Sunday assorted links


3. uh-oh: Facebook and Google capture roughly 70% of revenues from digital advertising (a higher percentage of the marginal dollar), so it's the "simple fact that [the NYT, the New Yorker, et al.] are just as guilty as Facebook [and Google] of leaking its readers data". This being Palm Sunday, I will point out that some Christians view all sins as equal, murder and lust for example. Do you believe that?

Equal in what way, Ray? In fact, your assertion sounds like nonsense- I don't know a single Christian that believes murder and lust are equal as sins.

You must not be a Christian; it's the difference between Catholics and Protestants.

Um, indulgences, confession, access to scripture without intermediaries?

Um, the Pope?

Um, the status of Jesus?

Sometimes I think you have the posting impulse-filter of Trump on Twitter.

Good Lord, do you really view Roman Catholics as Christians? You'll be telling me next that Mormons are Christians.

Catholics are the only real Christians because we kept the original Scriptures and we follow the Apostolic sucession. President Temer is Catholic, but switched rites.

Are you saying that Catholics believe murder and lust are equal? It certainly isn't what protestants have traditionally believed. Here is the Westminster Shorter Catechism on the topic:

Q. 83. Are all transgressions of the law equally heinous?
A. Some sins in themselves, and by reason of several aggravations, are more heinous in the sight of God than others.

So Ray, you're saying that Protestants think that murder and lust are the same? After all, people who take every word of the Bible literally might do so:

James 2:10-11:

For whoever keeps the whole law but fails in one point has become guilty of all of it. For he who said, “Do not commit adultery,” said also, “Do not kill.”

But I doubt it is a Catholic position.

“For he who said, 'Do not commit adultery,' said also, 'Do not kill'.”
And vice versa.

"I don’t know a single Christian that believes murder and lust are equal as sins."

I know lots and lots of self-called Christians who pretended to not understand what was the problem with the pedophilia reports involving priests. After all, homosexuals are not prosecuted anymore and could marry in some American states. So what is the difference?

how many cheese cake factories does it take for a post keynsian to screw up a chateau?

3a, yup. The behavior of the monitors, and how often they take control from the ai, is opaque to us. It makes all those "miles driven" a very non-uniform quantity.

3a, also evidence that these guys are not really "testing" self-driving cars. They are developing them in the live environment. Upgrade status of companies who build fake cities for that.

yes the big media hype/fiction that self-driving vehicles are a near term reality -- has been toppled by that Uber/pedestrian fatality. Much has been hidden from the public to support a rosy picture of autonomous vehicles to the public. New York Times article reveals some of the shadowy doings.

Autonomous vehicles are still slowly plodding thru the development stages with big problems not even close to solution. You will not see truly autonomous vehicles in common usage for another 30+ years.

In millions of miles of testing there has been one fatality, when a pedestrian jaywalked in front of a car in the dark.

It's tragic, and it was a failure of Uber's software, but wow is your take an overreaction.

As an engineer, I simply cannot comprehend the "millionns of miles" claim. We have X different systems, with Y revisions, and Z manual overrides. There is no real baseline.

It was certainly not a design freeze and then a million miles without system change or xriver intervention.

"Motor vehicle fatalities are measured in terms of “vehicle miles traveled”... In 2016, there were 1.18 fatalities for every 100 million miles that Americans drove. Since Americans drove nearly 3.2 trillion miles that year, that still added up to tens of thousands of deaths.

To know whether self-driving cars are safer than the traditional kind, you’d have to know how many miles they traveled before incurring this first fatality. And the answer is “ fewer than 100 million” — a lot fewer. Waymo, the industry leader, recently reported logging its 4 millionth mile of road travel, with much of that in Western states that offer unusually favorable driving conditions. Uber just reached 2 million miles with its autonomous program. Other companies are working on fully autonomous systems, but adding them all together couldn’t get us anywhere close to 100 million.

One fatality at these low numbers of road-miles driven does not suggest, to put it mildly, a safety improvement over humans. It’s more like dramatic step backward .... "
(Megan McArdle Washington Post 20 March)

It's more like a homeless person stepped off a median strip in the dark in front of a moving vehicle that wouldn't have been able to stop no matter who the driver was or wasn't. One cannot conclude *anything* about safety from this one incident.

I don't know to what degree you trust expert opinion, Larry, but:

"It’s more like a homeless person stepped off a median strip in the dark in front of a moving vehicle that wouldn’t have been able to stop no matter who the driver was or wasn’t. One cannot conclude *anything* about safety from this one incident."

In this accident, there was a clear failure of the autonomous vehicle system. The lidar (and radar) should have easily seen the pedestrian, Elaine Herzberg, and the vehicle should have been easily able to stop in time.

"I don’t know to what degree you trust expert opinion, Larry, but:..."

Oops. Sorry I missed that when I commented.

"As an engineer, I simply cannot comprehend the “millionns of miles” claim. We have X different systems, with Y revisions, and Z manual overrides. There is no real baseline."

That's not really different from the 1.1 deaths per 100 million passenger miles driven by humans. That number includes drunks, speeders, people texting, driving in good and bad weather, etc.

Autonomous vehicles are improving rapidly in both hardware and software, but they are also operating in increasingly challenging situations A more appropriate comparison will be whether autonomous vehicles get to 10 billion passenger miles before 100 deaths, or even whether autonomous vehicles get to 100 billion passenger miles before 1000 deaths. (I'm confident they will meet the latter benchmark.)

Mark, as far as your prediction about the future, I can't argue because who knows.

But I do disagree that test miles are like human miles. Sure some people are drunks, speeders, people texting, in unsafe conditions, but they do not have a "monitor" in the car who stops the test when things get sketchy.

In "self-driving" the monitor does not allow the car to continue to "logical end of test." He halts it at the first sign of misjudgment and so totally corrupts the data set.

"Mark, as far as your prediction about the future, I can’t argue because who knows."

Whenever autonomous vehicles get to either 10 billion miles or 100 deaths, or 100 billion miles or 1000 deaths, we'll all know. (And it will probably be fairly obvious to half of either of those conditions.)

"In 'self-driving' the monitor does not allow the car to continue to 'logical end of test.' He halts it at the first sign of misjudgment and so totally corrupts the data set."

That is true to a large extent at the present instant, but it's rapidly become more and more untrue. All autonomous vehicle companies are experiencing more and more miles between "disengagements."

For example, Waymo cars in California drove 30,000+ miles in November of 2017 with only one "disengagement":

And in less than five years, there will be mass-produced cars without any steering wheel or brake, so there will be no possibility for human intervention.

That is a good report. I encourage everyone who is interested to take a look.

It does show impressive progress at Waymo.


Deaths per miles driven is a poor metric. The majority of accidents are going to be in high-density (sub)urban areas where very few miles are traveled, whereas most miles are driven on interstates and are relatively safe (drive in a straight line) with virtually 0 pedestrians.

I don't know the ins and outs of their testing, but it would be kind of dumb if companies did the majority of their testing driving in a straight line on the interstate. They probably do a lot of testing in marginal conditions they're worried about so they can collect more data.

You need to integrate miles driven with commensurate human risk factor (say, place and time of day) to be able to tell whether autonomous vehicles are more dangerous than humans.

Human drivers in the United States (including drunk drivers / falling asleep drivers / bad weather) have 1 fatality / 100 million passenger miles.

Computer drivers have already killed one person after 5 million miles driven, and that is with backup human drivers running only on dry pavement, with the human drivers taking over in all tricky situations.

"Human drivers in the United States (including drunk drivers / falling asleep drivers / bad weather) have 1 fatality / 100 million passenger miles.

Computer drivers have already killed one person after 5 million miles driven,..."

That isn't a statistically valid comparison. The 1 fatality / 100 million passenger miles is an *average* over 30,000+ deaths. To get a valid comparison, you'd have to average the computer-driven vehicle deaths over the first 100, 1000, or 10,000 deaths. (And the problem with that comparison is that computer-driven vehicles are getting so much safer so quickly that the average over 1000 or 10,000 deaths would underestimate the present danger, and greatly overestimate the danger at the 1000th or 10,000th death.)

"Autonomous vehicles are still slowly plodding thru the development stages with big problems not even close to solution. You will not see truly autonomous vehicles in common usage for another 30+ years."

The automobile companies don't agree with you:

Those views are certainly optimistic, but on the other hand 30+ years is just as pessimistic. Within the 5-15 year time frame is probably more accurate. Either autonomous cars will be largely operational by 2028 (or clearly very close) or the technology plateaued before we got to that point. In which case, it will probably go the way of fusion power, forever, just 20 years away.

"Those views are certainly optimistic, but on the other hand 30+ years is just as pessimistic. Within the 5-15 year time frame is probably more accurate."

At least as far as I know, GM is still planning to release a mass-produced fully automomous--no steering wheel or brake--Chevy Cruise AV *next year*:

"Either autonomous cars will be largely operational by 2028 (or clearly very close) or the technology plateaued before we got to that point."

There is no sign that the technology is hitting a plateau. NVIDIA's latest autonomous driving package can process 1.6 gigapixels per second. Only about a decade ago, no computer *on earth* (no matter what size) could do that. And in 1988, only 30 years ago, all the computers on earth, *combined*, could not do that. Most people really don't understand how rapidly the hardware and software related to autonomous vehicles is progressing.

The human mind simply hasn't evolved to think in terms of exponential growth with very short doubling times, which is how the technology related to autonomous vehicles is progressing.

3a. Yes, it's the monitors' fault. Humans are unreliable, computers aren't. It's true that humans make bad decisions, like the decision to let autonomous vehicles travel on the same streets as pedestrians, bicyclists, teenage drivers, and soccer mom drivers. When did reason become obsolete?

4. Wait! Men in1948 were sexist and chauvinistic?!? OMG! Hold the presses!

Props for using the unfashionable term “chauvinist”. Just as all bigotry is now racism, so too all chauvinism is misogyny.

1. Chess is stupefying.

3. Was Paul Samuelson defending rationing in 1948?

1) No, it is not.
2) Yes, he was. Lots of people were.

I'm trying to figure out who is getting my consumer (70% of the economy) dollars.

I believe that there are venial and mortal sins. It's a matter of degree.

From the Apostle's Creed which contains the belief in the forgiveness of sins (Cardinal Virtue: Hope). Some, me included, believe that sins may be forgiven through repentance, Confession, penances, amendment of life, and good works for the greater glorify of God Almighty.

We are all sinners. We are all poor, banished children of Eve mourning and weeping, in exile, in this vale of tears. Some Christians believe our true home is in Heaven.

A humanist might believe that murder is very bad while lust, if it feels good and harms no one, is very good. A humanist may also believe that this one, carnal life is it and we must make the best/most of it. If it feels good, do it.

It is a shame genocide and slavement were neither venial nor mortal sins. The last 3000 years would have been a lot different. A non-humanist may try to appease Whomever is punishing hi for the sins of mythological forefathers. Try sacrificing goats. I heard it gets you good harvests and keeps the panthers away.

That is interesting because the Twentieth Century is a massive experiment in ignoring Christianity. The Germans and the Soviet Union were both ruled by Humanist regimes that rejected God. Both of them managed to produce genocide on a scale unseen before in human history. And both managed to re-create slavery - on a scale that meant in less than two decades both had more slaves than the entire Trans-Atlantic Slave Trade.

Human beings are clearly unpleasant people. One might almost say born in sin. Any belief in most Gods seems to improve them.

"Any belief in most Gods seems to improve them."
I will tell the Carthaginians. And the Aztecs. Maybe you are thinking about the gods of Egypt while the Jews were slaves. What about Allah? What about Shinto's gods? Should American soldiers thank them for the kindness of Pearl Harbor?
What about codified slavery at Moses' Law? Sorry, but I think Civilization is better and it is no coincidence the post-Enlightenment West saw the end of slavery. All Christ's apostles had to say was that slaves should not be treated in a harsh way... a needlessly harsh way, I mean.

And it is no coincidence Israel follows mostly Western-created laws instead of Moses' savage code. God gave us "His" law and it sucked. Sucked enough that his chosen people would rather try to forget the whole sorry affair. Except for rules banning eating pork, of course. And circumcision. Apparently, slavery, as holy as it may be, is not to make a big return in Israel.

Well, you can tell the Catholics who signed a treaty with Mussolini after failing to recognize unified Italy for decades without an end and signed a Concordate with Hitler after opposing the Kaiser and ignoring Weimar that they were on the wrong side of History. Catholics made more noise about Hillary than they ever did about Hitler until it was too late to make any meaningful noise.

Why don't you tell the Carthaginians? Can't hurt. What's your point? Civilization is linked to religion. And atheism to mass murder. It may be the post-Enlightenment West that saw an end to slavery was it was the religious believers who did it.

The Catholics tried to run governments their way but that failed. So they finally accepted the powers that be. So what? The Catholic Church was the first, loudest and longest critic of Hitler, his racism, and the Soviet Union. The fact is that the Catholic Church was on the right side of every moral issue in the Twentieth Century. They did not excommunicate people who joined Hillary. They did to people who joined the Nazis.

You might sober up before attempting another post.

Odd thing about 2: it appears to be new last week, and yet nothing in it is more recent than 2007.

'Fraser Nelson argues the Brexit process is proving manageable'

One can wonder who will tell the current prime minister that happy news.

6) If the British government keeps giving in to EU demands, Brexit might be pretty soft. Of course, it will also generate almost none of the results promised to voters. But maybe politics really is more about theater than outcomes, in which case I guess this is a win.

I seriously doubt most brits were that concerned with regulations on the size of pillows. If Brexit happens by only changing the things that Brits actually want and little else, then that sounds like a win to me.

So women's intuition beats Samuelson's expertise?

If May continues like this there's just going to be another referendum down the line.

#6...I was one of the people for Brexit on the grounds that the EU would not likely make the necessary changes needed to avoid similar slow -motion, leaderless, kick the can down the road, debacles, as we had, in the future. How's that going?

The interview with Ken Rogoff was interesting.

When he said that he no longer saw the chessboard quite the same clear way (after years away from chess - he was very good at chess when younger but gave it up for a career as an economist) you have to feel sorry for him (even if he was always exponentially better at chess than I was).

I was never good at chess. However, I recently saw, on a website called, an HD (High-definition) attempt at recreating what the Last Supper (relevant later this week, on Maundy Thursday, the day before Good Friday) of Da Vinci looked like when the paint was still fresh.

Everyone is good at something. I don't remember faces all that well, for example, and I cannot draw with any facility, but I can mimic (some) gestures and (some) facial expressions I saw, in interesting or idle moments, 30 or 40 or 50 years ago, and have not seen since. Other people remember other things better, maybe everybody has such skills, in their way.

Looking at this fascinating work of love on (i.e., looking at the HD version of Leonardo's last supper), and remembering all the years I thought "it would be nice to see what Leonardo's depiction of the Institution of the Eucharist" looked like when brand new (to give the picture the name such pictures have when viewed in the light of their portrayals of mysteries from the Rosary - in this case, a luminous mystery (the others are joyful, sorrowful, and glorious))- well, in a way, one understands a little better why Rogoff would be so sad that he does not see the chessboard the same way as he did when he was younger and more free. Oh well, Amor Omnia Vincit, eventually.

"Everyone is good at something."

Unfortunately, this is Gardneresque wishful thinking.

No it isn't (wishful thinking).

Probably you have read Aquinas on human potential and human infused genius.

I know more about dogs than anyone I have ever met and I have never met a dog that I could not make - given the time and resources that are so often needed, in this ugly but beautiful world - into the kind of dog that, as the Hungarians like to say, simply enchants like cold water on a hot day.

Feel free to disagree. Or feel free to think that I am some loser posting on the internet from my Mom's basement. Actually, I pay her mortgage, so that she can live out her retirement in a beautiful Southern home, while I live in a bland large cheap apartment and work in the kind of mid-Atlantic American city that will, one day, when the idiots and their idiocies are forgotten, be remembered as a near paradise on earth.

Yes, my friend, everyone is good at something.

(That being said, Gardner is not the sort of person you should spend time thinking about. He is not honest. If one day he decides honesty is important - well, then listen to what he has to say).

#6 A trend I tried to track is the impact of Brexit on the UK research performance after one year.

"""The UK received more from the EU for research than it contributed[190] with universities getting just over 10% of their research income from the EU.[191] All funding for net beneficiaries from the EU, including universities, was guaranteed by the government in August 2016.[192] Before the funding announcement, a newspaper investigation reported that some research projects were reluctant to include British researchers due to uncertainties over funding.[193] Currently the UK is part of the European Research Area and the UK is likely to wish to remain an associated member.[194]"""

There was report that even a Nobel Laureate on graphene could not get research staff from EU and all of his research grants came from EU,

The trends for most UK universities seems to be business as usual with no major break from the trends, except for Oxford Uni which is one of the top two universities in UK. The research outputs from Oxford succeeded in overtaking that from Cambridge Uni for two years (+0.67 SEest and +1.13 SEest) and then suddenly broke the trend and dropped below that of Cambridge in 2017 (1-year change of -2.33 SEest). Thus with only a singular example it might not be able to claim the effects of Brexit over the whole country.

UK surrendered things like offshore fishing rights to Spain and Portugal in exchange for things like scientific research grants. EU funded research centers in UK like the Nuclear Fusion research which received 283M Euro which will be replaced by 86M Pounds (97M Euro) from the UK government. The UK fusion research center is just south of Oxford.

One year ago the projection was that Oxford might overtake Stanford in less than 10 years to be the top global science research university but the new projection is that Oxford might not be able to overtake Stanford. That could be the price of Brexit.

Elites like to spend our money on expensive toys like applied fusion research that are very unlikely to provide any kind of cost benefits, never mind a return, so I am not sure that is a bad thing.

Britain gives more money to the EU than it gets back. If the British government wishes to support research, then Oxford will get as much or even more than it did before. If not, it won't. That is the choice British people have to make.

But in the end Oxford and Cambridge rose to greatness without the EU. They can and will thrive or fail without the EU as well.

Funding is a terrible measure for scientific health. I'd rather see data on the number of patents being issued, or the number and quality of papers being published.

5. Bach and sex.

This is from Counterpunch - a lying Trotskyite rag that makes the Yellow Press look good. However what is interesting is that they too got taken by the Russian Troll factory. Not that it seems hard in this case:

During the 2016 presidential election, CounterPunch published the writings of Alice Donovan who purported to be a freelance writer but was in fact a pseudonymous employee of the Russian government.[31] Donovan was tracked by the FBI for nine months.[31] According to The Washington Post, "she seemed to be doing the Kremlin’s bidding by stoking discontent toward Democratic front-runner Hillary Clinton and touting WikiLeaks, which U.S. officials say was a tool of Russia’s broad influence operation to affect the presidential race."[31] In late November 2017, the Washington Post contacted CounterPunch about Donovan; co-editor Jeffrey St. Clair said that Donovan's pitches did not stand out amongst the pitches that CounterPunch received daily.

I doubt he sees it but that last sentence is quite a criticism. Still, I am impressed by the utter silence from the pro-Hillary Hard Left about this. Being taken for a fool by the Russians - to put it as nicely as I can - is suddenly not of interest when it is the Left. If they did not have double standards .....

Off-topic -- this 'comment of the week' on SSC (about the effect of tax reform on corporate tax shelter strategies) looks like good raw material for MR:

Comments for this post are closed