First we have the guy in Kentucky who shot two people, then the mad bomber of the van who sent bombs to various people on the left of Trump, and currently the Pittsburgh synagogue shooter who's more right than Trump.
I find some solace in the idea these three events are examples of social contagion, of fads. It's similar to the spread of anti-vaccine theories, or the sudden popularity of a set of names for newborns. Somehow we humans are monkey-see, monkey-do (with my apologies to our simian cousins) people. I'm not sure whether we just like to follow the path beaten down by others or also we like to outdo each other.
Where does rhetoric come into play? I'm not sure. Maybe it's similar to a flu or measles epidemic. One condition, necessary but not sufficient, is the existence of an unvaccinated population, a set of people closely connected enough to support the spread of a disease. The other condition is the introduction of a carrier of a virus/bacteria which is infectious.
But the metaphor isn't good enough--there's just two conditions going on. With our recent events there's more conditions: the availability of guns, the availability of bomb technology (knowledge and materials), the existence of people somewhat (or very) nutty, the knowledge that others share the feelings and conceivably can be impressed by deeds, the triggering event, etc.
Saturday, October 27, 2018
Friday, October 26, 2018
AI, Dark Swans, Google Map Directions, and Moore's Law
Some further thoughts on the Google map episode I described yesterday.
It strikes me that AI in general will have problems with "dark swan" events. That's true by definition--if AI is trained by using a big database of past information, it can't be trained to handle events which aren't reflected in the database. In many cases, like autonomous cars, the algorithms can be set to do nothing. It would be similar to the "dead man switch" found on locomotives--if there's no longer an intelligence at the controls which can react properly to events, you stop the engine. But in other cases it may be harder to define such cases.
In cases like I encountered yesterday there may be salvation found in Moore's Law. Presumably Google samples current traffic volumes using some priority rules--sample most often the most traveled routes. But priorities are needed only when resources are scarce; as technology becomes cheaper it would be possible to sample everything all the time (which is almost what the human sensory system does).
It strikes me that AI in general will have problems with "dark swan" events. That's true by definition--if AI is trained by using a big database of past information, it can't be trained to handle events which aren't reflected in the database. In many cases, like autonomous cars, the algorithms can be set to do nothing. It would be similar to the "dead man switch" found on locomotives--if there's no longer an intelligence at the controls which can react properly to events, you stop the engine. But in other cases it may be harder to define such cases.
In cases like I encountered yesterday there may be salvation found in Moore's Law. Presumably Google samples current traffic volumes using some priority rules--sample most often the most traveled routes. But priorities are needed only when resources are scarce; as technology becomes cheaper it would be possible to sample everything all the time (which is almost what the human sensory system does).
Thursday, October 25, 2018
Victims of Google Map Directions
Had an interesting experience in NY last Saturday, which I blame on Google Maps and the limitations of AI.
We drove from Kingston where our B&B is located to Rhinebeck, where the NY Sheep and Wool Festival was taking place on the Dutchess County fairgrounds. That meant crossing the Hudson and taking 2-lane rural roads. I've an off-and-on experience with Google Maps--this morning it was on. So the friendly voice advised me to take a right at the first light after crossing the bridge. While my usual route would have continued on to the second light, I decided to cede my judgment to the Google gods.
All went well--we drove a few miles (maybe 8 or so) over a series of winding roads but with no traffic. Came out on Rte 9 just south of the fairgrounds but was able with a break in traffic to hand a left and get to the fairground.
Evening came and we're ready to return to Kingston. Each year they seem to organize the exit from the parking areas differently. This year I ended up heading south, not north, on Rte 9. Traffic was stop and go, mostly all stop very little go. Somehow I had to head north to the bridge. The Google voice advised me to take a right, retracing our route of the morning going the opposite direction. I did. Big mistake.
I think what the Google algorithm must do is periodically sample the times on alternative routes, and recommend the fastest. I suspect in areas such as we were in, they don't sample very often. Consequently, maybe at 4 pm the alternative route was marginally better than the main Rte 9 north. But the algorithm kept sending cars that way. The problem was likely not only the winding roads, but the light where the route met the road to the bridge. Since the big volume of traffic was on the main road (199 I think), the traffic light favored that, only permitting two or three cars at a time from the alternative to come onto the bridge road. The end result: immobility. At about the 1:30 mark I yielded to the advice of my better half, found a way to do a u-turn, and went back to Rte 9, which turned out at about 7 pm to be almost empty.
The problem IMHO is Google couldn't keep track of how many cars it had directed the alternative way compared to the carrying capacity of the road.
We drove from Kingston where our B&B is located to Rhinebeck, where the NY Sheep and Wool Festival was taking place on the Dutchess County fairgrounds. That meant crossing the Hudson and taking 2-lane rural roads. I've an off-and-on experience with Google Maps--this morning it was on. So the friendly voice advised me to take a right at the first light after crossing the bridge. While my usual route would have continued on to the second light, I decided to cede my judgment to the Google gods.
All went well--we drove a few miles (maybe 8 or so) over a series of winding roads but with no traffic. Came out on Rte 9 just south of the fairgrounds but was able with a break in traffic to hand a left and get to the fairground.
Evening came and we're ready to return to Kingston. Each year they seem to organize the exit from the parking areas differently. This year I ended up heading south, not north, on Rte 9. Traffic was stop and go, mostly all stop very little go. Somehow I had to head north to the bridge. The Google voice advised me to take a right, retracing our route of the morning going the opposite direction. I did. Big mistake.
I think what the Google algorithm must do is periodically sample the times on alternative routes, and recommend the fastest. I suspect in areas such as we were in, they don't sample very often. Consequently, maybe at 4 pm the alternative route was marginally better than the main Rte 9 north. But the algorithm kept sending cars that way. The problem was likely not only the winding roads, but the light where the route met the road to the bridge. Since the big volume of traffic was on the main road (199 I think), the traffic light favored that, only permitting two or three cars at a time from the alternative to come onto the bridge road. The end result: immobility. At about the 1:30 mark I yielded to the advice of my better half, found a way to do a u-turn, and went back to Rte 9, which turned out at about 7 pm to be almost empty.
The problem IMHO is Google couldn't keep track of how many cars it had directed the alternative way compared to the carrying capacity of the road.
Wednesday, October 24, 2018
A President Who Knows the US Seal!
I'm tempted to suggest that all presidential candidates take an exam on US iconography. (i.e., what was almost the role of the turkey, what's the significance of the direction the eagle faces, what does it hold in its claws, etc. etc.).
That would be unrealistic, although there is at least one president who could do well on the exam. Putin has got to be smart.
Thursday, October 18, 2018
Wednesday, October 17, 2018
MFP and Farmers.gov
Got a tweet announcing the latest figures on MFP applications and payments. I now can't find the tweet, not sure what's the matter.
Two things I'd like farmers.gov to do:
Two things I'd like farmers.gov to do:
- provide online access to FSA data, like the applications and payments. It seems to me that FSA administrators at each level should be watching the data. (That was true when I worked for them, but we never did. But with the centralization of the payment process it should be easy to do, and there's no privacy concerns that I can see.)
- provide a user-friendly interface to the USDA data silos. Does anyone outside USDA understand which data ERS has and which data NASS has? Damned few, is my guess. It shouldn't be too hard to present the data without regard to the organizational parents.
Tuesday, October 16, 2018
New Terms: Adult Orphans and Family Tree Completists
Learned two new terms today from reading Post and Times:
"Adult orphans". This refers to those of us, including me and my wife, who are getting old with no children, no parents, and essentially no support network. Applying a label makes the problem seem more concrete. Personally, on the one hand I'm tempted to say: "you made your bed, now lie in it." On the other hand, which I almost always have available, it's a real problem for us, and we need to figure out how to deal with it, most likely by moving to an assisted living complex which includes nursing care. BTW, googling the term results in 45,000 hits, so it's not that new.
"Family tree completists" is unique to the Times article on the ability of a site called "GEDmatch" to help identify suspects in a crime from their DNA by analyzing DNA matches from a database of relationships created by genealogical enthusiasts. For a while I was one of these--deriving great pleasure from adding another set of (remote) cousins to my genealogy. I still maintain an ancestry.com account, with a number of trees which someday I may return to
"Adult orphans". This refers to those of us, including me and my wife, who are getting old with no children, no parents, and essentially no support network. Applying a label makes the problem seem more concrete. Personally, on the one hand I'm tempted to say: "you made your bed, now lie in it." On the other hand, which I almost always have available, it's a real problem for us, and we need to figure out how to deal with it, most likely by moving to an assisted living complex which includes nursing care. BTW, googling the term results in 45,000 hits, so it's not that new.
"Family tree completists" is unique to the Times article on the ability of a site called "GEDmatch" to help identify suspects in a crime from their DNA by analyzing DNA matches from a database of relationships created by genealogical enthusiasts. For a while I was one of these--deriving great pleasure from adding another set of (remote) cousins to my genealogy. I still maintain an ancestry.com account, with a number of trees which someday I may return to
Monday, October 15, 2018
Sen. Warren--An Honest Reconsideration
I tweeted today that I was surprised by how much difference the DNA results on Sen. Warren made to me.
I'll expand here.
When Warren was coming into prominence, Megan McArdle had a blog post challenging the validity of her research on bankruptcy caused by ill health and lack of insurance. I think there was some counter from Warren's supporters. The specifics have long since vanished from memory, but it cast a shadow on my opinion of her.
Then there was the flap about whether her claim of Indian ancestry was correct and what part it played in her academic career. Again I've seen some back and forth on it.
Then she ran for the Senate and won,
So early in her political career I had formed an assessment of her as ambitious, smart, more liberal than me. And, mostly importantly, so ambitious she might have pushed the boundaries of academic research and made unfounded claims to advance in academia. I must also admit to possible chauvinism, though I'd state it as saying her personality struck me as unlikely to appeal to moderate and male voters (so it's their prejudice, I remain innocent. :-0) Taken altogether it made a package I was reluctant to support for the presidency.
But now I know Warren had a solid basis for claiming Native American ancestry. Somehow that makes me more comfortable with the idea, supported I think by Boston Globe reporting, that she never used the claim to advance in her career. (Though her employers may have used it in their EEO reporting.) That makes her less ambitious, or at least not breaking rules in her ambition, which makes me more comfortable in supporting her in the future. (It's possible, even likely, my standards are different for male versus female politicians.) And there may be a cascading effect--I'm now thinking about her senatorial career and positions more. And that helps her.
I've tried to be honest with the above. I don't know enough about Bayesian analysis to apply it to my changing position.
So, my preferred Dem nominee for 2020 is still Klobuchar/Hickenlooper, but if Warren runs and shows up well in trial runs against Trump in the polls, I'll be a more enthusiastic supporter.
But my bottomline is still: we must win in 2020.
I'll expand here.
When Warren was coming into prominence, Megan McArdle had a blog post challenging the validity of her research on bankruptcy caused by ill health and lack of insurance. I think there was some counter from Warren's supporters. The specifics have long since vanished from memory, but it cast a shadow on my opinion of her.
Then there was the flap about whether her claim of Indian ancestry was correct and what part it played in her academic career. Again I've seen some back and forth on it.
Then she ran for the Senate and won,
So early in her political career I had formed an assessment of her as ambitious, smart, more liberal than me. And, mostly importantly, so ambitious she might have pushed the boundaries of academic research and made unfounded claims to advance in academia. I must also admit to possible chauvinism, though I'd state it as saying her personality struck me as unlikely to appeal to moderate and male voters (so it's their prejudice, I remain innocent. :-0) Taken altogether it made a package I was reluctant to support for the presidency.
But now I know Warren had a solid basis for claiming Native American ancestry. Somehow that makes me more comfortable with the idea, supported I think by Boston Globe reporting, that she never used the claim to advance in her career. (Though her employers may have used it in their EEO reporting.) That makes her less ambitious, or at least not breaking rules in her ambition, which makes me more comfortable in supporting her in the future. (It's possible, even likely, my standards are different for male versus female politicians.) And there may be a cascading effect--I'm now thinking about her senatorial career and positions more. And that helps her.
I've tried to be honest with the above. I don't know enough about Bayesian analysis to apply it to my changing position.
So, my preferred Dem nominee for 2020 is still Klobuchar/Hickenlooper, but if Warren runs and shows up well in trial runs against Trump in the polls, I'll be a more enthusiastic supporter.
But my bottomline is still: we must win in 2020.
Sunday, October 14, 2018
My Preferences for 2020
There's a poll out on Democratic preferences for their presidential candidate for 2020--Biden leads.
He's not my preference. Based on what I know now, I'd prefer Sen. Klobuchar or Gov. Hickenlooper, who fit a pattern of moderate left, which is my sweet spot. It's not that I necessarily object to some of the more radical proposals on the left, but my priority is always the need to win the election. I usually feel that the very partisan people on the left, as on the right, overestimate the popularity of their ideas and that slow and steady beats fast and flashy.
So my bottom line for 2020--I want some one to win the nomination who looks likely to beat President Trump. IMHO it should be easy, but I've no confidence it will be. See this NYTimes piece on suburban white men rallying to his support, even though they recognize his personal failings.
Saturday, October 13, 2018
Uniformity and Diversity--Amazon's Kindle
I've mentioned my cousin's book, Dueling Dragons. As part of my help to her I've gotten a fair amount of exposure to Amazon's Kindle Direct Publishing operation. (The book was published iby CreateSpace, which Amazon bought years ago and now has dropped in favor of KDP.)
With the paperback version out, we now have to worry about the ebook version. This leads me into some thoughts about the whole publishing process. In the old world of publishing, say circa 1960, each hardcover book was handcrafted with lots of choices in its design and packaging. The paperbacks were a bit different with less variety, especially in the cases where a publisher had a series going. (I remember Ballantine's series of World War II histories as one example, or a series of John D. MacDonald's novels.)
I paid very limited attention to self-publishing. It was around, and advertised in the pages of the NY York Times Book Review. I think it required a rather hefty payment to get a batch of your book printed and available for sale.
These days with Amazon ebook publishing you have very limited choices in font and design. But what this standardization does, along with the support of software and the internet, is enable a much greater variety in the content of books, partially because the costs of publishing in ebook format are so low. Because the entry cost is low as long as you can live the with limited choices everyone and her brother can publish that book they've dreamed of.
This interplay of uniformity and diversity fascinates me, and I think you can find similar patterns in other areas.
With the paperback version out, we now have to worry about the ebook version. This leads me into some thoughts about the whole publishing process. In the old world of publishing, say circa 1960, each hardcover book was handcrafted with lots of choices in its design and packaging. The paperbacks were a bit different with less variety, especially in the cases where a publisher had a series going. (I remember Ballantine's series of World War II histories as one example, or a series of John D. MacDonald's novels.)
I paid very limited attention to self-publishing. It was around, and advertised in the pages of the NY York Times Book Review. I think it required a rather hefty payment to get a batch of your book printed and available for sale.
These days with Amazon ebook publishing you have very limited choices in font and design. But what this standardization does, along with the support of software and the internet, is enable a much greater variety in the content of books, partially because the costs of publishing in ebook format are so low. Because the entry cost is low as long as you can live the with limited choices everyone and her brother can publish that book they've dreamed of.
This interplay of uniformity and diversity fascinates me, and I think you can find similar patterns in other areas.
Subscribe to:
Posts (Atom)