your work can be multiplied by positing about it. it costs close to nothing to have a high signal channel amongst the noise -- the noise you add is below the background and improves the SNR, and the algorithm sadly is routing it anyway. So just post a lot. It's the one thing you can control, and it's a lot better than the media posting the same version of an article 1000 times and then ditching and burying you when that's out of fashion. (ask me how i know). Become your own platform. Be ungovernable. - nobody will be more resentful of your success than people who think "that should have been me." learn to read resentment from others as envy, and learn to read your own envy as a mistake. - remember that no matter how much money others have, no one as as much currency locally as the one who is aware and courageous enough to be truly present with a problem, and has enough command of themselves, their creative powers, and the resources they can draw own to make steps toward a real solution happen. never doubt that a good idea presented clearly and the will, effort and resource and convening power to make something happen, can. the world is a lot more malleable than you think. over time, the universe will shape itself around the vision that you induct into the realm of the possible. the will to power is real; it is the force from within emerging from all life maybe i'll share more ideas as they come to me. it's about cultivating special interests and their many touchpoints to everything you see. it's from developing a useful bag of tricks. share more ideas yourself from your super secret special bag of tools in the replies below! 🥰 this Feynman method is for the people! x.com/DanielleFong/s…

keyboard_arrow_left Previous keyboard_arrow_right Next

@gregisenberg It’s limitation of x.com due to very poor search, management and actually being able to use them. This is exactly why I made Chrom add-in to download bookmarks and offline viewer that allows to do much more with it: github.com/sytelus/xarchi…

@JoeBlogs685544 I have access to information and practices that directly contradict the above. Om is a literal sound and can be perceived with training. It is not just a pointer, it is the substance of manifestation itself

it's been a cyber security stacked week, looking forward to agenteering with confidence that won't suffer from these shenanigans going forwards

Socket @SocketSecurity

2 weeks ago

🚨 BREAKING: Active supply chain attack across npm, PyPI, and Crates.​io. Socket detected TrapDoor, a crypto stealer campaign hitting 34 malicious packages and 384 versions and artifacts, with attackers repeatedly pushing new releases across ecosystems. TrapDoor targets

133 423 2K 771K 737

@JoeBlogs685544 Indeed it is. The Vedas from India explain that all manifestation is made from the primordial sound OM

@JoeBlogs685544 I love it, both Silence and 🕉️ are perfect answers. It's great to connect with you!

Bob McGrew has a framework I keep thinking about: in the AI future there are only two jobs. The Lone Genius and the Manager. That's it. Everything else gets absorbed. The Lone Genius is the person sitting alone at a computer, amplified 1000x by AI. One person with taste, vision, and relentless focus who can now do what used to take a team of 50. The Manager is the person who becomes CEO of their own "firm" where most of the employees are AI agents. They define the goals. They decide what matters. They coordinate. The AI does the execution. The Marxists will hear "two jobs" and panic. "What about everyone else?!" But here's what they're missing: AI doesn't shrink these two categories. It explodes them open. More people get to be geniuses. More people get to be managers. The barrier to entry for both just collapsed. What actually gets eliminated? David Graeber called them "bullshit jobs." Graeber was no libertarian! He inspired Occupy Wall Street. His words: "Huge swaths of people spend their entire working lives performing tasks they secretly believe don't really need to be performed. The moral and spiritual damage that comes from this situation is profound. It is a scar across our collective soul." Graeber said bullshit jobs are "a form of spiritual violence directed at the essence of what it means to be a human being." They induce "hopelessness, depression, and self-loathing." This is who the left should be fighting for. Not to preserve those jobs. To liberate people from them and give them better ones. The dirty secret of the modern economy: millions of people sit in roles so pointless that even they can't justify their existence. Compliance layers. Reporting layers. Coordination layers. Meeting-about-the-meeting layers. They know it's meaningless. It eats them alive. AI eats those layers. Good. That's a jailbreak. What I love about Bob's framework is where it points. The Lone Genius used to require a PhD, a lab, institutional backing. Now a 19-year-old with taste and Codex can ship what took a research team a year. The genius bottleneck was never talent. It was access. The Manager used to mean you needed to hire 50 people, raise money, build an org chart. Now you can orchestrate a fleet of AI agents from your laptop. The management bottleneck was never skill. It was capital. AI doesn't concentrate genius and management into fewer hands. It distributes them into more hands. The working class kid in West Virginia. The single mom in Ohio. The 55-year-old who got laid off and now builds software for the first time. Those are some of Bob's future geniuses and managers. The best founders I see at YC are already living this. They toggle between both modes in the same day. Morning: lone genius, creative insight, the thing nobody else sees. Afternoon: manager, spinning up agents, steering, shipping. The cycle time between genius and manager IS the new productivity metric. So when someone tells you AI means "only two jobs and everyone else starves," quote Graeber to them, they’ll get it. Graeber knew the real violence was making people do meaningless work and pretending it was dignity. AI ends that. More genius. More agency. Fewer spiritual prisons.

This is beyond nifty 👏🏾

Alvin Sng @alvinsng

2 weeks ago

One of my favorite superpowers of agents is building classifiers. It’s insanely high leverage. Before AI, you needed a year-round team: - 3 ML engineers to build the models - 3 ML infra engineers to scale them up - 2 software engineers to integrate the parts - 1 data scientist

56 82 976 76K 2K

A little secret. About 5% of our production traffic is on the Pi harness, about another 5% is on OpenCode. Reminder you can use your ChatGPT account in a flourishing set of other tools. We’ll continue to make Codex awesome, but you have options.

Your karma is your dharma

Cody Savage-Dragon/acc @NaturallyDragon

5 months ago

@lichthauch You are your own purpose.

0 0 29 10K 1

@MichaelDell @DellTech Impossible is nothing 🔥

I started @DellTech 37 years ago with $1000. Revenues in 1984 were $6 million. Last year's revenues were $94.2 billion. Impossible is nothing.

Om 🕉️

φ @QuanticASI

2 weeks ago

how your email finds me Even though I walk through the darkest valley, I will fear no evil, for you are with me; your rod and your staff, they comfort me Psalm 23:4

6 19 103 3K 8

Just wrapped up the interview process of stealth infra startup (NYC based) Team had people from OpenAI and Anthropic and Hedge funds. 5 rounds in total. Use of LLMs were completely restricted in the first four rounds. Round 1 started with a grid DP problem. Basic state transition was straightforward, but the brute force complexity was too high for larger constraints. Interviewer kept pushing on optimization. After thinking for a bit, I noticed the feasibility condition was monotonic: if a solution worked for some value "k", it would also work for larger values. So instead of directly searching the answer space linearly, I wrapped the DP inside a binary search over the answer. That dropped the complexity pretty hard and the interviewer seemed happy with the direction. We also spent some time discussing my search engine project: indexing pipeline, retrieval flow, ranking heuristics, and some infra decisions behind it. Overall probably my strongest round in the loop. Round 2 humbled me a bit. We have a tree with letters on edges. Count paths where the letters along the path can be rearranged into a palindrome. Main observation there was: a path is valid if at most one character has odd frequency. So instead of storing exact frequencies, I represented parity using bitmasks. Each bit represented whether a character count was odd/even along the path. I initially went with small-to-large merging on subtrees to optimize the brute force approach. But halfway through the discussion I realized there was probably a cleaner way using prefix xor masks + frequency counting on the tree itself. Could’ve likely reduced the implementation complexity a lot by treating it more like: “how many previous masks differ by at most one bit.” One of those rounds where you know the core idea, but afterwards your brain keeps replaying all the cleaner approaches you could’ve explored. Round 3 was systems design: I had to design a database schema branching system. Basically: - isolated DB schema branches for migration testing - schema diffing - merge conflicts between branches - metadata storage - avoiding physically cloning huge datasets - garbage collection for abandoned branches - how to visualize merge conflicts without making UX horrible Really fun discussion, i worked alot on this stuff in college, Told him about the tournament tree to collect the fanout results and three way schema merging algorithm. I was satisfied with this round. Round 4 caught surprised me. No coding. Pure database deep dive. We discussed - MySQL replication - binlogs - failover mechanics - zero-downtime schema migrations on 500M+ row tables - online migrations - high-level discussion around my vitess internals knowledge like VTGate and VTOrc. I didn’t know every detail perfectly, but I reasoned through most of it from first principles. Final round was with the CTO. Super chill compared to the earlier rounds. Mostly talked about: - why I like building systems - long-term interests - what kind of problems excite me - company vision and where they see the infra space going we even pair-programmed a tiny 8085 microprocessor emulator using Codex. debugged a few weird issues together, and honestly it felt less like an interview and more like two engineers yapping. Probably my favorite round from the entire loop. Final Verdict: rejected. But honestly, this was probably one of the best interview loops I’ve gone through. I personally think the reason of rejection was second round. I fumbled and struggle there. There could be alot more reasons. I didn't got any feedback where I lacked. Kinda crazy how much you learn when smart people keep pushing the edge of your understanding for hours straight. "Until death every defeat is psychological"

there is no better time in tech than now to be a jack of all trades, master of a few. just make sure to keep adding to the few year over year, such that the cumulative breadth of expertise you collect becomes an increasingly rare combo. remember, if you're top 10% in 3 different areas, that already makes you top 0.1%. keep switching it up until you get to "your best", and then switch it up again (great for a particular flavor of people who don't enjoy resting on laurels, maybe not so great for others). question all institutional value and pedigrees, all traditional career paths or corporate ladders: the college industrial complex is getting shaken up, alongside a disappearing managerial class, so if you're pursuing either make sure you are fully internally aligned with why. social/political capital in a particular institution can feel incredible, but if you're spending all your energy on complex political people games, you're not a technologist anymore, you're an unelected politician. if you're ok with that, then all's well. critical thinking is more important than ever: take nothing at face-value, question everything and everyone. the equivalent of ai slop can be found in humans operating under misaligned incentives and interests. the sooner you're clued into disambiguating the talkers/larpers from the doers, the better off you'll be figuring out where and who to invest your time in. the anxiety of job displacement is very real, since a surprising amount of white collar work/prestige is built on a performative house of cards, significantly lacking in correlation with technical breadth, depth, and skill. as long as you keep learning, keep building, keep producing receipts, you will be fine. if all that sounds ok to you, welcome to the world of technology! it's truly one of the few places you can experience child-like wonder every few years, and be constantly humbled & excited by new adventures, as scary as they may seem at first. don't give up, drink your water, get your sunlight, and take breaks as needed. tech careers are notoriously nonlinear, so you might as well embrace it and enjoy the ride!

just checked out my /goal

@

57 years ago

0 0 0 0 0

claude is happy

Andrej Karpathy @karpathy

3 weeks ago

Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.

8K 11K 150K 27.4M 14K