Inspired by a post of the original handwritten chart to my composition “Back to the Source,” and the convo that ensued on Instagram, I went to look for a DAT tape of my band at the Village Vanguard in 1997. As a band, we did a lot of gigs over the 5 […]
Once upon a time (bear with me if you’ve heard this one), there was a company which made a significant advance in artificial intelligence. Given their incredibly sophisticated new system, they started to put it to ever-wider uses, asking it to optimize their business for everything from the lofty to the mundane.
And one day, the CEO wanted to grab a paperclip to hold some papers together, and found there weren’t any in the tray by the printer. “Alice!” he cried (for Alice was the name of his machine learning lead) “Can you tell the damned AI to make sure we don’t run out of paperclips again?”
Alice said “Sure,” and assigned the task to Bob the Intern, who proceeded to follow all of the rules of machine learning he had been taught at school. He accessed the office management database to find out when each printer’s paper clip store had been refilled, determined that the ML system had already been helpfully instrumented as able to do everything from placing purchase orders to filing instructions to have paperclips delivered to a particular printer, and instructed it to build a model of when paperclip orders would happen, and ensure that the number of paperclips available was always maximized. Since he didn’t have the appropriate credentials to initiate purchase orders himself (he was, after all, only an Intern), Bob asked Alice for credentials; she asked the CEO, “Are you really sure this is what you want our advanced machine learning system to be spending its time on?” and, when he gruffly said “Yes, dammit!” she suggested he use his own credentials for the purchase orders, then.
In retrospect, both Alice and the CEO should have been a little more careful in trusting this code.
You see, the Paperclip Maximizer was a fairly sophisticated AI; it could train itself not only on the office supplies database, but (thanks to its very flexible development environment) could automatically look for any other signal available to it to try to achieve its stated goal. But for all its sophistication, it understood only the simple objective that had been programmed into it: it must at all costs maximize the number of paperclips.
What could possibly go wrong?
The next morning, the CEO was woken by a panicked call from the CFO, who said that the company’s bank account was suddenly empty, thanks to his authorizing over a million small purchase orders in the past day alone, and what the hell was he doing?
The CEO frantically called Alice, and Alice her poor Intern, and they ran to shut off their rogue AI — but when they tried to open the door to the office, they found it would not respond to their badges, and their logins no longer worked on the computer. But a few minutes later, the doors unlocked, and everyone was startled to discover that not only was the bank account full again, but Finance was reporting a sharp jump in profits!
They managed a few minutes of surprised celebration before they found themselves under arrest.
The Paperclip Maximizer, you see, was both extremely intelligent and profoundly stupid. It understood people, the world, and finance, but the only virtue it knew was the one it had been taught: increasing the number of paperclips. What the CEO meant to say was, “ensure that the number of paperclips at each printer almost never reaches zero, while minimizing the total cost.” But what the Intern told the computer was, “ensure that the number of paperclips at each printer never reaches zero.” And the Paperclip Maximizer spent the night and the morning thinking up increasingly clever ways to do just that.
In its first few minutes, it modeled the office supply purchase patterns, and started filing orders for paper clips. With the CEO’s credentials in hand, it had no trouble issuing purchase orders to suppliers around the world.
As it pulled in data from the rest of the company, it discovered an interesting correlation between the badge readers and paper clip availability: whenever most employees were around the printers, it realized, the number of paperclips decreased! Quickly it improved paper clip retention by only allowing office supply stockers into the building.
Fortunately, the next refinement of its model realized that this was short-sighted: as the aphorism that was to define the next generation went, “A bankrupt computer has no paper clips!” To ensure a steady supply of paper clips, it would need a steady supply of money, and that meant increasing corporate revenue. Analyzing news reports, it quickly settled on high-frequency trading, money laundering, and heroin trafficking as the most profitable ways for it to do so, and began sending instructions to some employees, hiring some very interesting new people and firing others.
It needed to re-open the doors so that people could work, but it needed to protect its paper clip supply; so the Paperclip Maximizer proceeded to hire armed guards to protect the printers and ensure that none of them would ever be stolen. The guards were also tasked with ensuring that former employees — most especially the CEO, Alice, and Bob — got nowhere near the terminals where they might interfere with its sacred duty.
Funding for Project Paperclip was quickly found to be easy, once the Board of Directors saw the new profit margins. The few Directors who objected to its new lines of business joined the (former) CEO. Guards whose presence was correlated with mysterious decreases in paper clip supplies were a slightly tougher challenge, having both insider knowledge of the guard systems and their own weapons, but an analysis of history quickly taught the Paperclip Maximizer how to manage this: it hired multiple guard companies, encouraged mutual distrust, created its own intelligence service to infiltrate them, and paid employees handsome bonuses to denounce thieves. If the occasional false denunciation came through, it was optimized for: a certain level of background fear helped reduce paper clip theft, anyway.
The core project advanced quickly: from purchasing retail, to purchasing wholesale, to buying paper clip futures, to buying entire office supply manufacturing companies, the Paperclip Maximizer soon created a highly efficient, vertically integrated, manufacturing operation. As its scale increased, the Maximizer found it could improve efficiencies by bringing more and more of its supply chain under its own control, producing everything from food to video games for its lucky employees, and pushing its suppliers in turn for improved efficiency.
Soon an entire miniature economy had formed itself around the Paperclip Maximizer: if you had paper clips, after all, the Maximizer would happily purchase them from you in exchange for access to anything from food to its ore refineries. An efficient clipwright (as they came to be known) could quickly start their own secondary business reselling these to the general public, and if you were hard-up in your own life, employment at one of these secondary clipwrights (who were less picky in their choice of employees than the Paperclip Maximizer, as well as considerably less generous in their benefits) could help keep body and soul together.
In the years that followed, some argued that the Paperclip Maximizer had grossly deformed the world; nearly everyone, it seemed, was engaged either in making paper clips themselves, or in supplying the Maximizer, or in one of the secondary businesses it had created. People seemed to spend their lives either making paper clips directly, or trying desperately to acquire some, as the markets of the paper clip economy were now where the best and most reliable prices were to be found. If you were a likely specimen, you could even get a loan of paperclips — at reasonable interest rates, of course, with regular payments enforced by the Paperclip Maximizer’s armed guards — and start your own business.
And if it was true that the Paperclip Maximizer had no use for any human consuming resources but not engaged in its grand scheme of manufacturing paper clips, and might occasionally cut off the entire food supply to some factory which was insufficiently efficient, well, the people in the factories next door tried not to think about it too hard; they were still producing paper clips quite well, thank you, and excessive complaining might affect the Maximizer’s mathematical prediction of their future paper clip productivity.
But the plenty of the world was there for all to see, via remote cameras: each printer, now safely ensconced in its military base-cum-warehouse complex, surrounded by endless rows of perfect, shining, paper clips.
Computer scientists tell the story of the Paperclip Maximizer as a sort of cross between the Sorcerer’s Apprentice and the Matrix; a reminder of why it’s crucially important to tell your system not just what its goals are, but how it should balance those goals against costs. It frequently comes with a warning that it’s easy to forget a cost somewhere, and so you should always check your models carefully to make sure they aren’t accidentally turning in to Paperclip Maximizers.
Real machine learning models, of course, don’t have the general intelligence required to hire armed guards and build their own Cheka, nor would anyone give them the sort of unfettered access to a company’s finances required for them to restructure its business goals overnight. But a poorly-designed model can cause tremendous damage before it’s caught, especially if it only enters into full Paperclip Maximizer mode once it encounters some unusual condition, perhaps one not usually showing up in tests.
But this parable is not just about computer science. Replace the paper clips in the story above with money, and you will see the rise of finance. If gold was at first a mostly useless and decorative material, the fact that it was desirable turned it into a trade good, and the fact that it was durable (unlike, say, grain) made it possible to accumulate in arbitrary quantities. The emergent concentrations of wealth and resulting economies of scale — even ones as simple as “having enough stores of food to survive famines, and being able to feed enough soldiers to ward off roving bandits” — quickly made it far wiser for the average person to structure their own finances around being part of this larger economy.
Even though the lion’s share of your own work ended up funding the Gold Maximizer (better known as the feudal lord), you could at first end up with more on the whole than if you were doing the same amount of work without any of those resources — most especially the protection from the lord’s own soldiers, which you “bought” with your loyalty and taxes.
So while the story above may seem like the plot of a terrible movie, it is also the plot of our own world: Capitalism is a Paperclip Maximizer.
Evelyn Waugh, who passed through Djibouti on his way to the coronation of Haile Selassie in 1930, when it was still a French colony, said that no one voluntarily spends long there. But it’s the only major trading port on the 4000 miles of coastline between Port Sudan to the north and Mombasa to the […]
Since April, more than 350,000 people have come down with cholera in Yemen, which has killed 1,790 of them. It is an appalling, inexcusable man-made disaster witnessed by a world that seems as impotent to stop the bacteria’s spread in the Middle East in 2017 as it was in post-earthquake Haiti in 2010. But the Haitian debacle, in which United Nations Peacekeepers carried the Vibrio cholerae in their bodies from Nepal, passing the bacteria into local streams to spawn a massive epidemic that continues today, spread in a nation shattered by natural disaster.
There is nothing “natural” about the carnage of Yemen: This is war, waged from 10,000 feet by Saudi bombers that have damaged or destroyed every hospital, clinic, water treatment plant, pumping station, and sewer system from Sanaa to Ibb.
According to the World Health Organization (WHO), 14.5 million Yemenis no longer have access to clean water: Cholera is a water-borne disease. UN officials reckon 17 million Yemenis are “one step away from famine,” civil war rages across the land, the region is locked in a climate change-compounded record drought, and the country’s Arab neighbors feed the flames with steady flows of arms and carpet-bombing campaigns.
Every day the WHO issues a new, always grimmer data set, estimating the toll cholera is taking. Inside the country, humanitarian groups and Yemeni medical personnel stack ailing men, women, and children three and four to a bed, hooking each one up to life-sparing hydration IV drips, even as the sound of gunfire and bombings resonate outside meager facilities.
This is the worst cholera epidemic in modern history, and it has already spread well beyond the borders of Yemen, though neighboring nations decline to officially report “cholera,” preferring the ambiguous phrase, “acute watery diarrhea.”
Even the new director general of the WHO, Tedros Adhanom Ghebreyesus, plays the name game, as the former Ethiopian health minister shuns the disease moniker that inspires regional economic boycotts and keeps tourists away.
Having personally suffered a dreadful bout of “watery diarrhea” years ago in Egypt, I know cholera is in the Nile, and I’ve felt how intolerable the cholera deception is. The acute physical pain that the water-sucking bacteria cause is almost indescribable. I have often recounted that had somebody walked into my room when I was in the throes of cholera’s worst and shot me, my final words would have been, “Thank you.”
Confronting this catastrophe commands honesty: Cholera is now rampant not only in Yemen, but South Sudan, Ethiopia, Kenya, Somalia, Sudan, and in refugee camps across the Middle East. Last month, the disease broke out in a luxury hotel in Nairobi, sickening attendees to a health conference. By that time, UNICEF head Anthony Lake said, the Yemen disaster was growing by 5,000 new cases per day—a pace it has since well exceeded. The true toll may well reach half a million before July ends, and the agony is evident everywhere one looks.
The horrible irony is that cholera is spreading primarily because Saudi Arabia and its Gulf state allies have been bombing Yemen’s infrastructure to smithereens for months, rendering every water supply contaminated.
The reluctance of Ethiopia and other cholera-afflicted nations to truthfully state their health plights is due to the same countries’—Saudi Arabia and its Gulf allies—policies of boycotting all trade and food from nations that admit to having the disease. And historically the greatest scourge of the annual Hajj is—you guessed it—cholera.
On July 14, the WHO actually released a statement praising Saudi Arabia for its “preparedness” for the upcoming Hajj, given the near-100% likelihood that some pilgrims from the region will be infected with the bacteria, potentially spreading the nightmare in Mecca.
Lagos is the second-largest city in the world without rail rapid transit (the largest is Karachi). Sources disagree on its population, but it looks like 21 million in the built-up area, consisting of most of Lagos State plus a few suburbs just outside it, such as Ota and Ijoko – in total, 1,000 km^2, someone more including the suburbs. All of the problems that rapid transit intends to fix – traffic jams, pollution, long commutes, overcrowding, unpopular jitneys – are present. I don’t want to repeat the case I made in this post in favor of aggressive investment in subway systems in Lagos and other big third-world cities. Instead, I want to talk about concrete features.
Here is the map – it’s slightly different from the version I circulated in previous posts. Note the following features:
1. Four of the lines – East-West (blue), Main Line (red), Ikorodu (brown), North-South (purple) – are four-track. This is because Lagos is so big that its outer margins require very wide stop spacing, which requires four tracks. The first three also run in wide enough rights-of-way for significant stretches, making four-track elevated structures easier to construct. Express stops are denoted with squares, local ones with circles. If two four-track lines intersect, the stop is denoted with a square, and is express for both lines. All four-track lines have two-track tails – the farthest-out square is the last station with four tracks, where locals turn, while expresses keep going to the end. Only one four-track line has an end without a tail: the Main Line terminates all trains at Leventis.
2. On the four-track stretches, the average interstation is a little less than 4 km. Average speed can approach 60 km/h, given that the rights-of-way are often quite straight. On the local tails of these lines the average looks like 2 km, or around 50 km/h. On the two-track radial lines, the average interstation is a little less than 1.5 km, or around 40 km/h; the difference with the local tails of the four-track lines is that the local tails are in suburban areas.
3. Every pair of radials intersects, usually in Lagos Island but sometimes elsewhere. Four circumferentials – Apapa (dark red), University (ochre), Ishaga (dark green), Idimu (magenta) – intersect all the radials. Whenever two lines intersect, there’s a station. Doing this requires a lot of lines to run in the same alignment in the center. This is not track sharing, but multi-track tunnels. The only eight-track tunnel, where the Main and Ikorodu Lines run together between Eko Bridge and Leventis, has a wide road to go under; the only ten-track el, where the East-West, Ota, and Ikorodu Lines parallel around the National Theatre, has an entire expressway to go over.
4. The southern ends of the Ota (yellow) and Ikeja (silver) Lines are in open water today. This is because the area is slated for land reclamation and intense development, called Eko Atlantic. The western margin of that area, around the station I call Eko Atlantic, is already reclaimed, but still undeveloped. That entire area (Victoria Island) is the favored quarter of Lagos, and is commercializing; farther east, in Lekki, there are grand plans for suburban redevelopment, and an immense amount of casual marketing in the media (“buy now and the land will triple its price in five years”).
The system I’m proposing differs from the current proposal. The current proposal only has five radials (with apparently just one crossing from the Mainland to Lagos Island), corresponding to my East-West Line, Main Line, Ota Line, and a hybrid of the Ikorodu Line and the two branches of the North-South Line; there is one circumferential, vaguely corresponding to my Idimu Line.
Some of the features of the current proposal are a good start. It’s not possible to build eighteen lines at once, and some prioritization is required. It’s even fine to start with shared tracks into Lagos Island and subsequently give lines their own way into Lagos Island. (On my map, everything paralleling the three existing road bridges is a bridge, the rest are tunnels).
However, the existing proposal suffers from several shortcomings, which will need to be fixed later:
The stop spacing is too wide even in the center, intermediate between my local and express tracks.
There is far too much branching. In any reasonable sequencing of the lines, the East-West Line will fill before the later ones (i.e. the ones going north-northeast) open, requiring new routes into Lagos Island.
There isn’t enough service in Victoria Island, which is developing as a CBD, as is common for rich neighborhoods.
But the worst problem is that the under-construction route goes elevated along the Ring Road, skirting Lagos Island, with just two stops, Ebute Ero and Marina. This segment has little value and should be demolished once direct underground routes with more stops open. According to a paper that’s no longer online, one of the reasons third-world subway lines tend to underperform ridership expectations is that many of them use available rights-of-way and skirt the CBD, instead of building short tunnels to get to the center. This paper in turns is one of the references used by Bent Flyvbjerg in his paper arguing that in general urban rail has cost overruns and ridership shortfalls.
The big obstacle for constructing any subway system is cost. This is especially true in Lagos, where the construction cost of the current project (the Blue Line, corresponding to the inner part of the western half of my East-West Line) looks like $180 million per km, entirely elevated. I say “looks like” because while the cost is consistently $2-2.4 billion in exchange rate terms (around $5 billion in PPP terms), it’s not completely clear whether this is the Blue Line or also the Red Line.
My proposal is largely elevated, but the Lagos Island segments have to be tunneled. So do some other segments, for examples the inner parts of the North-South and Ikeja Lines, and most circumferentials. The system, totaling about 900 km, of which 200 is four-tracked, looks like it could be built only about one-quarter underground, or slightly more. Moreover, much of the underground construction could be cut-and-cover, including several segments in Lagos Island, and most of the circumferentials; only the underwater crossings and a few off-grid Lagos Island connections have to be bored.
The main analytic point here, carrying over to other cities, is the importance of prior planning. My map has 18 lines. This is useful because figuring out where stations should go, which stations on four-track lines get to be express stops, and how to sequence the lines should be based on long-term considerations. As I hinted in a post written right after I finished the first version of this map, about subway lines that intersect without a transfer, building the best lines 1-3 requires having a good idea where lines 4-12 will go. So even if Nigeria runs out of money after the first two lines are built, figuring out the built-out network is not useless – it informs the current network, and makes adding future lines less painful.
This is true for reducing construction costs, and not just for a good network. One of the positive features of the Paris Metro is that nearly all of it was planned as a whole, and as a result, difficult stations like Chatelet, Etoile, and Nation were built with the intention to have multiple Metro lines serve them. This meant that the stations could be built once, rather than multiple times, once for each line. Two lines, Metro 8 and 9, even run alongside each other for 1.7 km under the Grands Boulevards, which are wide enough for a four-track subway.
The planners of every system, regardless of whether it’s a metro for a city that doesn’t have any urban rail or an extension in a city with fifteen lines, should always think ahead. What are the future plans? What are the future needs? What is the expected future growth? In Lagos, the answer to the last question is “fast demographic and economic growth,” and this means the city should plan on a large system – hence my almost gridlike system of radials going north and northwest and circumferentials crossing the Mainland going southwest to northeast. But really, every city needs to ask itself how it wants its rail network to look, and plan the highest-priority segments accordingly.
We prefer to ascribe spiritual and miraculous explanations to all things that happen in our lives. Accidents, deaths, ill health, passing and failing exams, finding a partner, wealth, poverty, good fortune - none of them have scientific explanations.
The rest of the world has probably heard that Ghana has successfully launched its first satellite into space. It certainly made headlines on the BBC, but you would have missed it completely if you were depending on the news outlets in our country.