Brewster Kahle's Blog

There is a recent written attack on libraries that I find odd and somewhat dangerous– libraries overlap too much with new-book bookstores. At first I thought it was trolling, but I now believe it is sincere. More worrisome is that anti-library lobbying is generally growing in strength but their attacks on the concept of libraries have been more implied and disguised, until recently. Maybe it is time to think a bit about what a library is and what they are for.

First of all, there are lots of types of bookstores and lots of types of libraries, so a short post will not be comprehensive, but hopefully illustrative.

I love bookstores, and I am fond of online and offline bookstores, used bookstores, quirky bookstores, bookstores of bestsellers. Basically, I love books, I buy lots of them. Lots.

I love the book arts, where the form of the book is the art. My wife, Mary Austin, helped start the San Francisco Center for the Book. If you are unfamiliar with book arts, it is a treat– I suggest you dive in and create your book.

And I love libraries– academic and public, subscription libraries (yes, there are 21 of them in the United States– you should join one or many), digital libraries and special libraries. Ever been in a “special library”? Yes it is a thing. They are really fun, for instance the Linda Hall Library in Kansas City, Missouri.

A bookseller of new books wants their customer to find the right new book to buy, bring home, and read.

A library is all about context, a library helps their patrons come up with new ideas and connections whether by reading a whole book or a few pages of lots of books– new and old. Libraries offer historical newspapers and new newspapers, and from distant lands and opposing points of view. And libraries have experts on tap: reference librarians.

To do all this, librarians work hard at “information about information”, or metadata, to provide context. Libraries strive for comprehensive collections in their areas of expertise not just the popular– they keep old editions because they can be important context. Libraries typically offer many types of media: books, recordings, moving images, archives. 

In this way, a library helps people come up with ideas and connections, and if the ideas and connections are novel, and then get recorded, then they can be added to the library. Libraries help readers, many of whom become writers, whose works can then be made accessible, forever, in libraries.  The great circle of knowledge creation and preservation.

As Jim Gray said “Libraries are engines of research.” They are not the research, they power the research. To research something, a person does not need to buy all the books on the subject, new and old– rather, there is a more efficient way– they can go to the library.

Booksellers offer books for sale, and thank goodness books are still for sale (don’t get me started on restrictive licenses on ebooks).

Librarians offer access and preservation.

Both are valuable, both are necessary, and they are different.

-brewster
Digital Librarian
Internet Archive

San Francisco is best when it is inventing utopias (executing on it sometimes goes astray, but stay with me).

In our Covid/Pandemic world we are all pushed into virtual classes, virtual work, virtual relations.  Lets face it, we can only stretch virtual so far.

Let’s invent a physical world, and a desirable fabulous sharing wonderful world, in the age of Pandemic.   Together.

Burning Man is cancelled this year.  But it was an inspiration– I went in 1990, the first year in the desert– 80 people. It was called Zone Trip #4 in the paragraph in the Rough Draft physical newsletter– show up, be it, no spectators.  No tickets, no city planning, no directions, no rangers, no sherpas.   Make it happen, we were it.   It was so great that Mary and I invited our friends there to be our wedding in 1992 when there were 400 people and getting married at Burning Man was not a thing, nor would it be for years.

Ephemerisle has same “lets invent the world we want to live in” feel– 400 people on the delta trying to stay afloat, literally.  Inventive, dangerous, and wonderful.

Both early Burning Man, and Ephemerisle are all about adapting to a difficult environment, and with panache.

So…   how about Zone Trip #5: Pandemic?   In this case we start with a large chunk of desert, and then we invent a temporary utopia that works in the age of Pandemic– staying safe and loving it.  

With lots of inventors.   Everyone an inventor.

• Maybe someone brings a traincar of rice, a truckload of beans.   
• Maybe someone figures out what a school is, what a school means, what a school teaches.   
• Maybe there is a glass-cube restaurant.
• Maybe pandemic fashion takes a positive turn– a full-length clear isolation tube-dress 3ft in diameter.
• Romancing while social distancing?
• Virus free camp.
• Recovered camp.

What could you add to the world?

You should try this game.

I just invented and played a game today with my family, and it worked great– it was fun and we learned an important lesson. Similar to the game “Monopoly” which shows that one person ends up monopolizing even though you think the rules are “fair,” the game of Oligarchy shows that the “free market” leads inexorably to one person getting all the money and everyone else going broke. And fast.

The reason this is important is that it disproves one basic tenet of the free-market idea– that it is a game with many winners. Rather the free market, without redistribution, puts everyone (except one) in debt bondage, and quickly. We joked that those that were run out of money had to sell organs. Our game broke up into social classes– it was not worth it to the rich to play with the poor. It was all very real for a simple game.

All of this is based on a paper in Scientific American, Is Inequality Inevitable? Wealth naturally trickles up in free-market economies, model suggests, which has been a mind-blower for me: it is so simple and disproves the Libertarian premise.

We played 4 rounds with 6 people in about an hour:

• each gets a pad of paper and pen and writes $100 at the top, that is their pot of money
• each gets a coin to flip

Then each player picks another player agrees to gamble on a flip of a coin, they agree on which wins on heads, and tosses. The stake of the gamble is set at 50% of the lesser of the pots of that pair. So in the beginning, when everyone has $100, the gamble is $50.

Then the players pick another player (or the same player) to do another round. This proceeds.

What is amazing is that even through each toss is “fair” in that it is a 50-50 chance to win a straight amount of money, the results shows one player wins all the money, and really quickly.

Two nephews and their partners, Mary and I played 4 rounds in about an hour and we discovered social classes (we called the broke ones “organ sellers”), feeling of righteous empowerment based on being successful (even though it was completely random), but also that “free market” ended with all-but-one-of-us in a bad situation really quickly.

Try it, it is fun. And read the article, it is startling– free-market without redistribution goes to Oligarchy very very fast. This book on Sumerian and Babylonian economics shows it has always been this way, so people developed peaceful reset mechanisms with debt forgiveness and Julilee: …and Forgive Them Their Debts by Michael Hudson.

[And now you can see it played automatically! http://hackerfactor.com/oligarchy-game.php Go Neal!]

A Washington Post article talked about getting to interstellar rocketry by slingshotting around the Sun. I understood the idea of slingshotting required a planet or a moon– something that was orbiting something else because you are taking some of that acceleration of the planet for your rocket. Indeed this is called “Gravity Assist,” so what is this slingshotting around the Sun?  Well, this is where you get the Oberth Effect. The idea is you go really fast by getting close to the sun, burning your fuel, and you will go much faster even after you slow down again having left the proximity of the Sun.

But after reading and rereading the wikipedia article, talking with my mates, watching a youtube video I still did not get an intuitive grasp of it– in fact it seems to break the rule of conservation of energy. How could blasting your fuel while going faster make you go faster?  I think I have an approach to see how it makes sense, in general, but does not bring us to equations.

A rocket works by throwing exhaust out the back at high speed, thus pushing the rocket forward. For a given amount of fuel, you throw it out (relatative to the rocket) at a certain speed and certain weight to become the exhaust thus pushing the rocket forward with a certain amount of force.

Let’s say the fuel is thrown out the back at a certain speed, V. If you do this while standing still, the fuel goes one way at speed almost V and the rocket goes the other. If the rocket is moving at speed V, then throws the burnt fuel out the back (say it does it all at once for simplicity), then exhaust will be standing still and the rocket moving faster than it was.

The first case is like in a swimming pool you push against a floaty to go one way– the floaty goes the other way. The second case can be seen as you are in a swimming pool and you push just as hard, but you are pushing against the wall. The wall goes nowhere, and you go faster across the pool.

You would push against a wall if you could. And by speeding up the rocket you are effectively doing that. Same force, but the floaty goes the other way vs the wall which stays still. When you push against the wall you go faster in the direction you want to go.

In the rocket case, we have to get the rocket to go the speed of the exhaust V, and not take any energy to do that. The way the rocket is speed up is by “borrowing speed” by accelerating towards the sun. The sun accelerates both the rocket and the fuel it will burn equally– this is that tricky thing in high school physics where a feather and a bowling ball fall at the same rate (if there is no air resistance). So the sun pulls both in, the rocket and fuel are going really fast, say V, then the rocket blasts out the rocket fuel leaving it behind, and then decelerates as it leaves the sun. If the rocket got up to speed V, then the net effect is that instead of the rocket throwing its exhaust backwards, it would go nowhere, thus being the wall it pushed against.

The rocket decelerates as it leaves because of the gravity of the sun is pulling it back, but it is getting less force because it weighs less because it burned fuel. Again, this is that tricky bit about the feather and bowling ball. So the Sun applied more force to the rocket+fuel on the way in than on just the rocket on the way out. Another way of seeing how this effect works.

At least this makes sense to me. The effect is caused by the sun accelerating the rocket+fuel, then when the rocket blasts out its spent fuel, the exhaust is not going in the opposite direction as much, thereby offering more of a push to the rocket, even after it is decelerated.

Going to the airport this morning, I took a Lyft for $46. It us usually less, but it was rush hour, and still less than a taxi (usually $65) and more reliable than a taxi. The driver was cheerful, been driving for Lyft/Uber for a year and a half, and it was in a Chevy Bolt, so we were driving an electric car. I got there before it predicted, and all was good.

But I looked at the driver’s phone after we got out and it showed she made $23. Yikes. This does not seem fair.

The driver said that after Uber and Lyft went public, the “take” they were taking from the drivers went up.

50% is too high a fee for Lyft. The only higher fees I know are what book publishers take from authors.

In the old days, paying a highway toll meant I paid with cash often a tossing a quarter in a basket or smiling at a toll-taker and handing over a dollar. But now I get to speed through.

To do this, do I really need to have my license plate photographed? Do I need to hide my FasTrak in a conductive bag to keep from getting tracked other places on the highway without my permission? Do I need to have my every bridge crossing be logged into a database somewhere?

As a thought experiment: how do we high toll takers that are automated but not privacy invasive?   If we can figure this out, can we go on to make it happen?

Features I want, and maybe all of us want:

1. Speed through toll booths and paying the right toll
2. Able to deal with cars that have a device and ones don’t
3. Easy to refill the device with money as needed
4. Not be tracked any more than dropping a quarter in a basket (not creepy)

Goal 2 means that if the toll does not detect the car’s device, then we need to track it somehow and bill the owner. Currently this is done by photographing the license plate and sending a bill.  I do not have a way around this, do you

But if the car has a transponder then the car can put out a signal to pay the toll. I think with crypto-currencies, especially zcash, we have all the technology we need.

Basically the device would hold a key to the money I had transferred to it. Then the device would recieve and transmit a bunch of numbers to the toll as it was passing through, just as the current FasTrak does, but instead of saying the account number in the FasTrak, it would offer to pay the required toll by transferring the money into the government’s account. This could be done quickly, inexpensively, securely, and privately– all with technology that is currently here and open source.

Does this do the trick?

We need a low transaction cost billing system because tolls are low so the fee should be a small percentage of the toll. Currently it is 3% to 5% cost because of the visa card transaction to refill my FasTrack account. Bitcoin is often 7 cents per transaction, and zcash is much less than a penny. So current Bitcoin would work, but other systems are a better fit for this. But this works: check.

If the toll booth could detect if the payment was good and valid quickly enough then it would not need to take a picture in those cases. If the transaction is too slow for this (like current bitcoin), then the picture might have to be taken, but can be deleted when the transaction is complete, say in 30 minutes.

In terms of putting money into the account, this can be done from dollars, or credit cards, or crypto in any of a large number of proven wallets and websites. So that seems easy.

So cryptocurrency, bitcoin, etherium, or zcash can make a highway toll-takers more efficient than they are now.

Wouldn’t it be great if an early widespread use of crypto was to make our highways more efficient and less creepy?

Same product, same vendor, to same address, same day, same shipping, same billing address, but different customer, different price $8.63 vs $12.37. 43% higher price for me than for the Internet Archive. In fact I checked the price for me an hour before and the day after the Internet Archive bought it, so I do not think it is the vendor changing the price by this much.

Below are the receipt for the Internet Archive, offer to me, and offer to my home business account (same price as the Internet Archive). Turns out I stumbled upon “Business Pricing.” It’s a “feature.” Business Prime costs $179/year, as opposed to consumer Prime for $99/year.

Yikes. Shakes my confidence, and seems like we could be on our way to “redlining.”

What price to do you see for https://www.amazon.com/gp/product/B0089PBTV2 ?

I am wondering if the Internet Archive, and tech orgs more generally, could effectively hire teams in addition to hiring individuals.  I am thinking of 4 or 5 people that could work effectively with each other, be remote from HQ, but be really “hired” in the sense that they assimilate and integrate with the rest of the organization.  I have not tried this, so this is just random thoughts at this point.

Acquisitions usually fail, they say, and I have seen failure– culture mismatch, too much redundancy, so lots of loss.  The new term “acquihire” is interesting, so people are trying to hire functional groups, but so much is loss

What if a group defines themselves as a team and goes out to be hired.  They interview as if they were an individual– talking to managers in an organization to find out if there is skill, temperament, salary match etc.  If the team is hired, then it is trained by the org, and such, and then operates as a “super individual” for the first year or so. Built into the team can be different skills, maybe: communicator/manager, tech lead, prodmgr/testing, senior and junior dev.

What problem am I trying to solve?  Well there are several:

How to grow a tech company to meet a sudden opportunity: so you get funding, a grant, increased revenue– how do you grow?   Some big companies acquire other companies, but those teams often do not integrate well, take a long time to deal with and are expensive.  Most organizations hire individuals, maybe this would be effective step up from this system.

How to be effective remote workers:  San Francisco is a tough place to move to, but what if a team operating in another area or another country but then went to job hunt in San Francisco, and did some training there, but then mostly lived and worked in their home location. They could learn from each other to bring up younger workers and can recruit to replace people in their team. To integrate with other people and teams in the organization they should switch people’s projects otherwise silos could solidify.  This team would then really work for the organization, not just for the team.

How to teach younger workers in a remote workforce: I learned by watching and working closely with others– I think we all do. Doing that over a network can only go so far. Teams can have junior and senior people in it, people that know how to work together well and have the bonds to bring each other up.

The communications issues can ease management issues with remote workers: some people are better coders than communicators, but good communications is essential in remote work, and in modern tech work in general.  What if the team had a mix of different skill sets?

Robustness, resilience:  each team can handle their recruiting, training, and rejuvenation– each job function could be shared between people, so if any one leaves, there are those that can fill in. Key is that the team would not pick up and leave all together otherwise the organization is super stuck.  If longevity with the organization can be counted on, and the company resists the temptation to fire a full team at once, then this can be a big advantage.

Team formation can be a local phenomenon— maybe a team forms in hacker spaces, around schools, or in companies that are downsizing and then they market themselves as a unit. Maybe teams are created in outsourcing company somewhere.  It is possible that a team in a remote country could cost the same as an engineer in San Francisco.

—-
The Internet Archive has successful teams working as scanning centers around the world, so it works for this task.  Can it work for tech groups? Don’t know but maybe we should try.

Have you tried anything like this?