Alpha Takes All

This isn't a post about the adversarial neural network that was programmed to play a video game against itself until it took the Grandmaster Championship from the world's best (human) players. It's about a person who beat the computers, and took out a huge portion of the stock market in the process.

Come back with me for a moment, to the year 2010: Smart phones are just now in everybody's hand, the word "application" is just about to be replaced by the word "app," and an algorithm is still just an anagram of the word logarithm.

Wall Street, on the other hand, is just about to get rocked by an event called the Flash Crash of 2010. The market lost and then regained trillions of dollars in 30 minutes, due to a bunch of trading algorithms getting stuck in each other's code like a pack of leashed-up dogs trying to sniff each other's butts. It was described as one of the most turbulent periods in the history of financial markets, and for quite a bit after, nobody knew what the heck happened.

It was also one of the first alarms to be rung about the dangers of artificial intelligence.

"No place for a human" is how the stock market was now described after the takeover by high frequency trading algorithms.
-Jerry Adler, WIRED, 2012

That comment means something different all of the sudden, in light of the Boeing 737 Max's MCAS failure that killed almost 350 people in 2019.

But taking it back to the stock market, the story got a nice twist this year when we were reminded about the Hound of Hounslow. He's a human, but he doesn't exactly belong in the stockmarket either. By 2015, this self-taught quant was held on multiple counts of doing bad things to the stock market, and by 2020, he was given his (lenient) sentence.

Why wasn't he put in jail for life for causing a trillion-dollar market maelstrom? For one thing, he's helping authorities catch other HFT-wielding market manipulators.

The leniency also comes from the fact that he's running an Asperger cortex, and he saw the stock market as a video game, and like AlphaStar, he learned how to play that game really well, and how to "beat" his opponents.  He noticed that the other trading algorithms were doing strange things, and getting entrained into each other's code. He thought he could tap into that groupthink and orchestrate the whole lot, and he did, triggering a cascade of buy/sells that crashed the market.

Notes
Hounslow trader avoids jail in 'flash crash' case
Jan 2020, BBC News

Quantitative Financial Analyst

High Frequency Trading

Post Script
Circa 2010's commentary on high frequency trading:
Raging Bulls - How Wall Street Got Addicted to Light-Speed Trading
Aug 2012, WIRED

Keeping track of algorithms in the news:
Algos
Network Address, 2012

Deep Creep

Visualizing and Understanding Convolutional Networks -- This goes back to 2013 already, but I recently came across these images, and they are so mesmerizing I had to archive them.

The images pasted here are from a paper about convolutional neural networks. The researchers are able to train a network with tons of images, and then ask that network to classify new images it hasn't seen yet. Getting the detection error rate down to zero is the goal. This one does a good job.

But what makes this report special, is that we get to see how the system spits back what the different "neurons" see. The network develops layers or clusters that recognize different things; some are good at low-level features like lines and edges, and some are good at high level things like "bicycles" or "origami." Together they learn how to see.

^This is the first layer, it sees angles and colors.

^This is the second. This one's getting more complicated patterns. Notice the similarities, but also the differences. Of the 3x3 sets, which one is not like the other? The network saw all of those as similar. The network says that those images sit close together in image-space (the total space of all possible images, or at least all the images it was trained on).

It's a game to try and figure out the common denominator. We don't really know the common denominator, and we can't know, because we're just not computers. This is why they call it "deep." In the middle of this network, deep in there, is a layer with information that is just too complex for us. Collapse 33,000 dimensions into 3, and now try and communicate features of the 33,000 using only those 3 points of information. Can't do it. That's why it's mysterious.

^Now the third layer. Ok, so I see the people, I see the barcode/text motif, the honeycomb/diamond clusters, even the lady bug and the tomato -- it's a stretch, but I get it. But some of these are just nuts. White lower right corners? It sees white lower right corners?

^Fourth layer. No idea what this thing is talking about.

^Fifth layer. Here we go, people, dogs, flowers, now it's all making sense. And that was the point of creating this network -- you give it any picture of a flower, no matter how weird of a picture, barely looks like a flower, and this network will recognize it and classify it as a flower. (Mostly; it's not perfect.)

But there comes a point in the middle there, we have no idea what this thing is thinking about. And we never will. Just like other people, and how we can never really know what someone else is thinking. Which is to say that the computers are now a lot more like other people than they ever used to be. They have become mysterious.

Notes:
Zeiler and Fergus, Visualizing and Understanding Convolutional Networks, 2013 [ZFNet].
https://arxiv.org/abs/1311.2901

Not So Fast

Human cultural evolution found to be just as slow as biological evolution
Jan 2020, phys.org

I can't read the paper for paywall, so I'm copying Bob Yirka's review at phys.org:

To compare the rate at which human culture changes to rates of biological evolution, the researchers assigned variables to characteristics of several cultural artifacts—whether or not guitars were the major instrument in the average song, for example, or how car features such as size and power change over time, or the way references are tagged in scientific papers. Similar metrics for measuring the speed of evolutionary change have already been identified and measured by multiple scientific studies. The researchers chose to use some of the most well-known, such as the study of finches on the Galapagos Islands and moths changing color during the early industrial period in England in response to soot-covered tree bark.

The comparative analysis involved applying the Haldanes metric—it showed that human culture changed at very nearly the same pace as biological evolution. The researchers even suggest that cultural artifacts in a given society could be viewed as similar to organisms living in a given environment. Artifacts such as scientific papers, they note, when carried into society at large, either survive and become a part of the culture, or they die—just like natural selection. They acknowledge that there are instances in both cultural and biological evolution that change very quickly, such as smartphones or finch beaks, but overall, the rates come out nearly evenly.

I put this here to remind us that Kevin Kelly was saying the same thing in his 2010 book What Technology Wants, which said that Technology (as a cultural artifact) is an extension of the Tree of Life, and shares many similarities with a living organism under the pressures of natural selection.

image source: aero-mag

Notes
The pace of modern culture
Ben Lambert et al. Nature Human Behaviour (2020). DOI: 10.1038/s41562-019-0802-4

Neural-Interfaced Exoskeletons

Paralysed man moves in mind-reading exoskeleton
Oct 2019, BBC News

Progress marches on, this time wearing a neural interfaced exoskeleton.

This gentleman injured his spinal cord in a fall and lost the ability to walk. Scientists are letting him experiment with this pretty astonishing technology.

He first learned how to use the setup via an avatar; he would mentally control the video game character, making it move with only his thoughts of movement. Eventually he put on the actual suit and began making it move with his thoughts.

Caveats? The suit is attached to a ceiling harness, so he doesn't fall over. It's not that good yet. Also, this is super hi-tech, which means it's not available to pretty much anyone. And note that here we are talking about mind-controlled exoskeletons and yet around the world, 85% of people with a disability don't have a wheelchair.

image source: labreche_fondsdedotationclinatec

Dematerialization and Art's Identity Crisis

Erased de Kooning drawing by Rauschenberg, 1953.

Banksy MPs as chimpanzees painting sells for £9.9m
Oct 2019, BBC

I actually don't care about the price, but this mention here:

"The auction took place a year after Banksy himself intervened in a Sotheby's auction, when his artwork Girl with Balloon self-destructed as the gavel came down to become the newly titled Love is in the Bin."

I just thought it was really interesting how this painting, Girl with a Balloon, has to have two different names, like a person who transitions from one gender to another. There's a before and an after.

Private Places

Christo and Jean Claud Wrapped Reichstag, 1995.

How 3D technology is capturing the world
Oct 2019, BBC News

Photogrammetry - making 3-D models out of laser-photographs by simultaneously capturing visual and spatial information. Think of LIDAR, the sonar-like laser beams shooting from the Google maps car in all directions to create a 3-D model.

They're getting good at this, to the point that physical inspections are being done on the models and not the real thing. Which makes sense, what if you want to inspect a bridge for cracks, but it's spanning the Hudson River. You can now make a model of it, and inspect the model instead, and at any scale.

Imagine recreating via 3D printer a scaled version of something that needs to be inspected. Enemy territory, home interior, or just a bit of infrastructure that needs to be inspected for safety. Anything you want to look at, investigate, explore, or immerse yourself in, can be scanned and recreated either in 3D or in virtual reality.

Even other people's bodies? Are we going on tours of people's digestive systems? Somebody brings "back to life" Elvis's colon, models it in 3-D, and prints it out so we can walk through it like a museum?

Post Script
Living skin can now be 3-D-printed with blood vessels included
Nov 2019, phys.org

Bio-inks
In this paper, the researchers show that if they add key elements—including human endothelial cells, which line the inside of blood vessels, and human pericyte cells, which wrap around the endothelial cells—with animal collagen and other structural cells typically found in a skin graft, the cells start communicating and forming a biologically relevant vascular structure within the span of a few weeks.

They're taking two kinds of blood vessel cells, adding them to animal collagen, and printing them onto a substrate.

Within weeks, the cells start working with each other, and spontaneously generating blood vessels.

***
If you thought deepfake pornography was twisted...

Image source: Christo and Jean Claud Wrapped Reichstag 1995, photo by txmx 2

Words Out There

'Sadfishing' social media warning from school heads
Oct 2019, BBC 

Engrams emerging as the basic unit of memory
Jan 2020, phys.org

More hot neuro-pop: Ensembles, Synaptic plasticity

Giant surveillance balloons are lurking at the edge of space
Jan 2020, Ar Technica

Stratollites.

Post Script
Oldest living things on Earth?
Stromatolites

Socialbots Start Their Own Country, Declare War on the United States

Russia doesn't usually take the spotlight on the Fake Stage, but we've got a couple good ones here:

Russian trolls' chief target was 'black US voters' in 2016
Oct 2019, BBC News

People smuggler 'built fake Russia-Finland border posts'
Dec 2019, BBC News

Social media platforms leave 95% of reported fake accounts up, study finds
Dec 2019, Ars Technica

A questionably real user commented thus, and I can't agree more:

"If an algorithm wishes to identify as a human, who would be so insensitive as to delegitimize its perspective?"

Post Script
Google steps up battle on 'deepfakes' with data release
Oct 2019, phys.org

A dataset of fake faces has been standardized so it can be used to judge the fakeness of future faces. 

A 'Jackalope' of an ancient spider fossil deemed a hoax, unmasked as a crayfish
Dec 2019, phys.org

Local Chinese farmers skillfully craft a deceiving ancient spider fossil onto a not-so-ancient crayfish not-so-fossil. 

People do this with dinosaur fossils pretty often, as it can make good money. In this case, it's different because the fake spider is not just making money, it's making a new entry into the book of life -- this new spider was given its own scientific name, which was eventually taken back.

US and China clash over Uyghur harassment in Australia and 'fake police cars'
Dec 2019, Australian Broadcasting Corporation

"In an interview with ABC News Breakfast last Thursday, US ambassador Arthur B Culvahouse Jr said China was monitoring and intimidating Uyghurs living in Australia, and that this involved the use of fake Chinese police cars."

The Universal Flavor Network Emerges

Flavor pyramids for North American and East Asian cuisines, credit - Barabasi Labs, 2011

Global diets are converging, with benefits and problems
Jan 2020, phys.org

They found what sounds like a two-way shift between Asian and Western countries, where Asia is eating more animals and the West eating less.

This idea of the global diet reminds me of the Flavor Networks of 2011, where they took all the molecules in all the ingredients of all the recipes ever (56,498), and linked them together. You can see one of their network graphs above, and the rest in the paper.

The graph shown above is the part related to this global diet shift. You can see that Western recipes are populated by animals, bread and butter, and Eastern cuisine is likely to have more vegetables.

There is another way to look at this which includes the homogeneity of the ingredients. Western recipes rely on similar ingredients. Cows and their butter are related. Wheat, which is a grass, is eaten by cows, so it's in their meat and their butter. If these are the staple foods, and staple flavors of the West, then  it would make sense that those recipes are more closely related, like family.

For the East, there is less focus on animals and more on vegetables. Flavorwise, it's a more adversarial arrangement. Think onions and garlic and sesame oil and of course soy sauce. Note that soy sauce actually gets its strong animal-like flavor not from an animal but from the complex of fermented by-products that develop as a result of the browning reactions that occur after purposefully-inoculated molds and bacteria have digested and transformed the soy-energy for a long time.

The vegetables aren't eating each other and becoming part of each other's bodies like cows and grass. They aren't related in that way, so their flavors are different, and they have to stand out from each other.

And one step further, consider Jared Diamond's Guns, Germs and Steel idea. In it he points out that Asia has less medium-sized domesticable animals as compared to Western Europe. This matches the respective diets we're talking about. He also points out that the West had wheat and the East had rice as the staple grain. Finally, he reminds us that the East is predominantly north-south in its axis and the West (both Europe and America) are more on an east-west axis.

Who cares? Plants care. Plants in the north need a very different algorithm than plants in the south, because of the change in day length. Those plants will be different, and any culture that lives on a north-south axis, eating lots of different plants, will produce a diet like the one described above.

The side-to-side shape of the West allows all the plants to be on the same daylight schedule, and thus similar algorithms overall, similar plants, similar flavors.

It might be interesting to compare some data from global trade over the past 50 years to see how it matches with these diet changes (because we've pretty much obliterated any east-west or north-south biases of a culture).

Post Script:
Soylent Green is people mixed with soy and lentils and fed back to people.

Russian space lettuce was grown on the International Space Station for astronauts.

Notes:
Multidimensional characterization of global food supply from 1961 to 2013.
James Bentham, Gitanjali M Singh, Goodarz Danaei, Rosemary Green, John K Lin, Gretchen A Stevens, Farshad Farzadfar, James E Bennett, Mariachiara Di Cesare, Alan D Dangour & Majid Ezzati. Nature Food volume 1, pages 70–75 (2020). DOI: 10.1038/s43016-019-0012-2.

Flavor network and the principles of food pairing.
Yong-Yeol Ahn, Sebastian E. Ahnert, James P. Bagrow & Albert-Laszlo Barabasi. Scientific Reports, published 15 December 2011. DOI: 10.1038/srep00196

FurFuryl Mercapton, Abstract Foods, and Flavor Networks
Network Address, 2012