Not a bar, but plays one on TV

nitsua60

00:00

What's the cornerstone? (Aside from the obvious?)

sorry--gotta run. Videoconference starting.

BESW

When 'Abdu'l-Bahá visited America, He laid the cornerstone for the Chicago House of Worship Himself. There are a number of stories about how He used the occasion to inspire and instruct.

A lot of people have found it a powerful spot to visit.

doppelgreener

05:36

> Autonomous vehicles will likely become available in the near future. A new article published in Science raises a classical ethical question about how these cars should be programmed: should a car sacrifice its driver if doing so will save the lives of many pedestrians?

The article found that participants generally do want cars to be programmed in this way for other drivers, but they don’t want their own cars to work this way. It’s a potentially lethal form of “Not-In-My-Backyard” for our more automated future. -- http://arstechnica.com/science/2016/06/people-want-other-peoples-self-drivi…

This is surprising to me.

I think this also means, economically, most self-driving cars will not sacrifice their owner to save pedestrians, because very few people will want to buy those cars. They want a car that will protect them over others if forced to choose. So, cars that choose others over the driver won't sell as well and, accordingly, won't be made as much.

trogdor

@doppelgreener this,... has potential to become my white whale

doppelgreener

@trogdor in what way?

trogdor

well, first of all I am not of the opinion we could possibly have autonomous vehicles in "the near future" I don't know what definition is being used there to be fair

doppelgreener

@trogdor They already exist

They're pretty good, too.

trogdor

secondly,.... programming a car to sacrifice it's owner for pedestrians, regardless of ethics if it WERE to work the way intended, raises questions for me as to whether they could actually save said pedestrians every time

@doppelgreener my point isn't "can they drive themselves" it is "are they ACTUALLY ready for traffic yet?"

doppelgreener

05:47

@trogdor Yes. These things go through dense traffic and onto highways.

trogdor

@doppelgreener that does not automatically mean they are ready to do so on a mass scale

doppelgreener

There haven't been many studies on actual accident rates, but there isn't much of a large sample size.

Since they're being produced on a mass scale (Tesla will be producing 2-3 hundred thousand Model S cars over the next few years, which will all be self-driving) we're gonna find out pretty soon one way or another.

trogdor

don't get me wrong here, I don't hate the idea of self driving cars itself, but I do personally feel like a few too many people are expecting them to be a thing on a mass scale "really soon" and if that really soon is just a little too soon, people WILL die as a result specifically of these cars messing up

doppelgreener

It's a "ready or not, here it comes" scenario

trogdor

@doppelgreener that right there is the problem

anyway, I did say White Whale

doppelgreener

05:57

Also, the USA's... uh... road authority, whatever the name is, has required that the makers of self-driving cars have to report all accidents to the government. So far Google's self-driving cars have been all over the USA and had a small few relatively harmless bumps.

trogdor

part of my implication was that I might not be entirely unbiased here

and I only brought it up after an implication of a really heavy ethics problem related to them

doppelgreener

Fair 'nuff. Well since it is happening either way, and they've generally worked out pretty well so far, and they are in mass production and will be on the roads on a mass scale, we'll start gathering a lot of data.

trogdor

it's actually one thing if they can drive normally and avoid accidents, it's a completely different ballpark if we give them the equivalent of "choice" to kill one person, even if it is to save other people

doppelgreener

It's guaranteed people will die. People die on the roads all the time as a fact. Many of those deaths will be people inside self-driving cars and many of those deaths will be people who were hit by a self-driving car. I think the important thing to look out for is not "are people dying?" (that's guaranteed to be affirmed) but "how much are people dying, compared to ordinary non-self-driven cars?"

Also, all self-driving cars still have the pedal and wheel and must have a driver in the seat who can take over immediately. S'part of the laws for car manufacture. There are no cars yet that are completely self-driven.

... Actually that might've changed. Or might be changing. Not sure.

trogdor

@doppelgreener that is good, but how does that tie in to the ethics problem then? why is it even a question of them being able to do this thing if the driver can immediately stop them?

doppelgreener

06:01

@trogdor What if the driver's asleep, or reading a book?

trogdor

@doppelgreener fair enough

doppelgreener

definitely still a relevant ethical question, yeah

trogdor

@doppelgreener I was not implying it wasn't, I was literally asking why anyone would consider putting a feature like that in a car when the driver will obviously veto that decision out of pure survival instinct (and you answered that question)

doppelgreener

Oh right!

mm.

But I think the simple economical thing will be the decider though.

I don't want a car that will kill me instead of others. I want a car that will protect me, because I'm the only me I've got, and I would rather stay alive given the choice between me and some random people. (Sorry, random people, survival instincts are selfish.)

So, if I buy a self-driving car, I'll buy one that protects me.

Each other person feels this way. So, each other person buys a car that protects them instead of other people, given the necessity to make a choice.

trogdor

I want to reiterate, I am really cool with this idea of self driving cars if they save more lives than they take, the thing that bothers me is if we decide they are 100% ready when they aren't and more people die than have to as a result

doppelgreener

06:05

So, everyone has cars that protect them. The cars that don't protect the driver - that choose the pedestrians instead - sit unsold on the sales floor. They stop being manufactured as much, or being bought by dealers as much.

@trogdor Oh, yeah, definitely people don't think they're 100% ready yet. They're ready enough. (And, arguably, better drivers than an awful lot of humans who we happily hand licenses to!) But there's lots of improvements still to make.

trogdor

do I think we can make cars that can drive themselves well, even better than people drive? heck yes, do I think they are already ready to do so right now? I would have to say I am skeptical

@doppelgreener I agree that there are certainly people on the road who are way less qualified to drive than the licence in their pockets would imply, but I see no scenario where we give only those people these cars unerringly XD

regardless of the fact that theses cars are probably decent drivers right now, I would bet they could be better if they were to wait a few years, and in the meantime there are likely to be issues they have that a competent human driver would not have

(just the same as how they might do some things better than competent drivers might)

doppelgreener

Well, we can do both.

Release decent cars now. Ones that are at least as good as a decent human driver. Use the data we gather based on what they do to improve them. Upgrade the existing cars as we learn how to make better autonomous cars -- Tesla will almost certainly be doing this with its self-driving cars -- and newly developed self-driving cars come with those upgrades automatically.

Waiting just a few years so that all the problems gets solved isn't so viable, because we don't know what a lot of the problems are before these things are on the roads in large quantities.

You can only do and learn so much in theory and in a lab.

trogdor

06:40

sorry I was afk for a bit

@doppelgreener this is entirely fair, and in fact I was going to bring it up myself

trogdor

06:59

my only problem with that is that it is a catch 22, either we never get better information on how well these cars will do or we in effect use people driving on the road as test subjects when they didn't necessarily agree to it

neither of those things is particularly great from my point of view

but it also doesn't mean I am saying we should just scrap the whole idea

(even if we did somehow have the power to do that XD)

doppelgreener

@trogdor consider it like the development of hatchets. We didn't start out with well-wrought iron axes with finely crafted wooden handles and just the right kind of join. We had stone, then bronze, then began learning what didn't go do well, etc

You don't ever get the perfect product on release in a new field. There are always improvements to be made.

trogdor

I will leave this discussion off with the idea that I understand it needs to go through a trail of fire, I just don't LIKE that it does

doppelgreener

Gotta break a few eggs and all that.

trogdor

yeah those eggs are people though :(

BESW

doppelgreener

08:43

11

A: Can fighter use Commander's strike when he only have one attack?

Yes You appear to be confusing the Attack action with an attack. The Attack action is one of the things you can choose to do with your action; others include the Dash action, the Cast a Spell action and the Use an Object action. An attack is anything you do that causes you to roll a die to ove...

> You appear to be confusing the Attack action with an attack.

The Attack action is one of the things you can choose to do with your action; others include the Dash action, the Cast a Spell action and the Use an Object action.

It's almost like they should've just called it something more obvious and descriptive like a standard action, but were avoiding doing that for some reason...

Aether

09:49

@BESW It sadly doesn't help you right now, but GOG actually has some really good sales, too... even rivaling "Steam-cheap". The summer one just ended, unfortunately... keep an eye out for them though.

BESW

@Aether I play video games so rarely, I usually miss such sales entirely.

I only notice Steam summer sales because the Internet makes it hard to miss 'em.

...and I really want to play The Witness but I have no machine that can run it.

So I figure no point in buying it while it's still expensive--by the time I've got a computer that can run it, it'll probably be cheaper off sale than it is on sale now.

Aether

Fair enough.

BESW

(My professional computer is Mac, and Witness doesn't do Mac.)

Aether

I don't know if you can get GOG to email you when wishlist items go on sale, but if so, it's worth looking at...

(It's the kind of thing they really should provide)

You can always do both: wait for it to get cheap AND on sale...

(Which is what I generally do)

BESW

10:07

Absolutely.

But the instant I get a computer that can run it, I'll probably be too impatient to wait for a sale too.

...if Spacechem had a Mac version I'd be all over it right now.

Aether

10:39

Ah, sad that there's no Mac version. It's a good game, and I see is super cheap right now.

BESW

I played the demo, and it hit a very specific sweet spot in my brain.

Aether

I don't know if you do flash at all, but if so, you might want to take a look at The Codex of Alchemical Engineering, which is done by the same developer along somewhat similar lines.

BESW

I'll make a note of that.

Thanks.

BESW

23:37

Not a day goes by that I don't think about this vine. https://vine.co/v/hDjE3QnixKr

Tweeted by NowWeAreAllTom on June 24, 2016 at 10:43 PM

Transcript for

Not a bar, but plays one on TV