academia | advice | alcohol | American Indians | architecture | art | artificial intelligence | Barnard | best | biography | bitcoin | blogging | broken umbrellas | candide | censorship | children's books | Columbia | comics | consciousness | cooking | crime | criticism | dance | data analysis | design | dishonesty | economics | education | energy | epistemology | error correction | essays | family | fashion | finance | food | foreign policy | futurism | games | gender | Georgia | health | history | inspiration | intellectual property | Israel | journalism | Judaism | labor | language | law | leadership | letters | literature | management | marketing | memoir | movies | music | mystery | mythology | New Mexico | New York | parenting | philosophy | photography | podcast | poetry | politics | prediction | product | productivity | programming | psychology | public transportation | publishing | puzzles | race | reading | recommendation | religion | reputation | RSI | Russia | sci-fi | science | sex | short stories | social justice | social media | sports | startups | statistics | teaching | technology | Texas | theater | translation | travel | trivia | tv | typography | unreliable narrators | video games | violence | war | weather | wordplay | writing

Sunday, May 29, 2016

When a robot takes your life

I think one of the underappreciated potential revolutions is a revolution in the ability for "fast, cheap, and out of control" weapons to destroy any semblance of psychological and social stability.

If you think about it, we are extremely lucky that so far people have needed to do quite a bit of planning in order to kill large numbers of people, and the easiest way to do that comma with machine guns and handguns, seems to cap out at something like 50 murders per killer.

Our great mental limitation as humans is that we insist on seeing meaning in everything. That means if something bad would be limited by social and storytelling principles like just desserts and hubris, we underestimate its danger. We sort of assume that the universe wouldn't allow there to be the ability for disturbed or motivated individuals to kill thousands of people on a whim and then do it again the next day and the next day.

Weapons are coming that require no training, no ammunition, no forensic residue, no complex sourcing--they just exploit bootstrapping patterns of self-replicability.

What exactly am I imagining is coming? Well, I'm speculating about the convergence of several trends.

The model for the threat from replication is cancer. Can you produce a cheap microbot that runs on energy it collects from sun and the air? can you make a robot that builds them? can you tell the difference between them? Can you equip a fleet of cheap drones with cheap mini-guns? Can you project a razor-thin, nearly one dimensional thin beam of focused radiation with the right power source?

The engineering of all these things is tricky, but not the physics. And while we human animals are pretty physically robust to a small chance of major trauma, we're very physically fragile to a high chance of micro damage; and that damage need involve very little matter and energy.

You might be able to kill all large animals on earth with a few thousand watts and a few pounds of matter. You might end up doing that by accident, just because your efficient goal-focused AI organizes a replicating pattern of distributed intelligence and robotics that never stops replicating.

The specific scenarios aren't so likely at this point that we can describe them in detail, at least not casual observers like me. But there are no clear barriers to this threat; and if the threat is real, there's really no way to stop it from happening.

Labels: , , ,

How long has the "party of Lincoln" really been the party of Trump?

Trump's latest poll numbers have me considering the lessons this election is teaching me about what's been really going on in US politics for decades.
There's been chatter all year about this possibly being a "realignment" election like 1856 or 1964 or 1988, but that notion has been based on the presumption that a good half of Republican voters wouldn't support Trump, and that these rejects would want a party that represents them.
Now it's looking increasingly likely that Republican voters and politicians will support Trump after all, and that the election will break down under traditional partisan lines. Which raises the question: have liberals been right this whole time that conservativism and the GOP are based upon a reactionary desire for more oppression?
Even if Trump loses, a majority of politicians at all levels nationally have been elected by voters who now say they're willing to vote for Trump. Which is terrifying, and has been terrifying for years.
Compare the wide appeal of apocalyptic politicians like Trump, Cruz, and Palin--with policy proposals that are not even designed to make sense--to the absence of any comparable politicians on the left. The Republican voter base has turned out to include a shocking number of people who want these horsemen of the apocalypse in power.
Not that there haven't been some decent, Jack Kemp-style wonks involved over the years, but it turns out that the progressives were right: these have been little more than stalking-horses for a mass of ecstatically repressive, aggressively closed-minded voters.
And this apocalyptically destructive appeal has always been there. Look back at US electoral history since 1960 through the lens of the 85% of self identified conservatives who say they'll vote for Trump; it seems like a decades-long, extended struggle of pluralistic civilization against a party that most wants, deep down, to root out and purge that pluralism in holy fire.

Labels: , ,

Friday, May 13, 2016

PocketCasts UI criticisms

A lot of the problems I have with PocketCasts (which I do love overall) are with unnecessarily complex and counterintuitive aspects of UI and UX.


1. Cannot browse past episode titles and dates:


Generally, I cannot view past episode titles and dates without first turning off automatic cleanup, which makes no sense. Why not just store a bit of metadata--titles, length, show notes from the last 50 episodes of each podcast you subscribe to, and fetch anything older over the network if the user scrolls that far? As it is, I frequently leave the app and use other apps or the web just to see the titles and dates of past episodes! And/or I lose my cleanup settings over and over. (Not to mention that I can't browse these titles at all if i'm not in a location with good connectivity!)


2. Fiddly auto-download settings


I spend lots of time fiddling with the podcast-by-podcast auto download and auto cleanup settings. This is an enormous pain when you follow 25 podcasts like I do. As it is, if there is an episode that isn't automatically downloading that I see get my, and I want that podcast to automatically download and cleanup, it takes 10 taps (!) to make that change and navigate back to the feed where I was.


I would much rather have a total storage size and have the app intelligently manage the downloading and cleanup. This would take a bit of predictive logic, but that logic could do a pretty good job without being very complex. For instance, you could start by defaulting podcasts to automatically download and maintain just the most recent episode. Then, rank podcasts internally by time spent listening, and just map this ranking statically to a set of numbers of episodes to maintain and download. (Eg, the 4th most-listened podcast gets 4 episodes downloaded and cached at any one time... the 20th most-listened podcast gets 1 episode downloaded and cached. Historical minutes spent listening could expire after 3 months.)


You could improve this by allowing users to make any podcast a favorite by starring the whole podcast, with the most-listened 5 automatically receiving stars; then a user could un-star if they really don't want to download automatically, which would clear the listening count history.


3. Playlists too fiddly


Similarly, I would love to have smart playlists that just use that same priority ranking and play whatever is both high priority and recent. There are several podcasts where I listen to essentially every single episode. Shouldn't I be able to just open the app and hit "smart play" and have it start playing stuff i like? You could keep this limited to, say, 10 episodes, and many users wouldn't use it at all, but for heavy listeners it could be a killer feature.

Playlist UI never does quite what I expect. I can "play all" but I really just want to play stuff that doesn't have to stream over cell data.

4. Navigation mysteries

You can't get to the screen for a podcast from the screen where an episode is playing--there is no link to the podcast as a whole.

There's a bug in the playlist screen, screenshot attached.

The icon change that happens to episodes in the feed depending on playlist state is mystifying. If I have an empty playlist, the episodes all show a downloaded state or a streamable state. Fine. Then I chose play all, and now the icons all show a minus, to remove them from the queue. It's hard then to see the downloaded state. If I then later clear the playlist, the icons have a plus to add to the queue, instead of going back to the original normal state. It's all counterintuitive and bizarre.

It's never been clear to me how to get to the currently playing playlist screen from the episode feed screen. 

5. Partially played mysteries

It's silly for there to be separate unplayed and in progress feeds. That just doesn't reflect how I listen to podcasts. I want to see podcasts in progress mixed in with unplayed podcasts. And that's what PocketCasts does show me... sometimes! Sometimes a partially played episode leaves the unplayed feed and sometimes it doesn't. Why? Beats me.

Partial downloads do not present with consistent appearance and UI state; often I'll download half of an episode and then lose connectivity, only to find that the state of the episode has been shifted to an error, and all the content downloaded so far is inaccessible. Other times, a partially downloaded episode will present itself with a play icon.

6. Forgetfulness

The app forgets my place in a podcast often. This seems to maybe happen when playback is  interrupted by playlist actions for relegation to the in progress list?

7. Swiping sucks

I appreciate that swiping is a popular UI element in theory, at least for social apps we spend all day on. But there's a very good reason it hasn't been a bigger part of material design, and that is because there is too much ambiguity over what the adjacent screens are, and too much ambiguity over when a touch is intended as a full screen swipe vs a drag or slightly ham fisted tap. 

One of the things I do most often is play an episode and attempt to advance the place in the episode to some later point. Maybe I started listening to the podcast on Sonos and I'm switching to the app, maybe I want to skip to a part of the episode I saw in the show notes, maybe PocketCasts just forgot my position. More than half the time I do this it interprets my motion as a full screen swipe and takes me to a different view.

Labels: , , ,

Tuesday, May 10, 2016

When a robot takes your job, cont'd

I think there's a fundamental error in assuming that lack of productive employment among much of the population necessarilyequals poverty.
I actually think that the scenario where technology grows our collective productivity so well that few of us have marginally productive capability is a very good thing for the wealth of the poor--at least in the moderately democratic first world.
Why? Because it means there's way more wealth in general. Yes, in an autocracy the unproductive could be left to starve... but in any slight democracy there will be constant political work being done to share the wealth.
Much of that work will be ideological and identity-based cover for what is effectively a clumsy, haphazard socialism. Hence you have the anti-government Nevada racist rancher guy who happily takes government land subsidy checks and advocated for massive government seizure of labor wealth and private wealth to pay for a security state. 
Isn't this the old argument for supply-side, or trickle-down, economics? Not quite. That argument was an argument for redirecting wealth away from the welfare state and towards the already wealthy, with the justification that this would indirectly serve the working class even better than direct subsidies.
My argument is that overall productivity growth, which expands the whole pie, will make up for the reduction in wealth of the working class and poor due to replacement of unskilled and semi-skilled labor due to computerization. This process will be painful and problematic, but not apocalyptic.
We'll be amusing ourselves to death, but we won't be starving to death. At least in the first world, where there is a practical cost to too much poverty.

Labels: , , , ,

Blogger Benjamin Chambers on Tue May 10, 09:44:00 PM:
Well ... maybe. I just wrote you a longish and nicely-reasoned response, but my blinking browser chose to crash when I tried to post it. Perhaps tomorrow.
 
Blogger Ben on Wed May 11, 11:37:00 AM:
Sorry to hear that! I'd love to know your thoughts!
 
Blogger Benjamin Chambers on Wed May 11, 02:41:00 PM:
What I said in my now-vanished comment, basically, was that you seem to assume that automation will only affect unskilled and low-skilled jobs, but that all the commentary I've seen on the widespread automation of labor argues that almost *all* jobs will be eliminated (except caring professions like childcare and elderly care), including jobs we currently think of as impossible to automate. This is why universal basic income is being seriously discussed on both the left and the right.

I can't assess how likely this outcome actually is, but the automation of some aspects of high-skilled jobs like law and medicine makes me think it's at least a plausible scenario. But let's say it happens. Human beings need a purpose to their lives, and a sense that we're contributing to something -- for most of us, work fills that bill. It's obviously foolish to imagine that the vast majority of people will suddenly amuse themselves with philosophy and the arts; most of us aren't even that good at providing direction to our lives for long periods of time (though we all think we'd be good at unending leisure). So what do people do? I think a mass lack of purpose and sense of powerlessness are likely to lead to increased xenophobia, chaos, violence, and war ... a surefire recipe for poverty. (That's not even taking into account vast income disparities, probable mass movements of climate change refugees (the Syrian refugee crisis, according to one view, has its roots in the drought that drove massive numbers of agricultural workers into the cities), and the increasing scarcity of clean water.)

Maybe this is too dark a view -- it's always easy to see doomsday on the horizon, and would-be Cassandras are often wrong -- but I think our own democracy is less and less interested in "sharing the wealth" (even among those who would most benefit from it), and with the capacity (and willingness) of large corporations and the very rich to disconnect themselves from the United States, there's less and less political motivation to care about the disenfranchised -- especially if you need very few of them for labor.
 
Blogger Ben on Thu May 12, 11:00:00 AM:
I agree with your economic predictions but not your philosophical and social ones! There is so much room for a sense of meaning, purpose, direction and accomplishment without there needing to be an economic reason for your work at all. Just look at the army! Many army jobs are the equivalent of digging a hole and filling it up again. And yet the army is known as a place where people find direction. And if you don't think many more people could be satisfied as artisans of some sort, try asking well paid lawyers and finance and restaurant workers and the unemployed what they really want to do... very many would like to do boutique, artisanal curation or creation.

Cory Doctorow's Down and Out in the Magic Kingdom is a good argument in this direction.
 
Blogger Benjamin Chambers on Mon May 16, 01:38:00 AM:
(Part 1/2)In spite of the stereotypes -- all the wait staff who want to be actors, etc. -- and people's common avowals that they would love to write a book if only they didn't have to work, etc., I'm unconvinced that a vast number of people would take up some form of art or artisanship to give their lives meaning, in the absence of employment.

On the plus side, creativity comes in a million forms, and I think if people have an inch of luxury to do so, they do find ways to apply it or express it. On the down side, the disappearance of work would wipe out a lot of culturally understood avenues for leading a meaningful life, and I simply don't have the confidence that you do that very many people would be able to build that for themselves. (I haven't read Doctorow's book (Wikipedia's summary makes it sound a bit baroque) but I'm familiar with some of his essays and a recent YA book of his, and he strikes me, fundamentally, as an optimist about people's capacity to self-organize and self-direct themselves. I could be way off base about this.) There's a huge number of people who don't value intellectual or what we now class as creative work *as* work -- in their eyes, it's effete, it's pointless.

So I don't know that I have all that much more to add here, since we simply disagree. But as I've been thinking about this over the past few days, I recognize that I have, well, more thinking to do on the subject. Needless to say, my premise that universal basic income will destroy the world of work for most people has some problems. For one thing, I haven't read enough on the subject of basic income to understand how it's supposed to be funded. For another, it raises the question: if almost everyone is receiving basic income, will there be markets to raise the capital to fund that income? Will there be enough of the rare elements needed for processors that will run the AIs that are supposed to take our jobs? The entertainment industry might grow further -- we'll still need human actors and many of the supporting technicians; some direct care jobs in the health industry would likely still be human; and there are a large number of government jobs that would, I think, still probably be staffed by humans: public health, community justice (probation, parole, law enforcement), mental health, alcohol and drug treatment, and support of the elderly and developmentally delayed. These examples aren't a huge portion of the labor force, but I'm guessing that there are quite a few sectors of the economy where AIs either wouldn't do the job we wanted, or we wouldn't be likely to let them.
 
Blogger Benjamin Chambers on Mon May 16, 01:39:00 AM:
(Part 2/2) And from another angle: how do people who can't or don't need to work *now* fill their time and their lives with meaning? I'm sure studies have been done of what retirees or folks on unemployment, SSI, or other subsidies do, and how they feel, though I'm unfamiliar with them in any detail. Retirees have the advantage that there is a cultural niche for them -- I think many of them struggle with feelings of irrelevance and not contributing, but there is a strong belief that they have "earned" their retirement, that their status is something to be coveted, and I'm not sure how that changes when everyone is "retired," albeit without money to travel, golf, finally rebuild the engine in that 57 Chevy, become a patron of the arts, etc. The unemployed, I think, mostly don't want to be -- if for no other reason than they want to contribute, provide for their families, etc. -- but after a while, if you can't get work, what do you do with your time? I'm not sure of the answers here. All I'm really saying is that I don't know, but the answers might have a bearing on what would happen if the ranks of the unemployed/retired poor grew exponentially. It could lead to widespread unrest ... or not. I really don't know.

I just think that we're talking about a potentially enormous cultural shift in how we value ourselves and our roles in society -- we're talking about eliminating, for most people, potentially, the pathways they have to participate in culturally recognized and valued ways in that society, and not giving them anything in return except subsistence. The work of creating new roles that are valued will be, I fear, hard, chaotic and violent, and therefore, an economic disaster for most. But who knows? I don't.
 
Blogger Ben on Mon May 16, 02:11:00 PM:
You make some really great points, and I appreciate your perspective. I might be overestimating how many people dream of being some sort of business owning artisan. But my broader point is that being rich enough not to have to work is something that for most people is basically liberating, not basically oppressive. The people whose work is farthest from being artisanal -- unskilled service laborers -- are largely working mindnumbing jobs that they don't leap out of bed for in the morning. Of course these jobs do provide them a sense of being valuable, require them to structure their day and their appearance, etc., but I don't think there is something magically better for them than just being paid the money and given infinite days off :)

But I see your point that subsistence income is not a substitute for the income from a real job like, say, being a bus driver or a typical clerical worker. Here's where I appeal to the democracy, or pseudo-democracy, that a quasi-empire like the United States enjoys: the power structure is willing to spend money to buy off the population and increase social stability, in part because the people really are part part of the power structure. (Even antigovernment folks insist that Social Security and Medicare not be touched, so they aren't.) Maybe in our technological future that means there have to be all sorts of fake jobs concocted to keep people docile, but I think the power structure is completely capable of delivering that, especially with all of the added wealth that technological innovation brings.

So I expect inequality to keep increasing drastically, but because the whole pie is increasing so fast and because we have something of a "chicken in every pot" democratic politics, I think robot capabilities are not going to increase poverty in the United States or the rest of the first world.
 

Sunday, May 01, 2016

The guacamole recipe that launched 1000 ships

By popular demand... this is the guacamole recipe I started making as a bartender for Manhattan penthouse parties in college, and which I've evolved over the years through trial and error (and accident, as when I kept adding "salt" and couldn't taste it but wondered why the guacamole tasted so sweet!)
I guarantee that any group of 15+ people will finish it and scrape the bowl.
Get a large bowl, and mix the following ingredients:
* 5 or so avocados
Make sure you get avocados that are just ripe, which can be tricky. you want them to be a little bit smushy, but not so smushy that there's any dark brown color when you open them up. if your supermarket only has very hard avocados, the best thing to do is to buy them 3 days before you want to make the guacamole, and they will ripen just sitting on your kitchen counter. if you want them to ripen quickly, people say to put them together in a brown paper bag.
Cut open the avocados by slicing in a circle from the top to the bottom and continuing around until you reach the beginning of your cut again.
Twist the two halves and pull them apart; one of the two halves will have the pit in it. The easiest way to get the pit out is to hack a big kitchen knife right into it so it goes in pretty deeply, then twist the knife to the side (don't tilt it, twist it like clockwise or counterclockwise) and the pit will come out still attached to the knife. Then you can just whack the knife against your trash can and the pit should fly off into the trash. (Keep one pit to pit in the middle of the guacamole when you're done making it, as a decoration.)
Now take a large soup spoon and scoop out the flesh of the avocado into a big bowl. You can dig the spoon all the way against the avocado skin so you get all of the flesh. If you're doing this right it shouldn't take more than a few seconds to get most of the avocado flesh out of the skin. discard the skin. If there are any dark brown parts of the avocado, cut those away and discard them, or your guacamole will taste like it's gone bad.
* limes: one for every 2 avocados.
This is as unusually lime-heavy guacamole recipe!
You can juice the limes, or just cut them in half, stick a fork in, and crush the half around the form while grinding it back and forth; it's great to get some lime pulp in there.
Add the limes right away, because the citric acid will keep the avocado from browning in the air.
* 4 shallots (because of their mildness), or 1 medium white/yellow onion (because of its sharpness), or a combination (my preference)
Mince these very finely. The idea is that you want to taste them, but not have them be overpowering.
* several large cloves of garlic, depending on how garlicky you like it
I put in about half a small head of garlic -- maybe 5 cloves. Mince finely, or even better, mash it with the salt in a garlic press or with a mortar and pestle (or even just the bottom of a mug, and then mince it) so that the salt soaks up the garlic juice.
(Another option is to take a whole head of garlic, pull off most of the skin, wrap the cloves up in tin foil with a dash of vegetable oil, and

roast it in the oven on 400 degrees for 15 minutes -- the garlic will
be sweeter and have a deep, rich flavor. You might want to mince some fresh garlic as well to keep some bite.)

* salt: to taste, about 3 teaspoons
I recommend kosher salt, but any salt will do.
This is the easiest step to get wrong by using too much or too little, so keep tasting it and adding salt slowly. Ask other people to taste it.
Make sure you try it with the chips you are serving it with, because if they are very salty you don't need as much salt.
* sugar: to taste, about 1 teaspoon
This is a secret ingredient. You don't want to add so much that it is noticeably sweet, you just want to take the edge off the onion and lime a bit.
Also:
Mix with a fork. Don't mix forever, because it's fun when there are chunks of avocado.
You can add chopped tomato or sour cream or neither... totally optional. Don't add too much of either.
The bowl you mix it in will be pretty messy, so you might want to clean up the edge with a paper towel or transfer it to a more pretty bowl.

Labels: