The thing that inspires my writing is that the best sentences are self evident. Meaning you declare it without evidence and it feels so intuitively right to most people. It resonates, either being their lived experience, or being the inevitable conclusion of a line of thinking.
Making a sentence like requires deeply understanding a problem space to the point where these sentences emerge, rather than any "craft" of writing.
So the craft is thinking through a topic, usually by writing about it, and then deleting everything you've written because you arrived at the self evident position, and then writing from the vantage point of that self evident statement.
I feel that writing is a personal craft and you must dig it out of yourself through the practice of it, rather than learn it from others. The usage of AI as a resource makes this much clearer to me. You must be confident in your own writing not because it is following best practices or techniques of others but because it is the best version of your own voice at the time of being written.
If you are a data scientist or do anything with data... duckdb is like a swiss army knife. So many great ways it can help your workflow. The original video from CMU in 2020 [1] is a classic. Minutes 3-8 present a good argument for adding duckdb to your data cleaning/processing workflow.
And if you want to add a semantic layer on top of data, Malloy [2] is my favorite so far (it has duckdb built in):
I had to code something on a plane today. It used to be that you couldn't get you packages or check stackoverflow. But now, I'm useless. My mind has turned to pudding. I cannot remember basic boilerplate stuff. Crazy how fast that goes.
As a Spaniard, this also happens to me. You can either use a VPN or just switch DNS servers to one that doesn’t have anycast nodes in Spain.
Cloudflare’s authoritative DNS uses EDNS Client Subnet (ECS) to return different IP pools based on where the query originates. Spanish resolvers get IPs from a range that La Liga blocks. If your recursive resolver is physically outside Spain (or you use DoH/DoT to tunnel to one), Cloudflare returns a different, unblocked pool.
I know Arc AGI 2 has a private test set and they have a good amount of results[0] but it's not a conventional benchmark.
Looking around, SWE Rebench seems to have decent protection against training data leaks[1]. Kagi has one that is fully private[2]. One on HuggingFace that claims to be fully private[3]. SimpleBench[4]. HLE has a private test set apparently[5]. LiveBench[6]. Scale has some private benchmarks but not a lot of models tested[7]. vals.ai[8]. FrontierMath[9]. Terminal Bench Pro[10]. AA-Omniscience[11].
So I guess we do have some decent private benchmarks out there.
The future (if we keep using money to allocate resources) is something akin to feudalism but worse. If you are born at the bottom you will never rise to the top. It's bleak. Even worse, your labor will not be needed, nor will your intellectual abilities. There will be a few well off people with capital. The data centers will be guarded by automatons and drones. Everyone else will essentially live in a parallel economy that is borderline biblical. Countries like this already exist in the form of countries with excess access to a single natural resource. See the https://en.wikipedia.org/wiki/Resource_curse
The article itself does an excellent job spelling out the background:
> This style has a history, of course, a history far older than the microchip: It is a direct linguistic descendant of the British Empire. The English we were taught was not the fluid, evolving language of modern-day London or California, filled with slang and convenient abbreviations. It was the Queen's English, the language of the colonial administrator, the missionary, the headmaster. It was the language of the Bible, of Shakespeare, of the law. It was a tool of power, and we were taught to wield it with precision. Mastering its formal cadences, its slightly archaic vocabulary, its rigid grammatical structures, was not just about passing an exam.
> It was a signal. It was proof that you were educated, that you were civilised, that you were ready to take your place in the order of things.
Much of writing style is not about conveying meaning but conveying the author's identity. And much of that is about matching the fashion of the group you want to be a member of.
Fashion tends to go through cycles because once the less prestigious group becomes sufficiently skilled at emulating the prestige style, the prestigious need a new fashion to distinguish themselves. And if the emulated style is ostentatious and flowery, then the new prestige style will be the opposite.
Aping Hemingway's writing style is in a lot of ways like $1,000 ripped jeans. It sort of says "I can look poor because I'm so rich I don't even have to bother trying to look rich."
(I agree, of course, that there is a lot to be said for clean, spare prose. But writing without adverbs doesn't mean one necessarily has the clarity of thought of Hemingway. For many, it's just the way you write so that everyone knows you got educated in a place that told you to write that way.)
In my experience a culture where teammates prioritise review times (both by checking on updates in GH a few times a day, and by splitting changes agressively into smaller patches) is reflected in much faster overall progress time. It's definitely a culture thing, there's nothing technically or organisationally difficult about implementing it, it just requires people working together considering team velocity more important than personal velocity.
Amazing talk. Here's a quick writeup if you don't want to watch the full hour or don't have enough hardware knowledge to follow what Markus is talking about, as he goes very fast, in some cases too fast to even let you read the text on his slides. It's mandatory to use the pause key to understand the full details even if you have a deep understanding of every relevant technology, of which he explains none.
The Xbox uses a very advanced variant of the same technologies that also exist on smartphones, tablets and Secure Boot enabled PCs. When fully operational the Xbox security system prevents any unsigned code from running, keeps all code encrypted, proves to remote servers (Xbox Live) that it's a genuine device running in a secure state, and on this base you can build strong anti-piracy checks and block cheating.
The Xbox has several processors and what follows applies to the Platform Security Processor. When a computer starts up (any computer), the CPU begins execution in a state in which basically nothing works, including external communication and even RAM. Executions starts at a 'reset vector' mapped to a boot ROM i.e. the bytes are hard-wired into the silicon itself and can't be changed. The boot ROM then executes instructions to progressively enable more and more hardware, including things like activating RAM. Until that point the whole CPU executes out of its cache lines and can't use more memory than exists on-die.
Getting to the state where the Xbox can achieve all its security goals thus requires it to boot through a series of chained steps which incrementally bring the hardware online, and each step must verify the integrity of the next. The boot ROM is only 19kb of code and a few more kb of data, and can't do much beyond just activating RAM, the memory mapping unit (called MPU on the Xbox), and reading some more code out of writeable flash RAM. The code it reads from flash RAM is the second stage bootloader where much more work gets done, but from this second stage on it can be patched remotely by Microsoft. So if bugs are found there or in any later stage, it hardly matters because MS can issue a software update and detect remotely on Xbox Live servers if that upgrade was applied, so kicking out cheaters and pirates. The second stage boot loader in turn loads more code from disk, signature checks and decrypts it, sets up lots of software security schemes like hypervisors and so on, all the way up to the OS and the games.
Therefore to break Xbox security permanently you have to attack the boot ROM, because that's the only part that can't be changed via a software update. It's the keys to the kingdom and this is what Markus attacked. Attacking the boot ROM is very, very hard. The Xbox team were highly competent:
• Normally the bringup code would be written by the CPU or BIOS vendors but MS wrote it all in house themselves from scratch.
• The code isn't public and has never leaked. To obtain it, someone had to decode it visually by looking at the chip under a scanning electron microscope and map the atomic pictures to bits and then to bytes.
• Having the code barely helps because there are no bugs in it whatsoever.
So, the only way to manipulate it is to actually screw with the internals of the CPU itself by "glitching", meaning tampering with the power supply to the chip at exactly the right moment to corrupt the state of the internal electronics. Glitching a processor has semi-random effects and you don't control what happens exactly, but sometimes you can get lucky and the CPU will skip instructions. By creating a device that reboots the machine over and over again, glitching each time, you can wait until one of those attempts gets lucky and makes a tiny mistake in the execution process.
Glitching attacks predate the Xbox and were mostly used on smartcards until the Xbox 360, which was successfully attacked this way. So Microsoft knew all about them and added many mitigations, beyond "just" writing bug free code:
1. The boot ROM is full of randomized loops that do nothing but which are designed to make it hard to know where in the program the CPU has got to. Glitching requires near perfect timing and this makes it harder.
2. They hardware-disabled the usual status readouts that can be used to know where the program got up to and debug the boot process.
3. They hash-chain execution to catch cases where steps were skipped, even though that's impossible according to program logic.
4. They effectively use a little 'kernel' and run parts of the boot sequence as 'user mode' programs, so that if sensitive parts of the code are glitched they are limited in how badly they can tamper with the boot process.
And apparently there are even more mitigations added post-2013. Markus managed to bypass these by chaining two glitch attacks together, one which skipped past the code that turned on the MMU, which made it possible to break out of one of the the usermode 'processes' (not really a process) and into the 'kernel', and one which then was able to corrupt the CPU state during a memcpy operation, allowing him to take control of the CPU as it was copying the next stage from flash RAM.
If you can take control of the boot ROM execution then you can proceed to decrypt the next stage, skip the signature checks and from there do whatever you want in ways that can't be detected remotely - however, the fact that you're using a 2013 Phat device still can be.
>Although cholesterol is protective against oxidative and cytolytic damage, the chronic free radical exposure will oxidize it. During the low cholesterol turnover of hypothyroidism, the oxidized variants of cholesterol will accumulate, so cholesterol loses its protective functions.
>In the healthy organism, cholesterol is constantly being synthesized, and constantly converted into steroid hormones, and, in the liver, into the bile salts that are secreted to emulsify fats in the intestine. Thyroid hormone and vitamin A are used in the process of converting cholesterol into pregnenolone, the immediate precursor of progesterone and DHEA. Anything that interfered with these processes would be disastrous for the organism. The supply of cholesterol, thyroid and vitamin A must always be adequate for the production of steroid hormones and bile salts. When stress suppresses thyroid activity, increased cholesterol probably compensates to some extent by permitting more progesterone to be synthesized.
>In very young people, the metabolic rate is very high, and the rapid conversion of cholesterol into pregnenolone, DHEA, and progesterone usually keeps the level of cholesterol in the blood low. In the 1930s, a rise in the concentration of cholesterol was considered to be one of the most reliable ways to diagnose hypothyroidism (1936 Yearbook of Neurology, Psychiatry, and Endocrinology, E.L. Sevringhaus, editor, Chicago, p. 533). With aging, the metabolic rate declines, and the increase of cholesterol with aging is probably a spontaneous regulatory process, supporting the synthesis of the protective steroids, especially the neurosteroids in the brain and retina.
I remember using Pidgin in ~2009. A dozen chat networks, all on one app. Desktop software built with a native GUI toolkit. And, on top of all that: you could keep your chat logs forever. The world of yesterday.
I'm biased, as I lead the Zulip project. But I think this is a reasonable place for me to post some thoughts.
Given current events in the USA, I can't emphasize enough how worried one should be about the fact that a few companies like Discord, Google (Gmail), and Meta have databases with access to the private conversations of hundreds of millions of people with their closest friends and family members, linked up with their identity.
Some of the big strengths of running a self-hosted Zulip server for your community are:
- Zulip servers are operationally simple, highly stable and easy to upgrade.
- Zulip is much better than Discord or Slack for managing the firehose of busy communities. Or at least, a lot of people tell us that they prefer the user experience to everything else they've tried, after a few weeks of getting used to it. :)
- Your community leaders get to make the policy decisions about data protection, identity, etc.
- It's 100% FOSS software, with an extremely readable and maintainable codebase that ~1500 people have successfully contributed code to. I don't think you'll find modern alternatives with a comparable featureset to Discord that are more resilient to the sponsoring company being acquired or going out of business.
- We are a values-focused organization (https://zulip.com/values/) where providing a public service is important to us all.
- Each server is completely self-contained and independent, with the only centralized services needed from us being desktop/mobile app publication and mobile push notifications delivery (which is free for community use and soon to be E2EE).
You too can implement a slither of the features you need right now in an insecure form without hosting, support, maintenance or future development.
This reminds me of the people who think they can build docsend in a weekend. No, you cannot. You can build a wee throwaway app with some of the features of docsend. But that is not equivalent to what people pay docsend for.
Businesses and SaaS aren’t just a bunch of static code. Code is a part of them, but it’s actually a minority of the work and the service. It’s very common to see founders fall out because CTOs believe product and company = tech and CEOs do not.
If you’re a technical founder, learning this lesson will separate you from the pack.
- W. Brian Arthur, "Increasing Returns and Lock-In" (1989): small early advantages + network effects => path-dependent monopolies: https://www.jstor.org/stable/2234208
W. Ross Ashby, "Law of Requisite Variety" (1956): controllers need at least as much "variety" as the environment and it often pushes toward central coordinating mechanisms (or many distributed ones with enough capacity) http://pcp.vub.ac.be/books/AshbyReqVar.pdf
Gilbert & Lynch, proof of the CAP theorem (2002): in distributed computing you can’t have perfect consistency + availability under network partitions and real systems centralize/compromise to cope: https://groups.csail.mit.edu/tds/papers/Gilbert/Brewer2.pdf
== State power, legibility & infrastructure (why governments centralize)
James C. Scott, Seeing Like a State (1998): states seek legibility and large projects favor central plans and standardized populations/landscapes https://en.wikipedia.org/wiki/Seeing_Like_a_State (literally anything by Jim Scott -RIP- will be useful)
Elinor Ostrom, Governing the Commons (1990): shows conditions (clear rules, monitoring, graduated sanctions, polycentric governance) under which decentralized, federated management of shared resources succeed https://www.cambridge.org/core/books/governing-the-commons/A...
----
the literature making a counterpoint is abundant / overwhelming but that feels bleak considering when reading these works, systems thinking )the basis for "la technique") favors centralization
It's actually the buffering in this case that will get you dinged. The stated 110ms "lag" is probably the minimum time between keystrokes ever. If you have ever recorded data on the mean time between keystrokes you get a nice even distribution but for someone on a KVM it will look very skewed with most being under 110ms and zero below 110ms which is impossible for a normal human at a machine to replicate
Furthermore, there are a number of other side channel attacks here you could use to make things really inconvenient. Something super powerful would-be having a fido2 key such as a YubiKey and recording the mean time to human press keypress. Your average person who is present at the machine will touch the button in a number of seconds. A remote operator in NK will have to summon the homeowner which could take significantly longer.
Another technique you could use is look at the mouse movement data. You would also see the same truncated. distribution, I think a few people have put together a PoC for detecting cheaters in games based on mouse movements.
I do wonder also if the KVM devices they are using support HDCP. Showing media over HDCP on the screen that instructs the user to write an email or make a phone call instantly would be pretty cool.
Metabolism converts sugars and oxygen to carbon dioxide and water. Fire has the same inputs and outputs too! You can think of metabolism as releasing the same energy as fire but with more steps so our bodies can harness and use the energy via various other chemical reactions.
Most of the water from metabolism is released in your breath. Water you pee is mostly water you ingested via drinking and eating.
I used to do Buteyko breathing exercises. The main idea is that over breathing results in a loss of co2 and less co2 in the body results in less oxygen being released from the blood due to the Bohr Effect https://en.wikipedia.org/wiki/Bohr_effect.
No more Google. No more websites. A distributed swarm of ephemeral signed posts. Shared, rebroadcasted.
When you find someone like James and you like them, you follow them. Your local algorithm then prioritizes finding new content from them. You bookmark their author signature.
Like RSS but better. Fully distributed.
Your own local interest graph, but also the power of your peers' interest graphs.
Content is ephemeral but can also live forever if any nodes keep rebroadcasting it. Every post has a unique ID, so you can search for it later in the swarm or some persistent index utility.
The Internet should have become fully p2p. That would have been magical. But platforms stole the limelight just as the majority of the rest of the world got online.
Vitalik touched upon this briefly in an other-wise long and wide-reaching essay. I think its a good treatment of the topic that the author is talking about.
He categorizes the ecosystem broadly into 4 cohorts- [token holders] (which includes investors, speculators, etc.), [pragmatic users] (actual end-users who spend crypto to buy stuff), [intellectuals] (who give the vision and ideology), [builders] (of blockchains, apps, etc.) - These 4 groups come together but with different motivations and there is a gap in understanding between them. Indeed, there is even resistance against trying to reach an understanding - one which plays out in the comments section of every crypto-related post on hn. The author of this twitter-post clearly falls under [intellectual, builder] and has been disillusioned by the speculators from [token-holders]. Yet the [token-holders] are a vital component (as are the other groups) as they fund most of the development and adoption. Ultimately these 4 groups have more in common than not. The challenge going forward is to balance the occasionally conflicting needs of all the 4 groups, which includes checking the excesses of each group, while try to achieve a consensus. (Vitalik provides a nice diagram that maps out what that would look like). Crypto is an experiment in economics and economics is a science as well as a social-science. Anyone looking for a good solution must seek to understand and address the psychology of all the actors involved.
> Many investors focus on the right time to buy stocks because they don’t want to buy near a peak in case of a future market crash. I understand the feeling. With the market near all time highs in early 2017, it can be tempting to hold off until there is a larger negative adjustment in prices.
> The only problem with this approach is the market could go up for a significant period of time before a correction happens.
> For example, if you search “stock market overvalued 2012” on Google you can find plenty of stories discussing how overvalued stocks were in 2012. If you had started waiting for an S&P 500 correction then, you would have missed out on the ~70+% increase in prices from 2012 through early 2017.
And what has the S&P 500 done since 2017? Continuing:
> If I still haven’t convinced you, let me tell you a story. The story is about a man with possibly the worst luck in investing history.[0][1] He made a total of 4 large stock purchases between 1973 and 2007. He bought in 1973 before a 48% decline in stocks, bought in 1987 before a 34% decline, bought in 2000 before the dot com crash, and bought in 2007 before the Great Recession.
> Despite these 4 individual purchases that totaled a little less than $200,000, how did he do? He ended up with a $980,000 profit for a 9% annualized return. What was his secret? He never sold.
> That’s right. Selling out is literally selling out your future wealth. You need to hold on to your assets as you acquire more.
> This is the purpose of capitalism (i.e. acquiring capital). The only time you should sell your investments is for rebalancing (annually/quarterly, etc.) or in retirement. Otherwise, you already know the mantra.
And as Maggiulli shows, trying to buy the dip, even when you know when the dip will occur (which is impossible), gives worse results than simply putting away a little bit every month:
Put away a little every pay cheque, in a diversified low-fee fund (S&P 500, Total Market/Russell 3000), at a comfortable risk profile that allows you to sleep at night (0/20/40% bonds), and try not to pay attention too much about what The Market™ is doing. Most of us are investing for retirement, and if you do the above, you'll probably end up with a decent nest egg.
Save a little for the future, and enjoy the present.
My wife has to eat protein (in her case meat) with carbs. If she eats carbs with no protein she has anxiety. It's so awesome that we are finally studying things like this.
Damn, thanks for your insights. I recently discovered how much better I sleep when fasting. During my fasts I'm taking precise amounts of electrolytes (sodium, magnesium, potassium: https://en.wikipedia.org/wiki/Oral_rehydration_therapy). Being off my fasts I'm getting sloppy with my electrolytes intake and here we are: I'm getting the same symptoms you describe. Not severely, but noticeably.
Now I need to find a proper supplement in Germany. Most electrolytes that also have the recommended amounts of glucose contain artificial sweetener which is a big no-go for daily usage for me.
There's no single mastermind. This current wave of authoritarianism around the world is a consequence of not designing the Internet with democratic principles in mind. Online content discovery and moderation mechanisms are centralized and authoritarian in nature. And since most communication nowadays happens on the Internet on large platforms with millions of users (this is especially true after smartphones and social media were invented), the structure of human society in the real world is mirroring the Internet.
This can be solved, though. We have to move moderation and ranking mechanisms to the client-side, especially for search engines and social media. Each person should be able to decide what they post and see, but not what anyone else posts or sees.
It is interesting that most of our modes of interaction with AI is still just textboxes. The only big UX change in that the last three years has been the introduction of the Claude Code / OpenAI Codex tools. They feel amazing to use, like you're working with another independent mind.
I am curious what the user interfaces of AI in the future will be, I think whoever can crack that will create immense value.
While I agree with the larger sentiment, there is a rather substantial amount of diet and nutrition science with the core being quite an agreed upon set of foundational principles.
If I were to summarize:
1. We derive energy from three major macronutrients (protein, fat and carbohydrates)
2. Our body has certain macronutrients it cannot synthesize and NEEDS for healthy functioning (essential fatty acids and essential amino acids)
3. Our bodies are largely energy efficient (by which I mean the processes to produce energy are readily actionable with average hormone levels of insulin/ghrelin/leptin etc.) when consuming carbohydrates, but can turn to other nutrient sources when they are not to be found.
4. Our body composition can be manipulated by manipulating our energy expenditure either via manipulating the caloric content of food or via the preferential production of body tissue from horomonal influence (this is why muscle mass increases when using testosterone even if caloric intake is kept constant but above a certain macronutrient threshold).
5.There are a certain set of micronutrients we need on a continuous basis (the exact amounts of which are up for debate but the ranges seem to be reasonably agreed upon).
6. Our diet influences our gut fauna and this further influences how much nutritional value we get from our diet, it is quite a symbiotic cycle.
7. Barring certain specific horomonal imbalance diseases, everyone's body largely abides by the above principles, (i.e as special as everyone thinks they are, they're probably more average than not when it comes their metabolism).
N=1 but in 2017 I lost over 100 pounds in 8 months by changing to a keto + IF diet and I've kept it off. I lost 10 pounds in 10 days and 20 pounds the first month. At around six weeks I became 'fat-adapted', a long-term metabolic transition to primarily burning fat instead of carbs (glucose) for energy. I didn't start with IF but at around that point I sort of fell into intermittent fasting because it just felt right. I'd heard about IF but never had it as a goal because it seemed impossible since I'd been hungry my whole life. But limiting carbs with keto controlled my blood sugar to the extent I was almost never hungry which made IF trivially easy. So if you're trying IF and struggling with hunger pangs, try managing blood sugar by reducing carb intake.
The combination of Keto+IF worked so well for me, for a while my calorie tracking switched from the usual preventing eating too many calories to ensuring I was getting enough, which was certainly never on my bingo card. After a lifetime of being a slave to hunger it was liberating to suddenly feel effortlessly in control of diet and my relationship with food changed completely. Then at around 90 days my palate shifted, meaning I even lost my taste for carbie foods. If I tried a small bite of something carb-laden that I'd loved my whole life, it didn't even taste particularly good to me anymore. I also became hyper-sensitive to sugar. Sugar-soaked foods just taste poisonously over-sweetened (which they kind of are). A slice of apple now tastes as sweet as I'd ever want, like a dessert that has extra sugar-added.
In the 8th month I reached below my ideal 'dream' weight and even saw abs appear for the first time in my life! I transitioned to maintenance mode but stayed keto because being in a blood sugar controlled state felt so amazing and not just physically but also mentally and emotionally. At around a year I went from strict keto to low carb for life which I still am 8 years later. When I started that was unimaginable. I saw keto as an onerous regimen that I'd endure if it worked and stop the second I wasn't overweight. But during the journey my metabolism, palate and food preferences changed so dramatically, I was basically a different person when I arrived. Those first few months when I was rigorously tracking every calorie in an app and managing intake with measuring cups and a kitchen scale felt like a burden but were actually invaluable skill-building. After a few months all that process became automatic so I didn't need to constantly track and by six months I got to the point where I don't even think about it consciously. That early rigor helped me get so in sync with my body and able to sense where my metabolism is in its natural cycles that now I just eat when necessary and convenient for my schedule. This often ends up being IF but it's not intentional on my part, which makes me think maybe IF patterns evolved in the hunter/gatherer era as part of our natural biological rhythms. Due to habit and carb-laden factory foods I'd never been able to access those rhythms until I made the conscious effort to break the patterns I'd been raised in.
Making a sentence like requires deeply understanding a problem space to the point where these sentences emerge, rather than any "craft" of writing.
So the craft is thinking through a topic, usually by writing about it, and then deleting everything you've written because you arrived at the self evident position, and then writing from the vantage point of that self evident statement.
I feel that writing is a personal craft and you must dig it out of yourself through the practice of it, rather than learn it from others. The usage of AI as a resource makes this much clearer to me. You must be confident in your own writing not because it is following best practices or techniques of others but because it is the best version of your own voice at the time of being written.