Wednesday Round Up #33

The Super Mario Effect – Tricking Your Brain into Learning More

Why You Should Still Wear A Mask And Avoid Crowds After Getting The COVID-19 Vaccine

It may seem counterintuitive, but health officials say that even after you get vaccinated against COVID-19, you still need to practice the usual pandemic precautions, at least for a while. That means steering clear of crowds, continuing to wear a good mask in public, maintaining 6 feet or more of distance from people outside your household and frequently washing your hands. We talked to infectious disease specialists to get a better understanding of why.

Why Aren’t We Wearing Better Masks?

Don’t get us wrong; everything we said about the efficacy of cloth masks stands the test of time. Wearing them is much better than wearing nothing. They definitely help reduce transmission of the coronavirus from the wearer and likely protect the wearer to some degree as well. But we know that not all masks are equal, and early on in the pandemic, there was a dire shortage of higher-grade masks for medical workers. During those emergency conditions, something was much better than nothing. There are better possibilities now, but they require action and guidance by the authorities…

Fixing this problem is more urgent now that a new variant of the coronavirus, known as the B.1.1.7 lineage, is making its way around the world. This variant is believed to be about 50 to 70 percent more transmissible than earlier strains of the virus. Masks are an important part of the battle against this new variant because they decrease transmission by reducing the number of infectious particles spread by a mask wearer (known as “source control”) and by reducing the amount that a mask wearer inhales. The cloth masks that we focus on in our paper do a good job at source control, but on their own they do not protect the wearer as well as medical-grade respirators do.

Is It Really Too Late to Learn New Skills?

One problem with teaching an old dog new tricks is that certain cognitive abilities decline with age, and by “age” I mean starting as early as one’s twenties. Mental-processing speed is the big one. Maybe that’s one reason that air-traffic controllers have to retire at age fifty-six, while English professors can stay at it indefinitely. Vanderbilt cites the work of Neil Charness, a psychology professor at Florida State University, who has shown that the older a chess player is the slower she is to perceive a threatened check, no matter what her skill level. Processing speed is why I invariably lose against my daughter (pretty good-naturedly, if you ask me) at a game that I continue to play: Anomia. In this game, players flip cards bearing the names of categories (dog breeds, Olympic athletes, talk-show hosts, whatever), and, if your card displays the same small symbol as one of your opponents’ does, you try to be the first to call out something belonging to the other person’s category. If my daughter and I each had ten minutes to list as many talk-show hosts as we could, I’d probably triumph—after all, I have several decades of late-night-TV viewing over her. But, with speed the essence, a second’s lag in my response speed cooks my goose every game.

Social Allostasis and Social Allostatic Load: A New Model for Research in Social Dynamics, Stress, and Health

Theories such as social baseline theory have argued that social groups serve a regulatory function but have not explored whether this regulatory process carries costs for the group. Allostatic load, the wear and tear on regulatory systems caused by chronic or frequent stress, is marked by diminished stress system flexibility and compromised recovery. We argue that allostatic load may develop within social groups as well and provide a model for how relationship dysfunction operates. Social allostatic load may be characterized by processes such as groups becoming locked into static patterns of interaction and may ultimately lead to up-regulation or down-regulation of a group’s set point, or the optimal range of arousal or affect around which the group tends to converge. Many studies of emotional and physiological linkage within groups have reported that highly correlated states of arousal, which may reflect failure to maintain a group-level regulatory baseline, occur in the context of stress, conflict, and relationship distress.

Our Machiavellian Moment

The introduction to the English edition of The Art of Teaching the People What to Fear, written in June 2019 for readers in the United States, begins with the theme of fear in politics and an issue of Time magazine with Trump on the cover. Boucheron argues that the United States had entered a “Machiavellian moment”—“the dawning realization of the inadequacy of the republican ideal”—in the aftermath of the 9/11 terrorist attacks and that today, under “Trumpian America,” a fusion of politics and fiction has allowed for techniques of domination to be perfected, setting “a general disregard for the ‘actual truth of the matter.’” Referencing George Orwell’s 1984, Boucheron sees the United States as captured by a propaganda machine that has undermined reality and common sense—“that sixth sense Machiavelli spoke of, the accessory knowledge that the people have of what is dominating them.” Given the pervasive lack of realism in U.S. politics today, it is clear that the republic would appear to Machiavelli as a corrupt order, not because the powerful few break the rules or because a faction attempts to undermine the integrity of elections, but because the people have been “either deceived or forced into decreeing their own ruin.” Perhaps the most important part of Machiavelli’s wisdom for our own time is that republics tend to become oligarchic, giving the powerful few indirect control over government.

Marcus Aurelius on How to Motivate Yourself to Get Out of Bed in the Morning and Go to Work

So you were born to feel “nice”? Instead of doing things and experiencing them? Don’t you see the plants, the birds, the ants and spiders and bees going about their individual tasks, putting the world in order, as best they can? And you’re not willing to do your job as a human being? Why aren’t you running to do what your nature demands?

Banning Trump won’t fix social media: 10 ideas to rebuild our broken internet – by experts

*Hire 10,000 librarians for the internet

We need 10,000 librarians hired with the mandate of fixing our information ecosystem. This workforce would be global in size, but local in scope, focused on building systems for curating timely, local, relevant, and accurate information across social media platforms. Their work would be similar to the public interest obligation already applied to radio, where we legally require “broadcasting serve the public interest, convenience and necessity” as a condition of having access to the airwaves. Social media companies should tune into the frequency of democracy, rather than the static of disinformation.

*Fund training for teachers, our ‘informational first responders’

Social media feeds people lies, hate, and radicalizing messages. But people also bring lies, hate, and radicalizing messages to social media – or at least carry those interests with them when they go online, in turn triggering recommendation algorithms to serve up false, bigoted, and radicalized offerings. Something needs to be done about all those beliefs and choices too – because otherwise, even the most sweeping technological or regulatory fixes will be treating the symptoms, not the underlying causes.

We need funding for empirical, longitudinal research into media literacy education, and funding for the development of collaborative media literacy curriculum. We also need funding for research to assess what media literacy training K-12 teachers themselves need, and funding to ensure that teachers across subject area receive that training to ensure that they are equipped for their role as informational first responders. We do a great disservice to our teachers when the message is: here’s nothing, now go save democracy.

To Make Sense of the Present, Brains May Predict the Future

There are obvious technological applications for GQN, but it has also caught the eye of neuroscientists, who are particularly interested in the training algorithm it uses to learn how to perform its tasks. From the presented image, GQN generates predictions about what a scene should look like — where objects should be located, how shadows should fall against surfaces, which areas should be visible or hidden based on certain perspectives — and uses the differences between those predictions and its actual observations to improve the accuracy of the predictions it will make in the future. “It was the difference between reality and the prediction that enabled the updating of the model,” said Ali Eslami, one of the project’s leaders.

The Great Unraveling

For a while now I have thought of this period as a great unraveling — the unraveling of the old truths, the old political consensus, the old order, the old conventions, the old guardrails, the old principles, the old shared stories, the old common identity.

The metaphor of the unraveling is true enough, but it fails to capture the takeover and the unimaginable strength of the new powers that have superseded the old ones. My friend David Samuels has dubbed it the age of the machines and I think that gets it right.

“The machines ate us,” he wrote in Tablet last month. “We are all sick with the same disease, which is being pumped through our veins by the agents of a monopolistic oligarchy — whether they present themselves as the owners of large technology companies, or as the professional classes that are dependent on those companies for their declining wealth and status, or as identity politics campaigners, or security bureaucrats. The places where these vectors converge make up the new ideology, which is regulated by machines; the places outside this discourse are figured as threats, and made to disappear from screens and search results, using the same technologies that they use in China.”

The machines ate Ashli Babbitt, the 35-year-old Air Force veteran and Obama voter who slid into the gutter corners of the MAGA web and followed the siren song of Q to the capitol before bleeding out for the president in the people’s house.

We’re on the verge of breakdown: a data scientist’s take on Trump and Biden

In the US, he points out, there are two political chief executives, each commanding his own elite cadre, with nothing yet being done at a deep structural level to improve circumstances. “We’ve seen growing immiseration for 30 to 40 years: rising levels of state debt, declining median wages and declining life expectancies. But the most important aspect is elite overproduction” – by which he means that not just capital owners but high professionals – lawyers, media professionals and entertainment figures – have become insulated from wider society. It is not just the 1% who are in this privileged sector, but the 5% or 10% or even 20% – the so-called “dream hoarders” – they vie for a fixed number of positions and to translate wealth into political position.

“The elites had a great run for a while but their numbers become too great. The situation becomes so extreme they start undermining social norms and [there is] a breakdown of institutions. Who gets ahead is no longer the most capable, but [the one] who is willing to play dirtier.”

Foundations Built for a General Theory of Neural Networks

More recently, researchers have been trying to understand how far they can push neural networks in the other direction — by making them narrower (with fewer neurons per layer) and deeper (with more layers overall). So maybe you only need to pick out 100 different lines, but with connections for turning those 100 lines into 50 curves, which you can combine into 10 different shapes, which give you all the building blocks you need to recognize most objects.

In a paper completed last year, Rolnick and Max Tegmark of the Massachusetts Institute of Technology proved that by increasing depth and decreasing width, you can perform the same functions with exponentially fewer neurons. They showed that if the situation you’re modeling has 100 input variables, you can get the same reliability using either 2100 neurons in one layer or just 210 neurons spread over two layers. They found that there is power in taking small pieces and combining them at greater levels of abstraction instead of attempting to capture all levels of abstraction at once.

“The notion of depth in a neural network is linked to the idea that you can express something complicated by doing many simple things in sequence,” Rolnick said. “It’s like an assembly line.”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s