There was an "ask HN" a week or so ago about knowing how your personal machine was or was not compromised ...
I am imagining training a model on all of the PDFs and email and receipts and contracts that I have spent the last 30 years protecting ... and then exploding my home directory with 2M parallel copies.
Now what ?
Which is the real correspondence ? Which contracts are real ? Which invoices and receipts ?
Could one construct a framework where all data at rest is worthless and only witnessed transactions and traffic have any surveillance value ?
There's probably at least three different kinds of when we do this: (1) when a subthread is off topic but the parent comment isn't; (2) when a subthread is a step into flamewar or into more generic discussion, or otherwise breaks the site guidelines; (3) when a comment was posted as a child but isn't replying to anything specific in the parent—this happens a lot, especially when the original parent is the top comment at the time.
Edit: oh also (4) when a subthread is particularly good and on-topic and there's some problem with the parent (e.g. it's downvoted or downweighted) that is preventing the good subthread from getting the attention it deserves.
Edit: oh, also (5) when a subthread is extremely top-heavy and we're trying to prune it to make the tree more balanced.
Ultimately companies that abdicate their informatics operations like this will give their profits to their data-center operators, who will be empowered to charge them whatever price they want. Because what's their BATNA? Migrating from Azure to AWS when Microsoft doesn't want to let them?
Renting your information infrastructure is a great way to reduce startup costs, but down the road, that information infrastructure runs your company. Trying to outsource it is like trying to outsource upper management.
To be clear, I'm not saying that the optimal amount of cloud services for an established company like FedEx to buy is 0. They bring in management consultants, too. But it sure isn't 100%.
My Typesetting Markdown series[0] does a deep dive into how to keep content and presentation separate, which is much easier using ConTeXt than LaTeX, IMO. Part 8 of the series delves into typesetting annotations. As for TeX Live, I agree. The instructions for installing ConTeXt on Windows is basically: download, unzip, run two batch files.[1]
It's definitely a longform excellent (free!) article about how good content is paywalled, but the propaganda/lies are free.
-------------Excerpt-----------
But let us also notice something: the New York Times, the New Yorker, the Washington Post, the New Republic, New York, Harper’s, the New York Review of Books, the Financial Times, and the London Times all have paywalls. Breitbart, Fox News, the Daily Wire, the Federalist, the Washington Examiner, InfoWars: free! You want “Portland Protesters Burn Bibles, American Flags In The Streets,” “The Moral Case Against Mask Mandates And Other COVID Restrictions,” or an article suggesting the National Institutes of Health has admitted 5G phones cause coronavirus—they’re yours. You want the detailed Times reports on neo-Nazis infiltrating German institutions, the reasons contact tracing is failing in U.S. states, or the Trump administration’s undercutting of the USPS’s effectiveness—well, if you’ve clicked around the website a bit you’ll run straight into the paywall. This doesn’t mean the paywall shouldn’t be there. But it does mean that it costs time and money to access a lot of true and important information, while a lot of bullshit is completely free.
I wonder if a volume collective license agreement with ISPs would do it. I still find good, incisive reporting to be highly useful. The problem is the same as with subscribing to streaming services these days: there's too many. I might read from several sources but they all want payment information, authentication credentials, I have to be logged in, etc.
Would anyone notice if there was an extra $1.50 tacked on to your Internet bill? Some organization collects it from the ISPs and redistributes to the new sites based on viewership statistics from a trusted third party with a reserve for new/independent/grassroots/minority sources. Then I don't have to log in, renew my subscription to each site, and they still get funded for reporting.
I'm not super connected to the media publishing world or anything but it seems to me like a concentration-of-power thing is creating incentives for sensational/entertainment reporting than journalism. A system where billionaires wouldn't have a reason to buy up all of the sources for their circus would be nice.
"If people have the right to be tempted - and that's what free will is all about - the market is going to respond by supplying as much temptation as can be sold." - Eliezer Yudkowsky, LessWrong, "Superstimuli and the Collapse of Western Civilization, March 2007
There was a frightfully insightful recent twitter thread making rounds, saying that extractive industries are perfect to be headed/owned by mob bosses, because the more complex the industry the more leverage the technical employees have. I.e. in extractive industries they have close to none, in manufacturing they have a solid measure of respect, while in tech industries they get to dictate the culture of the company.
Which means a mob leadership (we're not pretending anymore Russia doesn't have one of those, right?) will favor extractive industries over everything else, not just because they're much easier to own and exploit, but because allowing other types of companies to grow creates power centers outside of their control. They actively prefer to outsource services abroad, even if they are more expensive and create reliance on foreign partners.
Which means Germany's bet was wrong on two counts:
1. It propped exactly the kind of industries where the money goes in mob's pockets
2. It was unnecessary, because Russia was making itself incredibly vulnerable to sanctions all by itself by avoiding to develop anything more complex than a well or a mine (normally I'd say defense excluded, but seeing the quality of their current armed forces, it's more like an argument that even defense is included).
The natural question here is if the blunder is an honest mistake... but honest mistakes and motivated reasoning go suspiciously well together, and by all accounts the German elites more than occasionally gained from trade with Russia.
> Where I have chemical structure X, and a system generates the process steps and chemicals needed to produce X.
Undergraduate chemistry students spend a fair amount of time learning how to look at a novel structure X and by disconnecting "backwards" it into simpler components, deduce a route by which it might be synthesed "forward" in the laboratory from readily available starting materials.
There's an excellent book on this, "Organic Synthesis: The Disconnection Approach", by Stuart Warren.
Money doesn't objectively exist; it's purely a figment of people's imagination. This ten-trillion-Zimbabwean-dollar bill isn't money anymore, even though it used to be, and it isn't physically changed in any relevant way. The only thing that changed was how people thought of it.
It might seem strange to say that a purely imaginary thing like money could be "the driver/catalyst for everything" in industrialization, but many strange things are true. I'm writing you this note in letters of lightning far too small to see, which are persistently trembling within a few tiny slivers of quartz that can thus contain the lightning because the vital air has been removed from it—stranger still, but true, because that is how DRAM chips work.
Still, let's see if we can analyze how industrialization happens in terms of objectively existing reality rather than shared hallucinations like money—surely the picture will be partial, but it may still be useful. Industrialization itself seems to be objectively observable: it doubles people's life expectancy at birth.
Let's start with an outside view of economic production as a whole. How would we explain it to a Martian?
People work, planting and harvesting and preparing food. If they do not eat food, they die in a few months from lack of various molecules crucial to their biochemistry and energy (collectively "nutrition"). They expend energy and water and damage their bodies by working, damages which they repair automatically if they have time to rest, but which will otherwise eventually kill them. So their capacity for work is limited. Not everyone has to work, but if the people collectively produce too little food from their work, then they will start dying from lack of nutrition, and pretty soon they will all die. One old book of poetry sums this up by saying, "In the sweat of thy face shalt thou eat bread."
On the other hand, if they can make a lot of food despite their limited work capacity, more than they need to survive—a "surplus"—many of them will live for decades, they will reproduce, and their population will increase, typically by about 4% per year. Groups that manage to do this come to vastly outnumber groups that do not.
(There are some other material necessities for human survival and reproduction besides food, but none of them are different in relevant ways, so for the time being I'll stick to food.)
The other conventional factors besides labor that affect productivity are land and something called "capital". Some land yields a lot of food and is easy to plant and harvest; other land yields very little. By "capital" is conventionally meant not money but durable goods that increase labor and land productivity, such as hoes and waterwheel-driven mills for making flour. If people don't need to spend their entire labor capacity on making enough food to survive, they can instead devote some of it to making these durable goods: cords, needles, ropes, bows and arrows, shoes, knives, hoes, axes, pots, and so on. Cords and ropes enable you to climb palm trees, hang meat over a fire, or bridle a horse; needles enable you to make clothing and all manner of sturdy, flexible fiber goods; bows and arrows enable you to kill dinner at a distance; shoes enable you to walk or run longer distances; knives and axes make cooking and woodworking much easier; hoes allow you to plant much more land; pots enable not only cooking but food storage and the storage of other goods; and so on.
So these are "capital goods": once you have a surplus, then by spending some of your work creating durable goods instead of satisfying basic necessities for survival, you thenceforth multiply their possessors' future production, or at least until the durable goods stop enduring. But a hoe or a blast furnace is produced in the same way as a vegetable garden: people work to turn raw materials into the desired end product, directed by their skills and knowledge. Instead of hoeing the field today, perhaps they're hammering the hoe blade into shape or molding bricks from fireclay. Most capital goods are more or less specialized to a particular sort of production—you cannot harvest more corn by dumping blast furnaces or gristmills into the cornfield—though some are more versatile than others.
I haven't mentioned skills and energy before, but they're crucially important. A person is skilled when they can direct their work to produce their intended results with ease and precision, and labor specialization develops skill much more highly.
Energy is a fundamental physical quantity which can be converted between many different forms, including heat and kinetic energy; many kinds of work require a lot of kinetic energy and a lot of heat, and this has historically often been the limiting resource for work productivity, particularly when all the relevant kinetic energy came via people's muscles from their food, and all the relevant heat came from firewood.
The available kinetic energy for production has greatly expanded five times in the past: ox-yokes (6000 years ago), horse-collars (1500 years ago, in the Sui), the steam-engine (250 years ago), and electric dynamos and motors (140 years ago). Photovoltaic cells seem at long last to be adding a sixth item to this list now that they have finally become cheaper than steam-engines.
The available heat energy for production has greatly expanded three times: fire (400,000 years ago), oil drilling (2300 years ago), and deep-shaft coal mining (250 years ago). Again, photovoltaics are probably a huge factor here today.
So, in these terms, we can analyze industrialization as a set of several major synergistic shifts.
First, enormously increased specialization, which would have been counterproductive to effective productivity in the absence of the worldwide markets created by the British Empire and by cheaper transport via canals. The ten workers in Adam Smith's pin factory could produce 48000 pins a day, far more pins than a village could use, or maybe even a city—perhaps each household might ruin five or six pins in a day as they repaired their clothing. Without easy shipping, the tonnes of pins thus produced would be as worthless as an asteroid belt converted into paperclips.
Second, enormously increased use of capital goods enabled by that specialization—only a few decades after Smith observed his pin factories of ten or eighteen men, the whole operation became automated by cam-driven pin-making machinery, further decreasing the labor per pin, and of course the rise of the spinning-machines and jacquard and other automated looms is even more notorious.
Third, as improved capital goods like the cotton gin and the combine harvester became abundant, the planting and harvesting of larger and larger amounts of land required less and less labor, so smaller and smaller amounts of farming work per person sufficed to prevent mass starvation. This surplus manifested as a mass population migration from the countryside to the cities, where increasingly specialized capital-intensive production of all kinds was carried on by large crews of people harnessing steam power, rather than by individual farmers plowing behind teams of horses.
Fourth, invention became central to economic production in the newly industrializing countries, as it had been a thousand years before in the Song; we can think of an invention as a skill that has been digitized, expressed in words or pictures so that it can be copied, rather than having to be learned by practice. New notions of precision, innovation, specification, tolerances, and standardization became central to work in a way entirely foreign to previous generations. New notions of measurement and metaphors of clockwork universes allowed the implicit to become explicit, legible, and reproducible.
Fifth, the advent of the steam-engine, which permitted pumping water out of deep-shaft coal mines, which provided a vastly increased fuel supply.
No, the real real problem is that in while there is some behavior that is obviously $BAD and others that are obviously not $BAD, there's a large range of behavior for which it's difficult to tell whether it's $BAD or not.
Consider the criminal justice system. Some people are obviously guilty and others are obviously innocent. But in between, there are lots of situations where it's difficult to tell whether the person is guilty or not. Vow to be more "tough on crime", and innocent people spend years in jail (or worse, end up executed). Vow to protect the innocent, and lots of guilty people get away scot-free. And there are criminals who are very good at exploiting this uncertainty.
There was a very insightful essay I saw many years ago which I can't find now unfortunately; but the main point was this: In superhero comic books and movies, the real superpower is certainty. The good guys always know who the bad guys are; it's just a matter of defeating them. In the real world, we have plenty of power to defeat the bad guys; it's just not always clear who the bad guys are.
So take the example from TFA, where the investor thought male founder A would be a better CEO than female founder B. Implicit bias is a real thing, and has been proven in dozens of studies. (For instance, where people are asked to rate the qualifications of a range of CVs, where the gender of the name on the resume is randomized.) Does the investor think A is better than B because of implicit (or not-so-implicit) bias? Or is A genuinely a better fit than B? It's basically impossible to know; even the investor themself may not know.
In the past, things swung very heavily toward "let the guilty go free", which meant implicit bias was allowed to stand unchallenged (leading to more men in leadership, leading to more implicit bias). "Cancel culture" is an attempt to swing things the other way. But it falls victim to the "certainty superpower" delusion: they think they know who the actual bad guys are, and end up taking down innocent people in the process.
What's the solution? In some sense there is no solution: until we have an Oracle of All Truth which we can consult, we will always have uncertainty; which means either punishing the innocent, letting the guilty go free, or some mixture of both. The best thing we can do is honestly acknowledge the situation and try to balance things as best we can.
Completely wrong. In 1900 a greater percentage of women than men graduated high school. Furthermore, while I couldn't find graduation rates by race for 1900, school enrollment by race shows white children were less than twice as likely to be enrolled than black children, 55% and 30%. Far from equal but also far from "virtually only white males".
Reminder that every DAO is a self-administering bug bounty for all of the value under its control.
Reminder also that you don't have to "hack etherum"; there are plenty of spots more vulnerable than the blockchain itself at which value can be stolen.
(I would however be interested to know where all this stolen value ends up, and how well it can ultimately be laundered into the real world, or if this is more like driving a truck into an ATM that causes far more loss than is actually successfully stolen)
> One of the Polaroid ID-2’s most important design features was a “boost” button that when pressed would boost the flash exactly 42%. Polaroid advertised this spectial feature for general lighting pruposes. However, researchers and artists assert that the ID-2 camera was actually created for and catering specifically to South Africa’s policies of Apartheid., and boost button, The white minority South African government largely used this camera for dompas, or the passbooks, which helped to sustain the Apartheid regime via surveillance. This was partially because the device was portable, fast at taking and developing photos, and created difficult to forge images because of its powerful lamination. But the most compelling feature was that the boost button increased the flash’s intensity by the exact amount it took to account for the extra light absorbed by black skin: 42%.
> Polaroid claimed only 20% of the film they sold in South Africa ended up being used for passports and according to Polaroid in 191, only 65 systems were sold before sales were stopped, and none of those systems were sold to government agencies. However, the Polaroid Revolutionary Workers Movement countered that sales were still going through indirect channels. Polaroid continued lying from 1971 to 1978, claiming that they had ceased supplying materials to the regime, when in fact there was an elaborate shell game which allowed them to sell through a third party (Caulfield 2015).
In an adversarial legal system, no one represents "externalities".
Parts of Australia have a deer problem, but despite cultivating a reputation for deadly animals, most of it lacks large terrestrial predators, so introducing wolves is an even harder sell than in the US.
It also has prior historical experience with introducing a species to address a previously introduced species, which then becomes a problem itself: foxes for rabbits; cane toads for cane beetles. cue Simpsons
The more complex the process you use to automate tasks, the more difficult it is to troubleshoot and maintain, and the more impossible it is to inevitably replace parts of it with a new system. https://xkcd.com/1319/ is not just a comic, it's a truism.
I am basically a Perl developer by trade, and have been building and maintaining customized Linux distributions for large clusters of enterprise machines for years. I would still rather use shell scripts to maintain it all than Perl, or Python, or Ruby, or anything else, and would rather use a system of 'stupid' shell scripts than invest more time in another complicated configuration management scheme.
Why use shell? It forces you to think simpler, and it greatly encourages you to extend existing tools rather than create your own. Even when you do create your own tools with it, they can be incredibly simple and yet work together to manage any aspect of a system at all. And of course, anyone can maintain it [especially non-developers].
As an example of how incredibly dumb it can be to reinvent the wheel, i've worked for a company that wanted a tool that could automate any task, and that anyone could use. They ended up writing a large, clunky program with a custom configuration format and lots of specific functions for specific tasks. It came to the point where if I needed to get something done I would avoid it and just write expect scripts, because expect was simpler. Could the proprietary program have been made as simple as expect? Sure! But what the hell would be the point of creating and maintaining something that is already done better in an existing ages-old tool?
That said, there are certain tasks i'd rather leave to a robust configuration management system (of which there are very few in the open source world [if any] that contain all the functionality you need in a large org). But it would be quite begrudgingly. The amount of times i've ripped out my hair trying to get the thing to do what I wanted it to do while in a time and resource crunch is not something i'd like to revisit.
Fun one to track down. Had to hop through a couple papers and books, but this is what everything winds up referencing:
> It was an organization designed not only to disseminate information and to clarify issues but also to arouse support for particular symbols and ideas. "The easiest way to inject a propaganda idea into most people's minds," said Davis, "is to let it go in through the medium of an entertainment picture when they do not realize that they are being propagandized."
Koppes, C. and Black, G., 1977. What to Show the World: The Office of War Information and Hollywood, 1942-1945. The Journal of American History, 64(1), p.2, <https://www.jstor.org/stable/1888275>
That paper then points to this ultimate source:
> Elmer Davis to Byron Price, Jan. 27, 1943, Box 3, Records of the Office of War Information, RG 208 (Federal Records Center, Suitland, Md.)
Dead end. I emailed the Archives but the entire facility is closed because of the pandemic. I also tried tracking down the paper's authors but both are retired.
For those that would like to know how to rebuild civilization, the book The Knowledge by Lewis Dartnell is a handy reference on things to know and do. General bibliography:
Deep dives on particular topics of agriculture, food and clothing, substances (charcoal, lime), materials (wood, cement, furnaces, glass), medicine, etc:
>Around the same time, from approximately 2008 to 2018, at the Named Sacklers’
request, billions of dollars were transferred out of Purdue as cash distributions of profits and
transfers of assets into Sackler family holding companies and trusts. Certain of these
distributions and transfers were made with the intent to hinder future creditors and/or were
otherwise voidable as fraudulent transfers.
Edit: Just going through this document makes me wonder how people can carry water for these demons
>The same day, David Sackler replied-all by email:
[W]hat do you think is going on in all of these courtrooms right now?
We’re rich? For how long? Until which suits get through to the family?
>In or about April 2008, Richard Sackler wrote a memorandum to Kathe Sackler,
Ilene Sackler, David Sackler, Jonathan Sackler, and Mortimer D. A. Sackler in which he
discussed limiting the Sackler family’s risk in the ownership of Purdue: “[T]he most certain way
for the owners to diversify their risk is to distribute more free cash flow so they can purchase
diversifying assets.”
These aren't people acting in ignorance, this is family essentially running a legalized drug cartel and has today just protected all their assets. I'm sure the negative externalities of these individuals will cause damage for decades to come in America's heartland.
Ben Shneiderman developed "TIES" aka "HyperTIES" at the University of Maryland Human Computer Interaction Lab, and here's what he recently wrote in response to a question about hyperlinks, in which he mentioned the origin of blue as a highlighting color.
Also here's a link to an article about the NeWS version of HyperTIES that we developed at HCIL, and some demos of HyperTIES and its Emacs based authoring tool, which had pie menus and embedded interactive PostScript "applets" in 1988.
John Gilmore via Internet-history <internet-history@elists.isoc.org>
Date: Mon, Apr 13, 2020, 11:56 PM
To: Brian, internet-history, Jeff
I forwarded this question to my friend Don Hopkins, who was a student of
Ben Shneiderman back in the day. Ben ultimately responded:
From: Ben Shneiderman <ben@cs.umd.edu>
To: Don Hopkins <don@donhopkins.com>
CC: John Gilmore <gnu@toad.com>, Ben Shneiderman <ben@cs.umd.edu>
Subject: RE: [ih] origins of the term "hyperlink"
Date: Mon, 13 Apr 2020 15:15:52 +0000
HI Don (and Jack Gilmore),
Thanks for including me in this conversation.
I do not have a claim for the term “hyperlinks” and don’t know when it
came into use. My claim is for the visual interface for showing
highlighted selectable links embedded in paragraphs. This is what we
called embedded menu items in that I think is an influential paper on
the topic, which was peer-reviewed and published in the CACM in April
1986.
While Engelbart had shown a list that could be selected by pointing
and clicking in 1968, I claim the idea of embedded highlighted
selectable text in paragraphs. This was implemented by grad student
Daniel Ostroff and described in:
Ewing J, Mehrabanzad S, Sheck S, Ostroff D and Shneiderman B (1986),
"An experimental comparison of a mouse and arrow-jump keys for an
interactive encyclopedia", International Journal of Man-Machine
Studies, Jan., 1986, Vol 24, pp. 29-45.
[Abstract] [BibTeX] [DOI]
Ostroff D and Shneiderman B (1988), "Selection devices for users of
an electronic encyclopedia: an empirical comparison of four
possibilities", Information Processing and Management, Nov., 1988,
Vol 24(6), pp. 665-680.
[Abstract] [BibTeX] [DOI]
I think the 1988 paper was the earlier study, but the publication took
a while.
My students conducted more than a dozen experiments (unpublished) on
different ways of highlighting and selection using current screens,
e.g. green screens only permitted, bold, underscore, blinking, and I
think italic(???). When we had a color screen we tried different color
highlighted links. While red made the links easier to spot, user
comprehension and recollection of the content declined. We chose the
light blue, which Tim adopted.
His systems with embedded menus (or hot spots), where a significant
user interface improvement over early systems such as Gopher. But Tim
told me at the time that he was influenced by our design as he saw it
in the Hypertext on Hypertext project that we used Hyperties to build
for the July 1988 CACM that held the articles from the July 1987
Hypertext conference at the University of North Carolina. The ACM sold
4000 copies of our Hypertext on Hypertext disks.
So in summary, I don’t know who coined hypertext, but I do think our
work visual and interaction design was influential.
Our Hyperties system was picked up by Cognetics Corporation (around
1987) who made a modestly successful commercial run with it, doing
dozens of corporate projects, most notably the Hewlett-Packard user
manual for their Laserjet 4 was distributed as a Hyperties disk.
Hyperties was the name we shifted to after we got a stop and desist
order from a lawyer because our TIES (The Interactive Encyclopedia
System) conflicted with an existing product. By then “hyper” was a
growing term.
Let me know if this helps, and what other questions you have…. Ben
Optimism is essentially the religion of Capitalism. This is because Capitalism is perpetually in debt to tomorrow. Every start up being discussed on HN, every fund people pump their 401k into, every new oil well being dug up, all of these require believing in the future.
The problem is that everything we have today depends on everything being better tomorrow. Optimism isn't just a nice feeling it's a core ideology that is necessary to keep the whole machine moving forward. If people en masse started to believe tomorrow might not be better than today, and that in the day after tomorrow might be even worse... the faith in our entire system starts to collapse.
People here talk about pessimism as if it's some rampant belief in society and the few optimists there are fighting against the hordes of the non-believer. But true pessimism remains a radical belief, outside of internet forums I have only occasionally met anyone who exhibits even the most mild form of pessimism. Even in academia the core works of most German philosophical pessimists remain untranslated!
Optimists also believe that someone questioning optimism it itself dangerous (which should be the first clue that something is not quite right with the dominant optimist world view). Whereas in practice it is optimism that allows us to perpetuate horrors on every scale without question because "tomorrow it will be better!". Ecological disaster is permitted because we will solve this tomorrow and everything will be okay. Sweatshops, child labor and modern slavery are all permitted because this is just a step in peoples inevitable rises to a better life. The murder of countless civilians around the globe are all justifiable because tomorrow we will have a better world economy, and democracy will spread through the region, making everything better.
Under optimism the shock from any horror is neutered and any atrocity can be trivially excused because the future is bright and we are heading in the right direction. We continue to burn every more fossil fuels because we are going to be fine, we'll figure it out, no reason for despair.
As the final insult to injury, questioning that future becomes heresy and so you cannot even voice your anguish as the world around you starts to collapse.
There was an "ask HN" a week or so ago about knowing how your personal machine was or was not compromised ...
I am imagining training a model on all of the PDFs and email and receipts and contracts that I have spent the last 30 years protecting ... and then exploding my home directory with 2M parallel copies.
Now what ?
Which is the real correspondence ? Which contracts are real ? Which invoices and receipts ?
Could one construct a framework where all data at rest is worthless and only witnessed transactions and traffic have any surveillance value ?