Fork me on GitHub
#off-topic
<
2018-03-20
>
qqq00:03:35

According to https://www.usgovernmentrevenue.com/total the current US total taxes is 6-7 trillion. I was reading an article about a $1000 / person / month basic income. 300 M 12 $1000 = 3.6 Trillion So that type of basic income -- would have to increase US taxes by 50% ?

gonewest81801:03:49

That assumes 300M people need full assistance to achieve the basic income, but current census puts it more like 50M assuming a poverty level of $12k per individual / $24k per household. Other federal assistance programs (WIC, Medicaid, social services) might also be playing an overlapping role and de-duplicating those programs might also offset some tax burden. I haven’t seen any studies on that so I’m clearly hand-waving. Seems likely you’re overestimating costs, maybe a factor between 3 and 5 higher than actual. On top of which, you could theorize other “knock-on effects” in terms of crime rate, education and bootstrapping effects, improved diet and wellbeing (therefore improved workplace productivity, etc) impacting other line items in the budget. Naturally there would be vehement disagreement whether such knock-on effects actually exist.

qqq01:03:05

In my limited understanding, the U in UBI stands for 'Universal' and the intent is to hand it out to everyone (to avoid issues of 'oh, if you apply for a job, you'll end up getting net less money due to losing your basic income')

tbaldridge01:03:17

Unless we saw a massive upheaval in the job market in the US, and by that I mean robots replacing a good 10% of jobs, I doubt UBI would ever take off. There's a quote I heard recently that I think applies:

tbaldridge01:03:36

>"Socialism never took root in America because the poor see themselves not as an exploited proletariat but as temporarily embarrassed millionaires." -John Steinbeck

tbaldridge01:03:05

In other words, most low income people (how my family was, growing up), don't see a need for such programs because they don't consider these people any worse off than they were or are. Growing up we were poor, but it was a matter of pride that "we're not poor enough to need food stamps".

fellshard02:03:28

That perspective has been waning over time, but it has greatly slowed socialism's ability to take root; more like one of the modern diluted European forms in the US than anything else.

gonewest81802:03:40

@qqq ah, I wasn’t following the reference.

tbaldridge02:03:58

The other issue with UBI, is really that it gives no incentive to change. I've lived very comfortably on $12/hr in somewhat remote parts of WI. Rent is about $650/mo, food for a single person can be around $350. Internet and utilities weren't that much (city ran both where I lived). And it's quite possible to have about $500 in disposable income assuming that UBI was A) untaxed and B) included a form of universal healthcare. Looking back on my life I'd have been really tempted to just do OSS "full time" or perhaps just not even bother and play games on Twitch instead.

tbaldridge03:03:32

Of course the argument is that at some point AI/robots will take over so many jobs that there won't be anything to do if you aren't a researcher or a maintenance person for the afore mentioned machines. But even then, that's a problem. Why would I want to become a robot-repair-person, when I could sit at home doing my own thing instead.

tbaldridge03:03:18

I think we may have UBI in the US someday, but I think it's 50-100 years off, depending on how fast tech evolves in that time.

fellshard03:03:21

I think we're observing what happens as that gets even remotely possible, though. Smaller and smaller 'social ills' become magnified much more greatly, well out of proportion once the deeper negatives are removed. Not that the alternate is better, but the point is that there's no free lunch, and those who become too zealous about those 'smaller ills' without realizing the alternatives may very well make it difficult to sustain UBI for any extended period.

donaldball03:03:18

It seems to be a fairly universal and fundamental human desire to be socially useful. I aver that among the problems that plague our society is that so much useful work is neither valued by the market nor paid for out of the treasury, to our net loss on many fronts. Fwiw I am not at all sure that a world in which Tim accepts UBI and does OSS work doesn’t have some real benefits over ours, in which it seems a sizable fraction on our collective software development effort is going into build surveillance capitalism almost by accident.

👍 4
qqq03:03:36

If the avg UBI recipient was as good as Tim at Clojure and writing OSS, I'd vote for UBI in a heartbeat.

sveri10:03:00

I think what most people dont see is that things, that are really needed, like elderly care worker, nurses, teacher etc will be paid much better because there is no more incentive to be working that many hours for such a shitty pay. So I would predict that ultimately the ones earning the most money are not the bankers / lawyers / etc but the people that have jobs with an actual purpose and benefit for the society.

sveri10:03:20

Also everybody (myself included) always predicts what is going to happen and depending on whats going to happen is pro or anti UBI. But I would say it is impossible to predict the outcome on such a large scale. This is one of the things that needs to be done before we can see whats actually going to happen. So I'd much rather say lets try it and hope for the best instead of being pessimistic from the beginning.

eggsyntax13:03:37

GiveDirectly (https://givedirectly.org/give-basic-income) is doing a large-scale long-term experiment in UBI. It's my favorite charity -- basically your dollars do double duty: lifting a human out of crippling poverty (for a dollar a day!), while funding an experiment that I think is really important right now. Couldn't resist repping that in light of the comments 😉

andre.stylianos10:03:10

The thing for me is that there's a tipping point where I see it breaking down. UBI requires money, which needs to be collected somehow so it's redistributed... of course there's taxes, inefficient government programs that could be replaced by UBI and so on; but the problem is that this works when 10% don't work and 90% work/generate the money, but when you tip the balance to the other side and say 10% work and 90% don't then you would need to be taxing the working ones so much that they would basically make nothing.

mpenet10:03:53

it can replace other measures

mpenet10:03:17

in switzerland one of the argument was that it could lighten the load to the various safety nets you get once you're unemployed

mpenet10:03:31

(they rejected the proposition btw)

mpenet10:03:18

then depending on the country you live in, there's no such thing or it's super low, so: it depends

gklijs11:03:53

On a global scale UBI is also tricky, how to prevent people getting into the country where there is a UBI, just for the IBU, or what if someone lives in a ‘IBU-country’ and working abroad, should he still gat the base income and/or pay extra taxes?

michaels13:03:24

@andre.stylianos in my experience, though thankfully not my present! - 10% work and 90% don’t in most jobs I’ve had anyway.

eggsyntax13:03:37

GiveDirectly (https://givedirectly.org/give-basic-income) is doing a large-scale long-term experiment in UBI. It's my favorite charity -- basically your dollars do double duty: lifting a human out of crippling poverty (for a dollar a day!), while funding an experiment that I think is really important right now. Couldn't resist repping that in light of the comments 😉

john13:03:30

Assuming UBI could work, I think it'd be smarter to give people something like a million dollar bond at birth, which starts paying out monthly at 18. That way everybody is incentivized in the success of the economy, regardless of social strata.

jgh13:03:22

that’s not a bad idea

john14:03:22

The average person is the one to thank for 8% growth every decade for the last 100 years anyway. Makes sense to have them cheering for a piece of that pie.

donaldball14:03:08

You could call it a citizenship dividend.

john14:03:35

You'd have to phase it in over the course of a half century though. Cause it'd be many, many trillions of dollars pumping into the economy.

hubert15:03:43

when we start UBI we also will save a LOT of money, because it replaces ALL social help, we do not need anymore all those (social agencies, people that work there, buildings, etc. etc.) system is simple, everyone gets UBI.

hubert15:03:05

"The Economics of Inequality" by Thomas Piketty and Arthur Goldhammer "Capital in the Twenty-First Century" by Thomas Piketty and Arthur Goldhammer

hubert15:03:27

but I believe in conspiracy theory 😉 that they will not allow to run it, because IMHO it will make much more easier for people to move to "upper classes" (please, do not think that I classify people )

bja16:03:27

Last week I decided that the internal admin (which I inherited) would be a lot easier to work on if I could just make calls to the internal endpoints for misc services directly. Previously, we had to add a custom wrapper as a flask controller for the admin to do any new thing. To that end, I added nginx proxying rules for the admin site to negotiate cors and upstream the various services with http basic in front of it. Now the admin will insert credentials into properly authenticated users admin panel and the admin js can hit the services using APIs that match what the swagger spec says. This has made my life infinitely better in the short term since it has come into being.

roklenarcic17:03:26

what's a good way or website to search for clojure jobs?

eggsyntax17:03:17

See #jobs-discuss, #jobs, and #remote-jobs (and a few other locality-specific ones IIRC).

eggsyntax17:03:47

I think maybe one of those has a pinned post that points to the major clj/s jobs websites.

eggsyntax17:03:11

If not, let me know & I'd be happy to help you find a central directory, I know there's one around somewhere.

qqq17:03:31

somewhat related to UBI: The progress of launching a SaaS has shifted over the past 20 years from 1. buy your own server, get your own network connection 2. use colocation services: just provide a machine + provide monthly rent 3. rent machines by the month; do your own admin work 4. rent machines by the hour, do your own admin work 5. rent machines by the function call, have AWS manage your dynamodb/aurora, s3, runtime -- just provide damn functions we've also seen another shift of cgi/C -> php/java -> ruby on rails and we've also seen the rise of mechanical turk question: how many years off are we, from a non-technical founder being able to build a SaaS by paying for mechanical programmers to write individual lambda functions on demand ? (related to UBI as this would eliminate many 'CRUD' programming jobs)

donaldball18:03:38

I doubt you’ll be able to build a non-trivial system that way without a well-defined system architecture expressed as one or more system-readable artifacts: code Moreover, writing specifications required for the mechanical programmers to be able to write the individual functions correctly and with the desired error handling behavior is not significantly easier than encoding those in code directly. Jobs will change, but I don’t think we’re yet on the cusp of a job-shedding transformation. Of course, we shouldn’t advocate (or not) for UBI based on self-interest.

qqq18:03:57

Let's change the problem slightly. Suppose the founder is Technical, leveraging somethin glike "mechanica turk for programmers" would he/she be able to build somehing of Pinterest/Snapchat/Instagram level complexity without any employees by just using 'mechanical turk programmers' ?

qqq18:03:15

This would still lead to massive loss of programming jobs via 1. new jobs not created and 2. old companies doing layoffs

Alex Miller (Clojure team)18:03:09

the percentage of novel code in things like Pinterest/Snapchat/Instagram is probably low

Alex Miller (Clojure team)18:03:02

there is creativity in business model and in certain technical aspects (like filters maybe?) but surely there is a big chunk of just boring web cruft

donaldball18:03:47

I don’t really think so; the bulk of the computer work I’d reckon those services require are in engineering the system, which is to say: ensuring the system is able to run with whatever constraints the business requires vis reliability, latency, cost-per-operation, etc. etc.

Alex Miller (Clojure team)18:03:20

can programs write the boring stuff? I don’t know. people have been pushing 4GLs since I started being a professional programmer, yet I still don’t feel that my job is particularly threatened by those

qqq18:03:16

@alexmiller: but you're doing compiler development right? It seems like alot of work is just "stateless functions in front of an eventually consistent db" -- and AWS will autoscale / manage that for you.

Alex Miller (Clojure team)18:03:49

I really mean “my job” as a profession not my actual job :)

Alex Miller (Clojure team)18:03:33

I don’t actually spend a very large percentage of any day writing code :)

Alex Miller (Clojure team)18:03:49

which is maybe instructive in itself

fellshard18:03:39

The amount of leverage that 'mechanical' programming can provide is relatively low. High-leverage systems require systems thinking and more than simply mechanical, procedural processes.

roklenarcic18:03:25

even if there was a solution that would make throwing together CRUD a non-programming task, various product managers and analysts still wouldn't be able to do it

john18:03:42

How many thousands of assembly programming jobs were killed by "high level" languages?

qqq18:03:48

Whether existing managers/analysts switch over is irrelevant if new startups come along and replace them with 1% the headcount.

john18:03:37

Seems like there'll always be a "higher level language" which is the one the human uses to tell the computer what to do.

fellshard18:03:00

And at some point, that involves natural language to tell the developer what to do, more or less. And that, too, can have intent lost in transmission.

john18:03:25

I just doubt we'll be able to automate that translation of human intention into code automatically, for any app

fellshard18:03:18

It takes rigor, volatility, back-and-forth, not unidirectional mechanical thinking. I follow the DDD line of thought on this quite heavily.

qqq18:03:33

I'm not saying "all programming jobs will be eliminated by cloud computing" what I am saying is: currently, some software company has 1000 employees by leveraging cloud computing, some founder develops a competitor using only 10 employees ==> 990 programming jobs eliminated

tbaldridge18:03:34

we can't even automate translation of business rules into code, really at any level.

tbaldridge18:03:58

@qqq I don't think that happens. Or more correctly it already has.

fellshard18:03:05

It's been tried, for decades.

john18:03:08

990 programmers can then do more as cloud devs, right?

tbaldridge18:03:58

Even in the cloud, you still have ops teams when you get to scale.

qqq18:03:18

@john Unlikely they will all get jobs at AWS/GCP.

qqq18:03:44

@tbaldridge: iirc, instagram was 13 employees at acquisation and whatsapp was 80 employees at acquisation, so the # of ppl required has drastically decreased

tbaldridge18:03:52

And as Kevin on Shark Tank is so fond of pointing out, if you start a business with a small team of 10, he'll hire 20 people and out-innovate you in a year.

fellshard18:03:08

Every innovation tends to generate new unforeseen complexities at the boundaries. Careful not to fall into the 'lump sum of work' fallacy.

tbaldridge18:03:11

@qqq right, and how much money has instagram made?

qqq18:03:42

@tbaldridge: if Kevin's logic was right, why did FB pay so much for instagram/whatsapp, instead of just "out-innovate" them given FB's head count?

fellshard18:03:44

That Shark Tank point has been proven time and again in Silicon Valley itself, especially if a startup is too slow on the uptake

fellshard18:03:52

Not all will, but most do

tbaldridge18:03:58

@qqq they bought a brand

john18:03:40

@fellshard exactly. programming is an inexact science because it is about worldly problems as much as computer problems. There are always more problems for the humans to work on optimizing.

qqq18:03:42

@tbaldridge: I think the phrase' out innovate' means different things to us (to me, it means -- win the market)

fellshard18:03:33

"Everything you can do, I can do better" - this applies especially when the idea is new, but the execution has obvious lessons someone else can learn from

fellshard18:03:34

Startup Alpha had this cool new idea, but was immature in its creation. Startup Beta now sees a fuller picture of the problem Alpha tried to solve and comes up with a much more strategic approach, undercutting Alpha in a heartbeat.

fellshard18:03:54

Not a new story by any stretch

tbaldridge18:03:37

We see it all the time in the gaming market. PUBG was created by a veteran in the industry (3rd Battle Royale game he made), and it had a huge following. But the game was buggy and lacked the funding to improve it much more. Epic comes along and releases Fortnite and in about 6 months utterly destroys PUBG. Biggest factor there was game produced by a huge company with capital and knowledge, vs the scrappy startup.

fellshard19:03:45

'managed copy and paste'

fellshard19:03:02

No central ownership of contracts, no way to prevent one team breaking another team; it's like a swarm of devs pushing to master, continuous integration without gating or ownership

seancorfield21:03:50

@mbjarland What Slack client are you using?

mbjarland21:03:07

@seancorfield “2.5.1 Direct Download - Walking on a dream” 😂 according to the about box, this is on osx. Ok trying to install the latest from app store…

mbjarland21:03:28

problem solved after client update

dominicm22:03:54

A weird idea I've been pondering. Does a rule engine make sense for cloudformation/terraform json generation? I find that it's more often that I have a set of facts. E.g. There's a server, and it's of type m2.small. And it's DNS route is "http://example.com". Then the rule would see the DNS route and create a corresponding route53 set of facts. Then you might say that 123 is the certificate id for http://example.com. And that the load balancer serves http://example.com is a fact. Then a rule will say it's ssl certificate is 123. I haven't thought this through very concretely. Does this seem logical and like it would solve anything?

dominicm22:03:33

The load balancer example should perhaps be that it points at the website, and it traverses the information from there.

fellshard22:03:33

So reducing specialized tools down to logic programming, instead?

dominicm22:03:58

Maybe. I seem to spend a lot of time talking about separate things in AWS, which are not how I think about my facts about the world.

jgh22:03:19

as a guy with a bunch of media servers and a serious bin packing problem that sounds like an interesting way of looking at it. Throw in rules around spinning up resources with prometheus and were cooking with gas