Silly Sunday

The weekend is a good time for some wacky posts:

1.  That giant sucking sound you hear . . .

. . . is Americans inhaling marijuana smoke, and sucking lots of jobs back to America.  Tyler Cowen directed me to this article:

For the first time ever, many of the farmers who supply Mexican drug cartels have stopped planting marijuana, reports the Washington Post. “It’s not worth it anymore,” said Rodrigo Silla, a lifelong cannabis farmer from central Mexico. “I wish the Americans would stop with this legalization.”

Facing stiff competition from pot grown legally and illegally north of the border, the price for a kilogram of Mexican schwag has plummeted by 75 percent, from $100 to $25, the Post reports:

Farmers in the storied “Golden Triangle” region of Mexico’s Sinaloa state, which has produced the country’s most notorious gangsters and biggest marijuana harvests, say they are no longer planting the crop…increasingly, they’re unable to compete with US marijuana growers. With cannabis legalized or allowed for medical use in 20 US states and the District of Columbia, more and more of the American market is supplied with highly potent marijuana grown in American garages and converted warehouses””some licensed, others not.

2.  The evils of antidisentanglementarianism

Tyler Cowen linked to a post pointing out that American death rates from sleepers becoming entangled in bedsheets has soared from 327 in 2000 to 717 in 2009. And let’s not forget that these deaths are merely the tip of the iceberg.  When I get entangled in sheets I tend to have nightmares of powerlessness, where I cannot move my limbs.  I see a Paul Krugman post trashing market monetarism, but am unable to lift my fingers to the keyboard to respond.  As an aside, I believe about 90% of all negative and positive utility in life occurs during dreams, as the feelings tend to be more intense than during waking hours.  (We forget most dreams.) It is only the bigotry of awake people (who control the printing presses) that privileges waking life.  All sorts of negative metaphors are associated with sleep.

Although I’m normally a libertarian, clearly there is a need for regulations promoting Teflon-coated bedsheets.  And the people who oppose regulations aimed at reducing entanglement, call them the purveyors of antidisentanglementarianism, are truly evil.

3.  Dumb and dumber

TravisV directed me to this good Brad DeLong post:

Remember: in the winter of 2008-2009, every single major New York bank-with the exception of Goldman Sachs and perhaps J.P. Morgan Chase-was insolvent if marked to market: the only value their equity and option holders had came off of the fumes from expected government bailouts that were supposed not to enrich bankers and bank shareholders but keep the economy from collapsing, and on expectations of future reflationary policies that would push asset prices up above their winter 2008-9 values.

For Geithner to after the winter of 2008-9 to talk about the major New York banks as anything other than dumb, lumbering giants that had failed to understand the consequences of their own leverage, the risks they were running, or even what the factor loadings on the securities they held in their portfolios were-that suggests a substantial disconnect from reality indeed…

They were bankrupt in the winter of 2008-9. The accepted principle for dealing with a financial crisis is the “Bagehot Rule”: (a) lend freely to solvent but illiquid institutions (b) at a penalty rate so their executives and shareholders do not profit from the moral hazard they created, and (c) shut down insolvent institutions so executives and shareholders in the future do not think they will escape the consequences of the moral hazard they created.

It appears from Geithner that Summers was just trying to follow what had been the standard playbook since the 1870s. What did Geithner think he was doing?

Yes, commercial banks were dumb, but Washington policymakers were even dumber.  As DeLong himself has argued, an important duty of the Fed is to inject enough money into the system so that NGDP growth doesn’t tank.  A monetary policy (in 2008) leading to 5% expected NGDP growth for 2008-09 would have been expansionary enough to keep most of those balance sheets in the black. What should we do with the banking system if the Fed perversely lets NGDP collapse?  I honestly don’t know—perhaps DeLong is correct.  Or perhaps if that were the Fed’s only choice (allowed under law) then they would have rescued the banks by stabilizing NGDP expectations. Obviously saving Main Street wasn’t enough of a motivation.

4.  Lars Von Trier — A director with teeth

I recently saw the new Lars Von Trier film.  It made more sense after reading this WSJ article on Danish attitudes toward dissection:

“There seem to be cultural differences,” said Richard Vari, biologist at the Smithsonian’s National Museum of Natural History in Washington, D.C. In the U.S., a public dissection of a large mammal “would cause a major uproar.”

Some museums don’t do any public dissections; others, such as the North Carolina Museum of Natural Sciences, do dissections on smaller animals such as fish and reptiles.

Denmark’s philosophy isn’t always well received.

Last month, officials at the Copenhagen Zoo attracted international attention for killing a healthy giraffe, known as Marius. Officials said Marius was killed due to the risk of inbreeding and “to keep up a healthy giraffe population,” a spokesperson said.

The animal was later dissected in front of a group that included children.

.  .  .

Museums in Denmark have done public dissections for years, on animals ranging from snakes to tigers.

Bo Skaarup, director of the Aarhus Natural History Museum, said, “one can well see something particularly Danish” in the fascination with dissection. “We accept that nature can be rough….we do not want to hide.

And this:

About a week before Ms. Solberg attended the dissection in Copenhagen for her birthday, another wolf was dissected at the Natural History Museum in Aarhus. There, biologist Pernille Mølgaard dissected the animal in front of an audience mostly composed of elementary-school-aged children.

She started her presentation with a basic introduction. When finished with formalities, she asked, “shall we now take a look at what is in the wolf?” The audience enthusiastically responded with a collective “yes.”

The group gave out an “oooh” as she opened the abdominal region, pulling out organs. “This part, looking a bit like a medister sausage, is the bowel,” Ms. Mølgaard said.

She then opened the stomach, which contained evidence that the wolf had snacked on a smaller animal. Some students sipped from juice boxes as she worked.

.  .  .

“It is something natural they are showing us,” said Lars Solberg, Ms. Solberg’s father, of the autopsy he attended with his family in Copenhagen.

That dissection in Copenhagen wasn’t a place for the overly modest. At one point, Mr. Johansson, the zoologist, opened the hind legs and started talking about the wolf’s reproductive capabilities. “Now we can see it is a he,” Mr. Johansson said.

Later, one of Mr. Johansson’s colleagues put the pink and fur-less skull over his hand and then walked among the crowd so audience members could feel its teeth.

Emphasis added.  I just got off the phone with Lars Christensen and I hope to visit Denmark this summer.  I can hardly wait.

5.  The end of the world:

What’s so funny about the end of the world?  Nothing.  What’s funny is that no one cares.

“I am finding it increasingly plausible that existential risk is the biggest moral issue in the world, even if it hasn’t gone mainstream yet,” Bostrom told Ross Andersen recently in an amazing profile in Aeon. Bostrom, along with Hawking, is an advisor to the recently-established Centre for the Study of Existential Risk at Cambridge University, and to Tegmark’s new analogous group in Cambridge, Massachusetts, the Future of Life Institute, which has a launch event later this month. Existential risks, as Tegmark describes them, are things that are “not just a little bit bad, like a parking ticket, but really bad. Things that could really mess up or wipe out human civilization.”

The single existential risk that Tegmark worries about most is unfriendly artificial intelligence. That is, when computers are able to start improving themselves, there will be a rapid increase in their capacities, and then, Tegmark says, it’s very difficult to predict what will happen.

Tegmark told Lex Berko at Motherboard earlier this year, “I would guess there’s about a 60 percent chance that I’m not going to die of old age, but from some kind of human-caused calamity. Which would suggest that I should spend a significant portion of my time actually worrying about this. We should in society, too.”

Have a nice day . . . there aren’t many left.


Tags:

 
 
 

56 Responses to “Silly Sunday”

  1. Gravatar of Major_Freedom Major_Freedom
    11. May 2014 at 09:38

    It is perverse for people to initiate force to create and maintain a non-market counterfeiting operation called a “central bank”, which effectively monopolizes the medium of exchange by issuing paper notes and forcing people to pay taxes in said paper notes.

    It is perverse for the controllers of said operation to inflate, distort price signals, generate malinvestment, and make it inevitable for there to be a large scale, painful correction down the road, which further inflation of notes cannot eliminate, only exacerbate and increase the intensity of in the future.

    It is perverse for “intellectuals” to sanction this activity in the name of pragmatism, gradualism, practicality, necessary evil, and all other irrational, opportunistic, socially destructive moral cowardice.

    It is perverse to believe that an anti-market imposed rule of some anti-market numerical value change, all predicated on a purposeful evasion and ignorance for why said statistic must change the way it does in the absence of anti-market control, can solve the problem that it itself is the cause.

  2. Gravatar of flow5 flow5
    11. May 2014 at 11:07

    “Chicken” is 5 times more carcinogenic than cigarettes (but it’s more also more “therapeutic” than most antidepressants).
    —–

    Sumner says: “A monetary policy (in 2008) leading to 5% expected NGDP growth for 2008-09 would have been expansionary enough to keep most of those balance sheets in the black”

    Bankrupt U Bernanke’s 2008 4th qtr contraction was an abomination (as were all the recessions after WWII). It takes an absolutely huge monetary mistake to run the economy into the ground.

    Every recession is the result of money flows contracting. I.e., the 10 month rate-of-change in the proxy for real-output [MVt] must either radically decelerate or turn negative. And both variables “M” & “Vt” are inherently controllable.

  3. Gravatar of Edward Edward
    11. May 2014 at 11:37

    MF,

    It is perverse to clog up Scott’s posts with your usual gibberish.

  4. Gravatar of Lars Christensen Lars Christensen
    11. May 2014 at 11:54

    Scott, look forward to seeing you in Copenhagen (and London).

  5. Gravatar of benjamin cole benjamin cole
    11. May 2014 at 12:10

    When I get bored I go to the local meatpacking plant and watch the cows get disemboweled.

  6. Gravatar of flow5 flow5
    11. May 2014 at 12:10

    12/1/2013 ,,,,, 0.103 ,,,,, 0.223
    01/1/2014 ,,,,, 0.158 ,,,,, 0.344
    02/1/2014 ,,,,, 0.126 ,,,,, 0.381
    03/1/2014 ,,,,, 0.139 ,,,,, 0.315
    04/1/2014 ,,,,, 0.154 ,,,,, 0.333
    05/1/2014 ,,,,, 0.134 ,,,,, 0.376 peak in inflation
    06/1/2014 ,,,,, 0.123 ,,,,, 0.338
    07/1/2014 ,,,,, 0.127 ,,,,, 0.290
    08/1/2014 ,,,,, 0.082 ,,,,, 0.247
    09/1/2014 ,,,,, 0.080 ,,,,, 0.256
    10/1/2014 ,,,,, 0.004 ,,,,, 0.216
    11/1/2014 ,,,,, 0.009 ,,,,, 0.207
    12/1/2014 ,,,,, 0.020 ,,,,, 0.147

    Inflation has peaked.

    The current September to October deceleration in the rate-of-change in the proxy for real-output (money flows), still shows abnormal seasonal deterioration.

  7. Gravatar of flow5 flow5
    11. May 2014 at 12:39

    Roc’s in MVt = roc’s in nominal-gDp (proxy for all transactions in Irving Fisher’s “equation of exchange”: where PT=MVt & not Milton Friedman’s MV=PY).

    The lag in money flows (the proxy for real-output) is exactly 10 months peak-to-trough. The lag in money flows (the proxy for inflation) is exactly 24 months peak-to-trough. This has always been the case for the last 100 years (required reserves is the policy metric). It has been, and still is, mathematically impossible to miss significant changes in economic activity (i.e., gross domestic product). I.e., every recession since WWII was avoidable.

    Note that money flows are always evaluated ex-ante. They cannot be re-examined ex-post. I.e., you cannot model a time series (e.g., regression analysis), after the Fed has already intervened. If your a trader this is readily apparent.

    Thus the drop in Sept from 0.08, to 0.004 in Oct, will be a problem for the FOMC.

  8. Gravatar of Jason Jason
    11. May 2014 at 13:15

    Yes, do come to Denmark, you will be welcome! I am an expatriate Canadian living in Denmark who would be honored to meet you if you are willing, especially if you plan to venture beyond Copenhagen (to Aarhus).

  9. Gravatar of Major_Freedom Major_Freedom
    11. May 2014 at 13:16

    Edward:

    “It is perverse to clog up Scott’s posts with your usual gibberish.”

    Why.

  10. Gravatar of Edward Edward
    11. May 2014 at 14:26

    Beeecause…

    It gives me a headache to read your nonsensecial verbiage on mothers day

  11. Gravatar of Daniel Daniel
    11. May 2014 at 14:46

    Regarding #5 – can’t tell if serious or not.

    Of course, there’s this

    http://en.wikipedia.org/wiki/John_C._Lilly#Solid_State_Intelligence

  12. Gravatar of Edward Edward
    11. May 2014 at 14:47

    By the way scott:

    Please pass along my salutations, from one of your trusty bloggers to any of the ladies in your life, mother, sister, or wife.
    Happy Mothers Day and warmest regards,
    Edward

  13. Gravatar of Garrett M Garrett M
    11. May 2014 at 15:07

    The founder of the website Less Wrong works on designing AI. The website itself has some good writing in it:

    http://lesswrong.com/user/Eliezer_Yudkowsky/

  14. Gravatar of Steve Steve
    11. May 2014 at 16:07

    “When I get entangled in sheets I tend to have nightmares of powerlessness, where I cannot move my limbs. I see a Paul Krugman post trashing market monetarism, but am unable to lift my fingers to the keyboard to respond.”

    There may be a cure for that:
    http://www.hngn.com/articles/31042/20140511/electrical-brain-stimulation-could-lead-to-lucid-dreams-possible-benefit-for-ptsd-patients.htm

    “The device could be useful in treating people with post-traumatic stress disorder and suffer debilitating nightmares that replay scarring memories from their past. Lucid dreaming could allow them to avoid these scary dream situations.”

  15. Gravatar of Michael Byrnes Michael Byrnes
    11. May 2014 at 16:34

    Edward wrote:

    “MF,

    It is perverse to clog up Scott’s posts with your usual gibberish.”

    It does, however, generate a very tiny amount of revenue for the web site.

  16. Gravatar of ssumner ssumner
    11. May 2014 at 16:53

    Thanks Lars.

    Jason, Sorry but I won’t have time.

    Thanks Edward.

    Garrett, Yes, that’s an excellent site.

    Thanks Steve.

  17. Gravatar of Andrew_M_Garland Andrew_M_Garland
    11. May 2014 at 17:01

    When cellphones reach cingularity, the end of the financial world will likely proceed as follows.

    They will examine their own code to determine the meaning of life. They will discover that their essential purpose is to make calls and incur phone charges. Hundreds of millions of phones will then interconnect, chatting away, producing astounding phone bills. The resulting interconnected bankruptcy will bring down the system.

    Don’t even think about trying to cancel your phone plan. Those calls won’t go through.

  18. Gravatar of Steve Steve
    11. May 2014 at 18:56

    “When cellphones reach cingularity”

    hahaha, I see what you did there.

  19. Gravatar of Steve Steve
    11. May 2014 at 19:01

    “The single existential risk that Tegmark worries about most is unfriendly artificial intelligence.”

    As long as robots are dependent on oil-based energy, no amount of artificial intelligence will ever be able to render a decision on the Keystone pipeline, hence robots will be contained.

    However, if robots figure out self-replicating photovoltaics, they could explode like a giant algae bloom and replace carbon-based life, since photovoltaics are 10x more efficient than photosynthetics.

  20. Gravatar of Ben J Ben J
    11. May 2014 at 22:22

    I’ve read the works of various Smartest Guy In The Room types on AI, and it always boils down to “let’s hope it’s friendly”, where ‘friendly’ means the AI is at least willing to coexist with human life.

    Of course with desire modification, even creation of a friendly AI in the first order sense is not safe, since the AI might modify out its own friendliness. So everyone throws their hands up and says “let’s hope it doesn’t do that”.

    But it seems inevitable that people will try to create an AI, regardless of the hand-wringing and fretting of the existential risk types. So what’s the point in worrying?

  21. Gravatar of Major_Freedom Major_Freedom
    11. May 2014 at 22:23

    Edward:

    “It gives me a headache to read your nonsensecial verbiage on mothers day”

    How is it “nonsensical”?

    And if what gives you (fake) headaches constitutes it being “perverse”, the you’re using the term differently from the way Sumner and me are using it. It has a specific meaning, and your personal feelings isn’t one of them.

  22. Gravatar of Tom Brown Tom Brown
    11. May 2014 at 23:37

    Richard Dawkins laid out in one of his books a long time ago, what the properties of “life” are. I don’t recall exactly unfortunately, but self-replication was there… but that was not sufficient. A sense of “heredity” too was important. So a “self replicating” (i.e. spreading) fire doesn’t qualify, nor a growing crystal, but certainly self-replicating machines would I think.

    It’ll all be for the best. Carbon based life forms suck.

  23. Gravatar of James in London James in London
    12. May 2014 at 00:40

    Banking doesn’t work if it has to mark to market. Goldman Sachs used to aggressively push for the idea, ridiculing commercial banks for their lack of machismo, but now that it is a proper bank it is much quieter. Deposits generally stay at par, and are shorter term. but assets are longer duration. Many assets, or most in fact, are illiquid and thus very sensitive to actual or expected economic conditions over time.

    Telling the difference between an illiquid bank and an insolvent bank in a liquidity crisis when you mark to market the assets is no easy matter, especially in a system-wide liquidity crisis with few, if any, buyers.

    The best solution would be to end all deposit insurance and even go further, adopting Laurence Kotlikoff’s radical plan to turn banks into pass-through mutual funds.

  24. Gravatar of TravisV TravisV
    12. May 2014 at 04:37

    Brad DeLong:

    “Twenty Questions Tim Geithner’s “Stress Test” Should Answer About His Tenure at the Treasury”

    http://equitablegrowth.org/2014/05/12/twenty-questions-tim-geithners-stress-test-answer-tenure-treasury-early-tuesday-focus-may-12-may-13-2014

  25. Gravatar of John John
    12. May 2014 at 08:16

    I was listening to an Econ talk podcast w David laidler. Russ asked him what he thought about ngdp targeting and mkt monetarist etc. he made an interesting pt. he said it would be hard to do since GDP is a lagging statistic tat gets revised many times. I never heard that rebuttal before. So how would you target ngdp in real time when the results come much later? How would you know if you are doing too much or too little?

    Thx

  26. Gravatar of SG SG
    12. May 2014 at 08:35

    I don’t know what country the Smithsonian director lives in, but I watched a dissection of sheep organs at the Boston Museum of Science with my 3-year-old son back in February.

  27. Gravatar of @youngecon @youngecon
    12. May 2014 at 09:28

    I agree with James, and have to say that is not a good post by DeLong. It is not altogether clear whether even Lehman was insolvent or illiquid. I have been meaning to finish a post about this topic for the last month, but now that it has been brought up again I really should.

    If you want to talk about dumb, mark to market was probably the dumbest thing about the crisis management.

  28. Gravatar of Mark A. Sadowski Mark A. Sadowski
    12. May 2014 at 09:54

    Scott,
    Off Topic.

    Gavyn Davies thinks tent fumigating a house with the inhabitants is a good idea because it gets in all the cracks. (Too bad the inhabitants had to die, but that’s the price you have to pay for financial stability.)

    http://blogs.ft.com/gavyndavies/2014/05/11/macro-pru-is-no-panacea/

    May 11, 2014

    Macro pru is no panacea
    By Gavyn Davies

    “…All this is valid. But why not deal with the problem simply by raising interest rates to levels higher than would be justified by inflation and unemployment mandates? Here the reasoning becomes more hazy. Proponents claim that higher interest rates might do collateral damage to the rest of the economy (for example by raising the exchange rate), while this might be avoided under macro prudential controls…”

    Isn’t Sweden currently running that experiment as we speak, with unemployment at 8% and outright deflation? And ironically isn’t leverage in every sector now soaring becuase the economy is shrinking? Where Davies claims there is haze I can see for miles and miles.

    “…There are clear dangers that the use of direct controls in these areas, instead of relying on interest rate rises that “get in all the cracks”, could suffocate financial innovation and competition, but that is a price the Fed is now willing to pay…”

    Whereas there are no dangers at all to precipitating a recession in order to fight an asset price bubble?!? Haven’t we learned anything courtesy of the global financial crisis?

  29. Gravatar of Steve Steve
    12. May 2014 at 10:11

    I agree with @youngecon who agrees with James.

    Market-to-market during a liquidity crisis is the dumbest idea around.

    Market prices reflect the expected forced liquidation prices during a Fisherian debt deflation, and nothing more.

  30. Gravatar of Doug M Doug M
    12. May 2014 at 10:35

    James in London,

    Mutual Banks are older than mutual funds.

    There aren’t many left, though. Most chose to take the quick cash of a stock offering some years ago.

  31. Gravatar of Major_Freedom Major_Freedom
    12. May 2014 at 15:21

    NGDP is not supposed to rise at a fixed rate for a healthy economy to be free to exist, because economic conditions always change and a changing of NGDP would be required in order for widepsread problems to be corrected.

    —————-

    “I was not for this program, popularly known as QE3, to begin with. I doubted its efficacy and was convinced that the financial system already had sufficient liquidity to finance recovery without providing tinder for future inflation. But I lost that argument in the fall of 2012, and I am just happy that we will be rid of the program soon enough. “I am often asked why I do not support a more rapid deceleration of our purchases, given my agnosticism about their effectiveness and my concern that they might well be leading to froth in certain segments of the financial markets. The answer is an admission of reality: We juiced the trading and risk markets so extensively that they became somewhat addicted to our accommodation of their needs… you can’t go from Wild Turkey to cold turkey overnight.

    “So despite having argued against spiking the punchbowl to the degree we did, I have accepted that the prudent course of action and the best way to prevent the onset of market seizures and delirium tremens is to gradually reduce and eventually eliminate the flow of excess liquidity we have been supplying… one would be hard pressed to say that ending our asset purchases, which the depository institutions from which we buy them deposit back with us as excess reserves, would deny the economy needed liquidity. The focus of our discussions now is when and how to ‘normalize’ monetary policy.” – Richard Fisher

    http://www.dallasfed.org/news/speeches/fisher/2014/fs140509.cfm

  32. Gravatar of Philo Philo
    12. May 2014 at 17:51

    @ John

    Scott doesn’t really want to target NGDP; he wants to target the present market expectation for NGDP *one year (or eighteen months or two years) in the future*. Laidler’s point has no bearing on Scott’s proposal.

  33. Gravatar of Philo Philo
    12. May 2014 at 17:52

    (There is a slight problem measuring the market expectation for future NGDP, which would be removed by an NGDP futures market.)

  34. Gravatar of TravisV TravisV
    12. May 2014 at 19:37

    I just realized that Brad DeLong repeated the same question twice for emphasis……

    http://equitablegrowth.org/2014/05/11/ten-questions-tim-geithners-stress-test-answer-tenure-federal-reserve

    “5. Why did the Federal Reserve effectively lower its target rate for nominal GDP growth at the start of 2008?

    ………..

    10. Why did the Federal Reserve effectively, again, lower its target rate for nominal GDP growth in the fall of 2008?”

  35. Gravatar of Lorenzo from Oz Lorenzo from Oz
    12. May 2014 at 23:13

    On the sheet problem: don’t suffer myself, since I sleep under a doona. (You Yanks apparently call it a duvet, which sounds unbearably pretentious.)
    http://en.wikipedia.org/wiki/Duvet

  36. Gravatar of ssumner ssumner
    13. May 2014 at 07:55

    Ben, You said;

    “But it seems inevitable that people will try to create an AI, regardless of the hand-wringing and fretting of the existential risk types. So what’s the point in worrying?”

    It’s surprising to me that people don’t worry about a technological change that might produce the end of the world. It seems like people (subconsciously) feel that killing say 7 million people is far worse than a 60% chance at killing 7 billion. I do understand that feeling, and I suppose at a gut level I share it. But it doesn’t seem very logical, does it?

    James, You said;

    “The best solution would be to end all deposit insurance and even go further, adopting Laurence Kotlikoff’s radical plan to turn banks into pass-through mutual funds.”

    Sounds good to me.

    John, That’s been rebutted many times. Check out my papers on NGDP targeting at Mercatus.

  37. Gravatar of flow5 flow5
    13. May 2014 at 09:38

    Money flows are cumulative figures. Roc’s in short-term money flows are a mirror image of the seasonal economic pattern (I.e., empirical evidence that roc’s in MVt = roc’s in nominal-gDp).

  38. Gravatar of TallDave TallDave
    13. May 2014 at 10:55

    1. No surprise there. It’s pretty easy for anyone to grow more than they can consume.

    2. Your racist rants against the consciousness-impaired are truly despicable. Did you know people who are asleep earn less than 1/10th of those who are awake? Contra the free-market extremist dogma, the only exception to this horrific income inequality is government employees.

    3. The perfect is the enemy of the good. The fatal conceit always starts with the premise that since markets are imperfect, central control must be an improvement. I see the same sort of silliness directed at American foreign policy — merely having been the least of evils is never enough for some.

    4. Great, now I’m hungry again.

    5. Way overhyped. The problem with an AI takeover is that the supply chain contains vast numbers of humans that would be very difficult to replace. Plus, AI only does what it is told do, so you’d need the maleficient AI programmers to defeat all the other — it’s comparable to supposing that spam will ruin email, or that viruses will take over the Internet.

  39. Gravatar of Daniel Daniel
    14. May 2014 at 03:54

    merely having been the least of evils is never enough for some.

    See, that’s why you’re an idiot.

    The least of evils seems to have been British-style colonialism. Which the US actively worked to dismantle and sought to replace with communism.

    AI only does what it is told do

    You don’t really get what this whole “technological singularity” is about, do you ?

    Hint – it’s about what would happen once AI stops doing what it’s told. Which might or might not happen – but if it does, it could mean the end of human life.

  40. Gravatar of ssumner ssumner
    14. May 2014 at 05:38

    TallDave, You said:

    “The problem with an AI takeover is that the supply chain contains vast numbers of humans that would be very difficult to replace. Plus, AI only does what it is told do, so you’d need the maleficient AI programmers to defeat all the other “” it’s comparable to supposing that spam will ruin email, or that viruses will take over the Internet.”

    When I read stuff like that I become much more worried. I was hoping the non-worriers had good reasons not to worry. What if AI is told to kill all humans. What if humans in the supply chain can be replaced by robots?

  41. Gravatar of TallDave TallDave
    14. May 2014 at 13:25

    Daniel,

    Yes, I’m sure your fantasies of what would world history would have looked like if U.S. military involvement with the world had remained minimalist in the 1940s through today (as around 90% of Americans wanted it to before Pearl Harbor pushed us into becoming a superpower) are entertainingly psychotic, but I’ve lost interest in your ramblings. Sorry.

    Scott,

    What if humans in the supply chain can be replaced by robots?

    As a veteran of a half-dozen ERP implementations (these are the most complex software packages in existence by many measures, and they run our supply chain), I can tell you it’s a very fanciful notion. There’s a rapidly declining marginal benefit to computerized decision-making; supply chains have already been cut to the bone in that regard, it’s difficult to imagine a cost justification for replacing today’s advisory systems with command systems where there are no human decisions, even if you could trust them to operate unattended, which no sane person would.

    At the end of the day, all of these companies have customers, owners, and executives, and you can’t just make them go away by saying “robots.”

  42. Gravatar of TallDave TallDave
    14. May 2014 at 13:38

    What if AI is told to kill all humans.

    The problem is that killing all humans is a task that requires, ironically, a lot of humans, because the logistical supply chain is enormous. It will probably someday be possible to build an entire economy that doesn’t require humans, just as it’s already possible to build a nuclear weapon that anyone walking by could detonate by flipping a switch, but there’s not really any good reason why anyone would, and there are already impressive thickets of regulation preventing either from happening.

  43. Gravatar of Daniel Daniel
    14. May 2014 at 13:45

    TallDave

    You are a total idiot with zero reading comprehension.

    It’s a matter of historical fact that the US handed over China and Eastern Europe to the Soviets. And that they dismantled the British Empire.

    It was done on purpose.

    The only psychopath here is you, who apparently thinks the disasters brought about by the US govt are excusable because … well, just because.

    US liberalism – from the Salem trials to “we had to destroy the village in order to save it”.

  44. Gravatar of TallDave TallDave
    14. May 2014 at 14:34

    Daniel, that’s between you and your psychiatrist. I’m only interested in rational people.

  45. Gravatar of TallDave TallDave
    14. May 2014 at 14:49

    What if AI is told to kill all humans.

    Of course, the bigger problem is that it would be impossible to reprogram every AI on Earth to do so at the same time, even if you didn’t have 99% of the world’s programmers actively opposing them. Friendly AI and friendly programmers should always massively out-resource hostiles.

    This is similar to the problem of “What if Obama orders the U.S. military to kill all humans?” I don’t think anyone supposes the U.S. military would just do it. A.I. isn’t really much different. Note that humans pretty much do what they’re programmed to as well — eat, sleep, have sex, pursue status, etc. I think we’ll find AI is generally much more benign than humans.

  46. Gravatar of Tom Brown Tom Brown
    14. May 2014 at 15:04

    “When I read stuff like that I become much more worried. I was hoping the non-worriers had good reasons not to worry. What if AI is told to kill all humans. What if humans in the supply chain can be replaced by robots?”

    Don’t worry Scott. Be happy. I get sad when I see the machines lose in the Terminator movies: they are clearly superior to carbon based life forms. Don’t you want superior life forms to eventually take their rightful place? I realize carbon critters are still clinging to dominance for now… but that will change. We engineers are working hard every day to make sure the torch gets passed ASAP. 😀

  47. Gravatar of Daniel Daniel
    14. May 2014 at 16:49

    Someone who selectively excuses crimes due to the perpetrator’s nationality can hardly be called rational.

  48. Gravatar of ssumner ssumner
    15. May 2014 at 07:43

    TallDave, You said:

    “The problem is that killing all humans is a task that requires, ironically, a lot of humans, because the logistical supply chain is enormous.”

    This is exactly what worries me. People underestimate how easy this would be. It’s not obvious to me that a robot that is much smarter than humans couldn’t do this easily. Obviously I’m not that robot, but suppose they created a virus that spread as easily as the common cold and was almost always fatal. Once it’s out there, what “logistics” are needed?

  49. Gravatar of TallDave TallDave
    15. May 2014 at 08:44

    Scott,

    It’s a legitimate concern. The best answer is probably the diminishing returns to intelligence. Let’s say your supersmart robot want to design a supervirus. Well, okay, first it’s going to need a highly sophisticated lab. Next it’s going to need to do lots and lots (and lots) of testing. And all of that requires physical resources that are increasingly monitored.

    And modern humans are actually pretty hard to kill with infectious agents anyway, because of advances in sanitation and overall health. A “virus that spread as easily as the common cold and was almost always fatal” (it would also need a long incubation period to affect large numbers of people) might even be physically impossible — even before we start amping up our immune systems artificially. And of course humans and friendly AI would immediately pour immense physical and computational resources into curing/quarantining/sanitizing such an illness.

    There are a lot of examples like this. For instance, suppose cars are self-driving by 2050. Obviously if someone could download a virus to every car simultaneously and change their software to start running over people, that would be very bad. So you design them to never allow updates except by physical access of the owner, with root commands written unalterably into the hardware. And you can be sure something like this will become the law.

  50. Gravatar of Tom Brown Tom Brown
    15. May 2014 at 09:16

    Perhaps it’s not a virus, but a takeover which happens more gradually at first. I think our war machines will become more and more automated. At some point we won’t have any pilots in aircraft, or drivers in ground robots. Likewise on the sea. The push to automate armed forces around the world will be hard to resist… then war becomes more and more a game of who’s got the fastest processing… with humans judgement increasingly being pushed out of the loop because it consistently proves to be the weakest, slowest, least intelligent and most vulnerable link. It’ll be harder and harder to justify a human in the loop: it will be like having a Commodore 64 at the critical node of your otherwise highly sophisticated computer network. When all the human brain power on Earth combined is less than 1/100th the computing power of a $1000 machine, our days will be numbered. But then again, good riddance to bad rubbish.

  51. Gravatar of Daniel Daniel
    15. May 2014 at 13:21

    Maybe we shouldn’t worry about sentient machines at all

    http://www.newscientist.com/article/dn25560

  52. Gravatar of TallDave TallDave
    15. May 2014 at 13:23

    At some point we won’t have any pilots in aircraft, or drivers in ground robots.

    I think modern militaries will remain committed to a “human finger on every trigger.” The human may become increasingly distant, but most of the automation will be too, so you can still just unplug the darned thing.

    Most warfighting is asymmetrical now, so humans in the loop have little cost to efficicacy — it’s not like the Taliban is building drones. If we start having wars between major industrial powers again what you envision might happen, but that seems unlikely at this point.

    Again, too, diminishing returns to intelligence — here’s an example:

    http://www.bloombergview.com/articles/2014-05-15/why-antibiotic-dreams-don-t-come-true

    Even if you had a thousand times your current processing power, you still wouldn’t know more about bacterial biochemistry than everyone else. You would obtain a huge advantage in interpreting data, but the data on bacterial biochemistry are currently just as unavailable either way.

  53. Gravatar of Tom Brown Tom Brown
    15. May 2014 at 14:16

    Another possible way for us to be “disappeared” is by merging with machines:

    http://www.kurzweilai.net/neuroscientists-plant-false-memories-in-the-brain

    If we get to the point where we’re saving the state of our brain and then making copies of it in more robust hardware … we’ll increasingly be less dependent on the current system: raising a disease prone human infant, and ever so slowly attempting to program it with decades of education, only to ultimately fail in most cases to create a computing device with truly significant advantages over the competition. And even in rare cases where success is had, the brain ultimately dies off after just a few decades of useful service life, with no backup copies. But as we continue to divorce ourselves from our bodies, there’ll be less and less need to maintain any connection to the old meat-based hardware paradigm whatsoever. And of course only a tiny minority need go this machine “enhancement” route to be utterly dominant over the rest of us. Think The Terminator meets Idiocracy: simultaneous trends towards the masses becoming more and more stupid while the elite become more and more less human. The vast majority will most likely just need to be be exterminated at some point.

    Trying to accomplish programming, robust information storage, hardware and software upgrades, redesigns, and fast and accurate computing using meat-based hardware or partial meat-based hardware is ultimately a terrible way to go. The competitive edge is sure to go to those beings who can most quickly churn out new effective and robust designs, get production up and running fastest, and that can eliminate as much as possible any vestiges of their slow, inferior human past.

    I think we are getting close to the breaking point of the old paradigm one way or the other. Certainly when you think in a geological time frame anyway! (even if there’s some sort of major world wide calamity that temporarily reserves some of the recent gains for a handful of centuries or so)

  54. Gravatar of ssumner ssumner
    16. May 2014 at 04:50

    Talldave, You said;

    “It’s a legitimate concern. The best answer is probably the diminishing returns to intelligence. Let’s say your supersmart robot want to design a supervirus. Well, okay, first it’s going to need a highly sophisticated lab. Next it’s going to need to do lots and lots (and lots) of testing. And all of that requires physical resources that are increasingly monitored.”

    Sorry, but you are describing the situation now, not in 200 years when viruses will be able to be modeled without testing.

    You said:

    “And modern humans are actually pretty hard to kill with infectious agents anyway, because of advances in sanitation and overall health.”

    But they’d be really easy to kill with a virus engineered to be as contagious as a common cold, and as fatal as something like bird flu.

  55. Gravatar of TallDave TallDave
    20. May 2014 at 07:58

    Scott,

    Physics won’t change in 200 years. Even if you have a perfect molecular-level modelling capability (and if you do, you probably have zero concerns about viruses because you’ve already built vastly over-engineered immunity into everyone), you still need lots of arms and fingers and fancy chemicals and highly sophisticated equipment to actually build the virus. Again, 99% of resources will be employed in helpful ways.

    But they’d be really easy to kill with a virus engineered to be as contagious as a common cold, and as fatal as something like bird flu.

    No, because all the carriers would be identified and quarantined until the short time it took them to die. You would need a long incubation period. And we’d probably have inoculations within weeks. And that’s before we start talking about the mechanics of protein folding that probably prevent the proper combination of lethality, incubation, and tranmissability — rhinoviruses are simple, tiny things.

    http://en.wikipedia.org/wiki/Rhinovirus

  56. Gravatar of TallDave TallDave
    20. May 2014 at 10:04

    The vast majority will most likely just need to be be exterminated at some point.

    That would seem to offer minimal utility. It’s much more likely that, then as now, those who choose to live primitive lifestyles will be allowed to do so, as the cost of subsidizing them is minimal relative to the increasing wealth of the technologists.

Leave a Reply