Search This Blog

Monday, October 28, 2019

And in Other Sports News...

The National Ultimate Frisbee Finals were held in San Diego last Sunday and an a capella quartet I was in sang the national anthems of the United States and Canada at the beginning of it. I've always wondered where they get the random people to sing the anthems at these sporting events and now I know - it's people like me! There's no recording (alas!) but here's a picture of us intrepid singers (it was windy - that's why it looks like I'm having a bad hair day):


The men's finals was really exciting. The Seattle Sockeyes won 13-12 in sudden death overtime against the Chicago Machine. The Sockeyes were up 11-6. Then the Machine scored 5 in a row to tie it up, then it was tied again at 12-12 with the next goal deciding the game. Both teams went all out for the last point for an exciting finale!

I used to enthusiastically play Ultimate so it was really fun to watch the sport. I think it's unfortunate that the sport hasn't really ever caught on at the professional level because it's really an exciting spectator sport.

Friday, October 25, 2019

Will Trump Make it to the End of His First Term?

In 2017, in a comment to one of Clovis' posts, I wrote:
And, by the way, I think Trump's chance of finishing his first term is less than 70%. He's old and someone his age has about a 7% chance of dying in the next 3+ years. Nobody's been more hated as a president, and while he's a pretty hard target, I think he has a 10+% chances of being assassinated - all you need is one traitorous secret service agent and Trump's a goner. I think there's a 5% chance he just gets sick of it all and quits (resigns). And I think there's around a 10% chance he is removed from office because they do find something bad enough about him and Russia or whatever.
I think the odds are still roughly the same. He probably won't die of natural causes and he almost certainly won't resign because "he just gets sick of it all." On the other hand, the chance of his being impeached and removed from office is probably more than 10% at this point.

I still think there's a significant chance he'll be assassinated, possibly along with Barr and other investigators in Trump's administration. If those investigators really begin indicting people and if there really is a "deep state," then they'll start killing people to stop the investigations, and I think Trump would be the primary target. And the likelihood of assassination will go up substantially if it looks like Trump is likely to win in 2020.

Whether or not there is a "deep state" is either a matter of definition or degree. There are around 2 million federal employees (far more if you include military) and most of them keep their jobs even when the parties of the president and/or congress change. These bureaucrats are in some sense a "deep state" in that they're entrenched very deeply and nearly impossible to get rid of. However, they're not necessarily a "deep state" by the more commonly used definition of people conspiring to rule over the people while ignoring elected officials and perhaps to use any method, including assassination, to guarantee their continued power.

But do they need to conspire to unseat of assassinate Trump? Or at least conspire on a grand scale? I don't think so. I believe that it is true that the vast majority of federal employees would love to get rid of Trump and I believe that each of those is willing to do at least a little bit in order to achieve that goal. Some percentage is probably willing to do quite a lot more than a little bit. Multiply that percentage, even if pretty small, by 2 million people, and suddenly you have an awful lot effort directed at bring down Trump, perhaps tens if not hundreds of thousands of people.

As a result, no conspiracy is necessary to have Trump assassinated and I think there's about a 10% chance of that happening. From large yet largely unconnected groups of people, unpredictable behaviors will emerge.

Thursday, October 24, 2019

In The Latest Critter News ...

... Rats taught to drive tiny cars to lower their stress levels.




The most important part of this bizarre story: "The rats were not required to take a driving test at the end of the study."

However, I'm wondering, did any of the rats suffer from road rage?

Tuesday, October 22, 2019

In the Wilds of San Diego

I ran into my friend Tom while mountain biking today ...

... My friend Tom T. Tarantula.

(It be more accurate to say I almost ran over Mr. Tarantula).

Friday, October 18, 2019

Quote of the Day

"Men are qualified for civil liberty in exact proportion to their disposition to put moral chains upon their own appetites,—in proportion as their love to justice is above their rapacity,—in proportion as their soundness and sobriety of understanding is above their vanity and presumption,—in proportion as they are more disposed to listen to the counsels of the wise and good, in preference to the flattery of knaves. Society cannot exist, unless a controlling power upon will and appetite be placed somewhere; and the less of it there is within, the more there must be without. It is ordained in the eternal constitution of things, that men of intemperate minds cannot be free. Their passions forge their fetters." -- EDMUND BURKE, “Letter to a Member of the National Assembly,” 1791.—The Works of the Right Honorable Edmund Burke, vol. 4, pp. 51–52 (1899).

This quote was embedded in an interesting speech by US Attorney General William Barr.

Thursday, October 17, 2019

Epic Construction Projects

There are train tracks directly behind my office. It used to be a single track, but a 2nd one was built and is about a mile long.

It took more than 2 years from the start of construction to finish that mile of track. At that rate, it would've taken 4,000 years to complete the US Transcontinental Railroad. So I was amused when an article by the historian Victor David Hanson (VDH) asked the question:
Does anyone believe that contemporary Americans could build another transcontinental railroad in six years?
Oh I'm sure someone, somewhere believes that it could be done again, but it seems beyond implausible to me. You'd think that with all of the technology we've developed, we could just snap our fingers and voila!, new railroads and regular roads and bridges and ... would appear in no time. But no, not even close.

VDH notes other typical nearly absurdly slow projects:
Californians tried to build a high-speed rail line. But after more than a decade of government incompetence, lawsuits, cost overruns and constant bureaucratic squabbling, they have all but given up. The result is a half-built overpass over the skyline of Fresno — and not yet a foot of track laid.

California’s roads now are mostly the same as we inherited them, although the state population has tripled. We have added little to our freeway network, either because we forgot how to build good roads or would prefer to spend the money on redistributive entitlements.

When California had to replace a quarter section of the earthquake-damaged San Francisco Bay Bridge, it turned into a near-disaster, with 11 years of acrimony, fighting, cost overruns — and a commentary on our decline into Dark Ages primitivism. Yet 82 years ago, our ancestors built four times the length of our singe replacement span in less than four years. It took them just two years to design the entire Bay Bridge and award the contracts.

Our generation required five years just to plan to replace a single section. In inflation-adjusted dollars, we spent six times the money on one-quarter of the length of the bridge and required 13 agencies to grant approval. In 1936, just one agency oversaw the entire bridge project.

California has not built a major dam in 40 years. Instead, officials squabble over the water stored and distributed by our ancestors, who designed the California State Water Project and Central Valley Project.
Contemporary Californians would have little food or water without these massive transfers, and yet they often ignore or damn the generation that built the very system that saves us.

America went to the moon in 1969 with supposedly primitive computers and backward engineering. Does anyone believe we could launch a similar moonshot today? No American has set foot on the moon in the last 47 years, and it may not happen in the next 50 years.
VDH wonders if a new mythology will be born based on our forebearers being able to construct wonders far beyond our modern day capabilities:
Many of the stories about the gods and heroes of Greek mythology were compiled during Greek Dark Ages. Impoverished tribes passed down oral traditions that originated after the fall of the lost palatial civilizations of the Mycenaean Greeks.
Dark Age Greeks tried to make sense of the massive ruins of their forgotten forbearers’ monumental palaces that were still standing around. As illiterates, they were curious about occasional clay tablets they plowed up in their fields with incomprehensible ancient Linear B inscriptions.
We of the 21st century are beginning to look back at our own lost epic times and wonder about these now-nameless giants who left behind monuments [such as the transcontinental railroad] that we cannot replicate, but instead merely use or even mock.
I do see his point. Who isn't frustrated with traffic being badly slowed for years while crews patch a few holes at a snail's pace?

However, VDH did leave out a few details that I think are important. First, the working conditions were really, really bad for most of those epic projects. Around 1,200 people died building the Transcontinental Railroad. The construction of the Golden Gate Bridge was noted for how "safe" it was - only 11 people died. Things are much, much more comfortable now. Almost nobody would be willing to work in those conditions and take those risks (especially for what they were paid) and even fewer in power are willing to let them take those risks.

Yet before we blame those running the projects for the death toll, we need to keep in mind that those horrible working conditions were often a step up from what the workers were previously experiencing. For example,
Many more workers were imported from the Guangdong Province of China, which at the time, beside great poverty, suffered from the violence of the Taiping Rebellion. Most Chinese workers were planning on returning with their new found "wealth" when the work was completed. Most of the men received between one and three dollars per day, the same as unskilled white workers ... A diligent worker could save over $20 per month after paying for food and lodging—a "fortune" by Chinese standards.
Second, though he does grudgingly admit it, VDH glosses over the fact that working with modern technology very often creates more value than building yet another road. Instead of concrete, we build most of our roads with glass fiber and electrons and both the market and the taxpayer think that's more valuable.

So to me, it's not so much that we were once competent at building immense material things and now we're not. Instead, it's that once upon a time we were very poor and the best we could do was work high-risk construction jobs for the "fortune" of net $20 per month whereas now we can do oh-so-much better doing other things. And those that still work construction jobs (reasonably) demand orders-of-magnitude higher pay, far better working conditions, and far better safety.

VDH ends his article with:
Our ancestors were builders and pioneers and mostly fearless. We are regulators, auditors, bureaucrats, adjudicators, censors, critics, plaintiffs, defendants, social media junkies and thin-skinned scolds. A distant generation created; we mostly delay, idle and gripe.

As we walk amid the refuse, needles and excrement of the sidewalks of our fetid cities; as we sit motionless on our jammed ancient freeways; and as we pout on Twitter and electronically whine in the porticos of our Ivy League campuses, will we ask: “Who were these people who left these strange monuments that we use but can neither emulate nor understand?”

In comparison to us, they now seem like gods.
Perhaps we do "mostly delay, idle and gripe." But we can afford to, our ancestors could not. To me, our ancestors seem far less like gods and far more like people desperately impoverished compared to us trying to do the best they could. I thank them for taking the risks and building our comfort, I really do, but gods? Not so much.

Thursday, October 03, 2019

Flynn Effected

The Flynn Effect is one of the most cited topics in the debate over nature versus nurture regarding intelligence:

The Flynn effect is the substantial and long-sustained increase in both fluid and crystallized intelligence test scores that were measured in many parts of the world over the 20th century.[1] When intelligence quotient (IQ) tests are initially standardized using a sample of test-takers, by convention the average of the test results is set to 100 and their standard deviation is set to 15 or 16 IQ points. When IQ tests are revised, they are again standardized using a new sample of test-takers, usually born more recently than the first. Again, the average result is set to 100. However, when the new test subjects take the older tests, in almost every case their average scores are significantly above 100.
Test score increases have been continuous and approximately linear from the earliest years of testing to the present. For the Raven's Progressive Matrices test, a study published in the year 2009 found that British children's average scores rose by 14 IQ points from 1942 to 2008.[2] Similar gains have been observed in many other countries in which IQ testing has long been widely used, including other Western European countries, Japan, and South Korea.[1]
This effect is strong evidence against intelligence being overwhelmingly heritable (though since IQs run from less than 50 to 200+, there's still a fair amount of potential room for nature). As a result, James R. Flynn (for whom the Effect was named), has been somewhat of a hero for those who discount the heritable nature of intelligence.

It seems, though, that Mr. Flynn's hero status has waned substantially. He recently tried to get a book published (In Defense of Free Speech: The University as Censor), but it was rejected out-of-hand by the publisher. The reasons for the rejection were explained in an email (the whole article is interesting) from the publisher:
I am contacting you in regard to your manuscript In Defense of Free Speech: The University as Censor. Emerald believes that its publication, in particular in the United Kingdom, would raise serious concerns. By the nature of its subject matter, the work addresses sensitive topics of race, religion, and gender. The challenging manner in which you handle these topics as author, particularly at the beginning of the work, whilst no doubt editorially powerful, increase the sensitivity and the risk of reaction and legal challenge. As a result, we have taken external legal advice on the contents of the manuscript and summarize our concerns below.
There are two main causes of concern for Emerald. Firstly, the work could be seen to incite racial hatred and stir up religious hatred under United Kingdom law. Clearly you have no intention of promoting racism but intent can be irrelevant. For example, one test is merely whether it is “likely” that racial hatred could be stirred up as a result of the work. This is a particular difficulty given modern means of digital media expression. The potential for circulation of the more controversial passages of the manuscript online, without the wider intellectual context of the work as a whole and to a very broad audience—in a manner beyond our control—represents a material legal risk for Emerald. ... [emphasis added]
The ironies are frightening (to me), yet delicious. The first is that a book arguing for free speech is censored. That's kinda gettin' near the end of the road for free speech, isn't it? The second is that a progressive hero is censored. As long as he was willing to research and write stuff that supports that which all right thinking people are certain is correct, he's a hero and is cited incessantly. Write something a little different and bzzzzt, throw the bastard out.

The truth may be dangerous and now we're at a point where trying to find the truth is even more dangerous.

Wednesday, September 18, 2019

Another Topic Too Dangerous to Discuss?

I find the topic of sex, gender, identity, power and social constructionism very interesting. And here's an interesting article on the topic with the following catchy excerpt:
I basically just made it up.
Human characteristics generally have a basis in some mix of nature and nurture (or DNA and memes if you prefer). Topics like the above are dangerous to discuss because if it can be interpreted that one is putting just a little too much emphasis on nature (for example that the contribution of nature/DNA is non-zero) then one can get in a lot of trouble.

I sometimes wonder if the study of biology and particularly genetics is going to be shut down in the future. The problem is that it's seemingly increasingly at odds with social science. Biologists are finding more and more correlations between genes and human traits like intelligence and various behaviors via Genome Wide Association Studies (GWAS) and are starting to propose mechanisms for the genetic basis of those traits while Social Scientists clearly assert that what biologists are finding simply cannot be correct.

Perhaps not all of biology will be banned - just those topics that have to do with things like intelligence, behavior and identity. Nonetheless, it seems like we might be headed for a different sort of Creationism - not one that's deity based, but rather social science based.

Tuesday, September 17, 2019

Richard Stallman Resigns

There have been topics that I've wanted to write about but have been hesitant to do so. For example, I found the Epstein phenomenon to be fascinating (though awful), from his motivations to his operations to his (apparent) suicide. However, it was moderately clear that writing even one word about the subject that could possibly be interpreted by anybody as not being politically correct could be devastating to me.

I met Richard Stallman, a MacArthur Fellowship Award (Genius Grant) recipient and quintessential MIT nerd a few times when I was at MIT, both at CSAIL and at parties. He was, in my opinion, quite opinionated and could be very abrasive, but he was also very smart, very talented, extremely productive and seemed to overall have a good heart as far as I could tell.

He was recently forced to resign from various positions:
In 2019, Stallman was reported by colleagues to have made statements by email in defense of Marvin Minsky, then deceased, against allegations of sexual abuse in connection with Jeffrey Epstein's alleged child sex trafficking operation.[114] In the resulting furor, Stallman resigned from both MIT[115][116] and the Free Software Foundation.[117]
I'm not totally sure, but my recollection is that Minsky was at least somewhat of a mentor to Stallman, so it's not surprising that Stallman might be inclined to try and defend his dead mentor and given that he's the quintessential MIT nerd also not surprising that he'd lack the filters to realize it would be a really bad idea to do so.

Anyway, if I needed confirmation that Epstein was yet another topic I should stay way away from, this was it.

My question is: what can I write about that won't get me in trouble? I guess more science and math stuff so that's what I'll focus on.

Wednesday, September 11, 2019

Interesting Abstract


Here is an abstract I found interesting:

Technological innovation can create or mitigate risks of catastrophes—such as nuclear war, extreme climate change, or powerful artificial intelligence run amok—that could imperil human civilization. What is the relationship between economic growth and these existential risks? In a model of endogenous and directed technical change, with moderate parameters, existential risk follows a Kuznets-style inverted Ushape. This suggests we could be living in a unique “time of perils,” having developed technologies advanced enough to threaten our permanent destruction, but not having grown wealthy enough yet to be willing to spend much on safety. Accelerating growth during this “time of perils” initially increases risk, but improves the chances of humanity’s survival in the long run. Conversely, even short-term stagnation could substantially curtail the future of humanity. Nevertheless, if the scale effect of existential risk is large and the returns to research diminish rapidly, it may be impossible to avert an eventual existential catastrophe.

This has been my intuition for a long time. My metaphor is this. Humanity/civilization is on a runway in a scramjet accelerating towards a brick wall. If we go full pedal to the metal we might, just might, be fast enough to lift off the runway in time to clear the brick wall. If we don't, we won't reach a high enough speed to to clear the wall but unfortunately our forward momentum is too great to stop and we're sure to hit the wall and that will be the end.

Thursday, September 05, 2019

YASP (Yet Another Sunset Photo)

As usual, from my apartment. This time with a little foreground rain. No rainbow though.

Wednesday, June 26, 2019

Capitalism on Parade

Socialists have a long standing argument against capitalism: it commodifies human relations, trades lives for money, and exploits the brown working class for the pleasure and benefit of the white rich.

I give you Exhibit A, which actually isn't trying to be Exhibit A.

I was going to summarize the video, except it is so well done as to well reward the 20 minutes spent watching it.

If this shows socialists to be right, why are they, nonetheless, wrong?

Wednesday, May 22, 2019

Burning Down the House

Those of a certain age, towards the trailing edge of the Baby Boom, very likely vividly remember Dr. Seuss, aka Theodor Seuss Geisel.

When the first of his books came out, the pantheon of primary readers was anchored by a trio of pallid characters: Dick, Jane, and some damn dog Spot. They did boring things — run! play! — in various boring ways, enervated by cold porridge prose.

Then along came Dr. Seuss. Off kilter drawings, quirky rhythms, and hare-brained adventures.

He is how I learned to read. I made my parents read those things to me until I had completely memorized them, a process made easier not only by their novelty, but by the cadences shoving the words into my brain. Having made the connection between sounds and words and letters, learning to read came effortlessly.

So how to view Dr. Seuss?

Poisonously, of course:


In the fall of 2017, there was a furor involving Dr. Seuss, the first lady and a school librarian that many people found surprising and disconcerting. In celebration of National Read a Book Day, Melania Trump had sent a parcel containing 10 Seuss titles to a school in Massachusetts.

What could possibly be wrong with this, providing such a powerful tool to help children read?

Lots, apparently.

At that point, we were well into the first year of the “Resistance,” and the librarian, Liz Phipps Soeiro, wanted to make various political points. Attacking Dr. Seuss was one of them. “Dr. Seuss is a bit of a cliché, a tired and worn ambassador for children’s literature,” she wrote in an open letter to Mrs. Trump, adding: “Dr. Seuss’s illustrations are steeped in racist propaganda, caricatures, and harmful stereotypes.”

Being woke must be a real burden; but then, the savior business is never easy.

To be completely fair, Mr. Geissel was not without sin.

… in Geisel’s juvenilia, his early political cartooning and some of his first books for children, he evoked ethnic and racial caricatures that were common in the early 20th century and that, by the lights of the early 21st, appear shocking and shameful.



Is the full body of Geisel’s work fatally tainted by “harmful stereotypes”? Do the origins of the hat-wearing cat really lie in minstrelsy, as Kansas State University professor Philip Nel and others believe? And if so—assuming these transgressions are detectable to the civilian eye, which is not a sure thing—do they outweigh the joy and love of reading that Dr. Seuss brought to all sorts of children and families?

The author, Mrs. Gurdon, misses a critical point: the librarians presumption of an inborn moral superiority superpower.

This librarian is not alone. Everyone of these wokelings, the ones who want to tear down statues, rename buildings, or rubbish people like Mr. Gessel are, must be, asserting that their wokeness is timeless. That had they been alive in Dr. Seuss's time, they would have been just as enlightened as they are now. We must trust their judgment not as some post hoc virtue signaling, but rather as coming from a deeper place accessible only to the vanguard, give them special dispensation to decide for the rest of us which parts of our culture must be excised.

Thus, the first question to be asked of the this librarian, and every one of her ilk: Just who the hell do you think you are?




Thursday, May 09, 2019

Only the Best and Brightest

While I have quit wasting $20/month on the NYT — at which even the least discerning puppies turn up their noses — I still get their daily news summary. From today's comes this bit of journalistic excellence:

Indonesia: A group has begun translating the Quran into sign language, helping millions of deaf Muslims get access to their holiest book for the first time.

Sunday, April 14, 2019

The Book I Should Have Written

I'm old enough to remember those halcyon days when Earth was going to be burdened with so many people that some would get pushed off the edge.

Ummm. Not so much.

In the recently published Empty Planet, The Government and UN Experts are — shocking, I know — lagging the fight.

The great defining event of the twenty-first century,” they say, “will occur in three decades, give or take, when the global population starts to decline. Once that decline begins, it will never end.

For roughly thirty years, fertility has been declining, starting with developed world. Since the 1990s, global fertility has plummeted far faster than anyone has predicted, and may well go below replacement rate within a decade.

The UN Population Division has systematically over estimated fertility, with projections out of date almost as soon as they are published. For example:

The U.N.’s most recent population forecasts suggest that the average U.S. total fertility rate from 2015 to 2020 should be 1.9 children per woman. In reality, CDC data shows U.S. fertility has averaged about 1.8 children per woman from 2015 to 2018. In 2019, early indications are that fertility will probably be nearer 1.7 children per woman.

Contrary to expectations, instead of recovering along with the economy, the US total fertility rate has continued to drop, now standing at 1.76. That amounts to 125 fewer daughters out of 1000 women per generation.

And the disconnect isn't limited to just the US — it is nearly global. UN population forecasts are almost certainly wrong, and not by just a little, but by billions.

Population decline isn't unique in history. The bubonic plague decimated Europe in the 1300s. War and famine have caused temporary, smaller, declines.

However, the recent, relentless, decrease in fertility during times of unprecedented peace, health, and material comfort is wholly unprecedented. So far, there is no indication that women, given a meaningful choice, choose to have enough children to prevent steady, relentless, global, population decline.

(Territory we covered here at Great Guys in 2013 and 2016.)

Yet, somehow, the woke are completely eaten up by GlobalWarmingClimateChangeChaos. Amazing.


Thursday, February 14, 2019

Justice Kavanaugh and Global Warming

What do they have in common?

When Dr. Christine Ford's accusation against now Justice Kavanaugh arrived with all the subtlety of the stricken Hindenburg, there was one thing that near as dammit to certain: the correlation between political proclivity and assault assessment.

Which is, or should be, beyond odd.

After all, in as much as they occupy entirely different realms, judicial philosophy and inclination towards coerced sex don't have any obvious correlation.

Yet when Dr. Ford's accusation came to light, the correlation between attitude towards constitutional originalism and Dr. Ford's credibility was nearly one. Progressives almost without exception found Dr. Ford credible; conservatives, incredible.

Same goes for Anthropogenic Global Warming. Conservative ≅ disdainful. Progressive ≅ dainful. Yet AGW, as an objective fact, just as Dr. Ford's accusation, is completely independent of judicial philosophy or political priors. These strong relationships shouldn't exist, yet there they are, nonetheless.

Welcome to motivated reasoning.

Clearly, a great many people simply do not think things through independently of their desire for a preferred outcome. Kavanaugh is to be resisted, therefore any impeachment of his character is true, and to heck with that bothersome evidence nonsense.

And just as clearly, should one have settled on individualistic free markets as the sine qua non of human flourishing, then AGW cannot, must not, be true.

Of course, as should be transparently obvious to even the most casual observer of reality, I am uniquely immune to motivated reasoning.

No matter that I agree with constitutional originalism, I am certain that Dr. Ford is a moral cretin.

And completely disregard the fact I am an individualist, AGW is nothing more than scientistic catechisms.

My reasoning is entirely unmotivated.

Now you know.


Tuesday, February 05, 2019

Deep Learning and Emergent Deception

With all of the processing power available, all kinds of Deep Neural Network learning topologies are possible with tens of millions of connections or "parameters" (which are similar in purpose to synapses in a biological brain).

One of the more interesting nets to me are Generative Adversarial Networks (GANs) which are two (or more) connected networks that fight to win in a game to "outsmart" the other network. I've written about synthetic face generation before, and those applications use GANs. One network in the GAN learns to distinguish between real faces and synthetic faces and is called the discriminative network. The other network learns to generate synthetic faces and, not surprisingly, is called the generative network. The generative network is "rewarded" when a synthetic face is so realistic that it fools the discriminative network and "punished" when the discriminative network correctly identifies that the face is synthetic and not real. And when the generative network is rewarded, the discriminative network is punished and vice-versa. The two networks are locked in this zero sum win at all costs struggle, each trying to be rewarded and avoid punishment. If the GAN is set up correctly (being correct is mostly guesswork and trial and error), it can provide really impressive results as with the case of the synthetic faces.

But deception is an inherent part of the generative network. After all, it's designed to try an fool the discriminative network and ultimately us humans. Recently, a generative network went well past the bounds of deception expected by its creators. The application is this: transform aerial images into street maps and back to automate much of the image processing for things like google maps.



The above images show the process. There's the original aerial photograph (a), the street view (b), and the synthetic aerial view (c) that's reconstructed ONLY from the street view (b).

But wait! Looking at image (c), which is constructed from ONLY image (b), how on earth did it guess where to put the air conditioning units on the long white build? Or the trees? None of those details are in the street view image (b), right?

It turns out that the network "decided" to cheat:
It learned how to subtly encode the features of one into the noise patterns of the other. The details of the aerial map are secretly written into the actual visual data of the street map: thousands of tiny changes in color that the human eye wouldn’t notice, but that the computer can easily detect.

In fact, the computer is so good at slipping these details into the street maps that it had learned to encode any aerial map into any street map! It doesn’t even have to pay attention to the “real” street map — all the data needed for reconstructing the aerial photo can be superimposed harmlessly on a completely different street map...
In other words, the street view map has gazillions of minute variations that aren't visible to the human eye that encode the data required for the remarkable aerial reconstructions.
This practice of encoding data into images isn’t new; it’s an established science called steganography, and it’s used all the time to, say, watermark images or add metadata like camera settings. But a computer creating its own steganographic method to evade having to actually learn to perform the task at hand is rather new.
Note the last sentence. The generative network wasn't very good at generating the reconstructed aerial view the way it was supposed to. So instead, it figured out how to encode the data it needed so it didn't have to learn how to do it the right way.

The thing I find most interesting is the emergent deception. Nobody predicted this would happen (since it wasn't a desired result) and I don't think anybody could've predicted it.

We're currently able to use multiple networks with hundreds of millions of connections and we're already seeing emergent behavior that can't be predicted. Every ten years gives about a factor of 100 increase in processing power and network complexity.

It will be interesting to see what emerges when thousands of networks with billions of connections interact.

Tuesday, January 29, 2019

Happy 60th Birthday to the Transistor!

And what a momentous invention it has been:
The invention of the transistor-based logic engine, the integrated circuit, turned 60 this year. Today, humanity fabricates 1,000 times more transistors annually than the entire world grows grains of wheat and rice combined. Collectively, all those transistors consume more electricity than the state of California. The rise of transistors as “engines of innovation” emerged from Moore’s Law. And we’re still in its early days: paraphrasing Mark Twain, recent reports of the death of that Law are greatly exaggerated.

Monday, January 21, 2019

Artificial Deception

I've written recently about state-of-the-art creation of synthesized faces and I concluded:
I think that the day is coming within my lifetime when there'll be no need for human actors. Any screenwriter will just be able to work with AI based tools to create and produce movies. 
But what if the "screenwriter" isn't creating a work that's meant to be viewed as fiction, but rather a fictional story that's intended to look like news? In other words, what if the screenwriting wants to purposely create fake news? And what if those creations are ever more indistinguishable from real videos of real events?

It's actually beginning to happen:
Lawmakers and experts are sounding the alarm about "deepfakes," forged videos that look remarkably real, warning they will be the next phase in disinformation campaigns.
The manipulated videos make it difficult to distinguish between fact and fiction, as artificial intelligence technology produces fake content that looks increasingly real. [...]
Experts say it is only a matter of time before advances in artificial intelligence technology and the proliferation of those tools allow any online user to create deepfakes.
As a sort of expert in this area, I believe that to be true as well.

Pornography is one the biggest areas where deepfakes are developing at the moment. For example:
Deepfakes are already here, including one prominent incident involving actress Scarlett Johansson. Johansson was victimized by deepfakes that doctored her face onto pornographic videos.
“Nothing can stop someone from cutting and pasting my image or anyone else’s onto a different body and making it look as eerily realistic as desired,” she told The Washington Post in December, calling the issue a “lost cause.”
Ms. Johansson is wise enough to realize that trying to do much about it is a "lost cause." The problem is that the software to "understand" Ms. Johansson's face and to manipulate it realistically to replace the face of someone in a video, porn or otherwise, is actually fairly trivial, widely available, and getting easier and easier to access and use. The genie is out of the bottle and there's no way to recapture it.

Besides, porn is probably fairly far down in the list of things to worry about, even if it will be driver of the technology. Other sorts of fake news will generally be more of a problem:
Other cases have resulted in bloodshed. Last year, Myanmar's military is believed to have pushed fake news fanning anti-Muslim sentiment on Facebook that ignited a wave of killings in the country.
And as the fakes get better and better, inciting mobs will be easier and easier.

Of course governments, which like to regulate everything under the sun, are working to legislate against this sort of technology use:
Farid said First Amendment speech must be balanced with the new, emerging challenges of maintaining online security. [...]
Other countries are already working to ban deepfakes.
Australia, Farid noted, banned such content after a woman was victimized by fakes nudes and the United Kingdom is also working on legislation.
Unfortunately (or maybe fortunately, depending on your point of view), my guess is that there's very little governments can do to stifle this sort of thing. Pretty much anybody with a high-end graphics card and a little too much time on their hands will be able to create these sorts of things.

In the end, I believe that the main reason fake news, including deepfake news, is a problem is that we're too damn gullible. The reason fake news works is because we want to believe it:
“We have to stop being so gullible and stupid of how we consume content online,” Farid said. “Frankly, we are all part of the fake news phenomenon.”
My guess is that after the first couple of outrageous deepfakes that catch us unawares, we'll quickly learn to be more skeptical. Hopefully, the first deepfakes don't drive us to nuclear war or anything completely catastrophic first.