417

LESSWRONG
LW

416
Machine Intelligence Research Institute (MIRI)
Frontpage

51

Singularity Institute is now Machine Intelligence Research Institute

by Kaj_Sotala
31st Jan 2013
1 min read
96

51

Machine Intelligence Research Institute (MIRI)
Frontpage

51

Singularity Institute is now Machine Intelligence Research Institute
86Xachariah
10lukeprog
8Kawoomba
10wuncidunci
4[anonymous]
2Xachariah
0wuncidunci
8Xachariah
0FiftyTwo
0gwern
6lukeprog
-1MugaSofer
56Mitchell_Porter
23Larks
8cody-bryce
0Ben Pace
22Exiles
19Rain
7lukeprog
7[anonymous]
5lukeprog
19Ben Pace
1private_messaging
12wedrifid
9novalis
9timtyler
7buybuydandavis
0Eugine_Nier
6A1987dM
12Kaj_Sotala
3amacfie
0loup-vaillant
2A1987dM
2FiftyTwo
2David_Gerard
6RomeoStevens
8Baughn
5Tuxedage
22Qiaochu_Yuan
10Alexei
2JoshuaFox
0Document
0curiousepic
0bentarm
2Kaj_Sotala
0JoshuaFox
7John_Maxwell
1MichaelAnissimov
5[anonymous]
1MichaelAnissimov
2JoshuaFox
-1The Dao of Bayes
2John_Maxwell
-4The Dao of Bayes
2John_Maxwell
-2CronoDAS
-3[anonymous]
-6Nornagest
0fubarobfusco
1Nornagest
2fubarobfusco
4BlazeOrangeDeer
0Andreas_Giger
8gwern
22Kaj_Sotala
17Paul Crowley
1gwern
37Alex_Altair
4Kevin
22lukeprog
17Alicorn
6Paul Crowley
16David_Gerard
15Rain
0pleeppleep
11Kaj_Sotala
1Kevin
-3Shmi
5hankx7787
1Shmi
2Baughn
12Shmi
4drethelin
1Shmi
2Baughn
0drethelin
-1timtyler
0loup-vaillant
2gjm
-2Shmi
0gjm
0Shmi
-1David_Gerard
2fubarobfusco
0Shmi
-18Thomas
New Comment
96 comments, sorted by
top scoring
Click to highlight new comments since: Today at 10:20 PM
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings
[-]Xachariah13y860

The new acronym for the Singularity Institute is MIRI.

The first google hit is the wikipedia page for the Star Trek: TOS episode Miri (S1E8). It's about how 90% of the population of not-Earth was destroyed by an existential threat leaving nothing but irrational children. The crew find themselves under a death sentence from this threat and try to find a solution, but they need the children's help. However the children think themselves immune and immortal and won't assist. In the last seconds, the crew manages to convince the children that the existential threat cannot be ignored and must be solved or the kids will eventually die too. With their help, the crew saves the day and everyone lives happily ever after. Also, the episode was so ahead of it's time that even though it was reviewed as excellent, it got so many complaints that it was never rebroadcast for 20 years.

I think my symbolism detector just pegged off the charts then exploded.

Reply
[-]lukeprog13y100

For the record, we totally didn't know this before you posted this comment. I remember seeing the Google search result but didn't read the plot summary on Wikipedia.

Reply
8Kawoomba13y
You should still replace his broken symbolism detector, if only symbolically.
[-]wuncidunci13y100

When I searched the first hit was the Malaysian town called Miri. Looks like an example of filter bubbles.

Reply
4[anonymous]13y
ddg agrees on "miri" being the Malaysian town.
2Xachariah13y
What do you get when you use incognito mode? I checked with that and got the same Star Trek result. I think incognito pops Google's filter bubble, although I'm not certain. Though other search engines do give me the Malaysian town.
0wuncidunci13y
The same, though Star Trek comes up second. Though google uses a lot of other info about your computer to determine the results (like IP-adress and browser details).
8Xachariah13y
I had overlooked that I was trying to hide from Google using their own software. Of course they wouldn't have provided it if it worked against them. Silly of me in retrospect. A handful of proxies verify that it is indeed Malaysia first and Star Trek second.
0FiftyTwo13y
Same, followed by "Mid-Infrared Instrument "
0gwern13y
Funny, I get Star Trek, Malaysia, language, Star Trek, musical instructor, and only hit #6 is yours.
6lukeprog13y
Near the end of the episode, Kirk is trying to persuade the irrational children to help him save the planet, and the children just keep yelling "Blah Blah Blah!" Kirk says:
-1MugaSofer13y
I knew I'd heard that name somewhere ... suddenly feel a lot more respect for whoever picked it.
[-]Mitchell_Porter13y560

MIRI's number-one goal will be the discovery of a Consequentialist Anticipatory Logic that can save the world (codename, MIRI-CAL).

Reply
[-]Larks13y230

I think this is a good change. Bravo!

It will be hard not to refer to "Singinst" anymore as it did trip off the tougne, but I guess "The Singularity Institute for or against Artificial Intelligence, depending on which seems like a better idea at the time" was never really on the cards.

Reply
8cody-bryce13y
There is an important tradition of people using the former names of things for decades.
0Ben Pace13y
We could call it the Machriarchy. (MACHine intelligence ReaseARCH Institute, and just stick in the 'I')
[-]Exiles13y220

We wish, Dmytry.

Reply
[-]Rain13y190

I like it.

Does this mean I have to put both on my tax forms for donations this year?

Reply
7lukeprog13y
We'll be sure to inform all 2013 donors of the proper procedure.
7[anonymous]13y
And what about paypal monthlys? Will they redirect properly or do I have to exert agency?
5lukeprog13y
We'll be in touch with everyone about that when the time comes.
[-]Ben Pace13y190

This is an interesting example of how the changing definitions doesn't change reality. 'Singularity Institute for Artificial Intelligence' sounded all future-y to me, and 'Machine Intelligence Research Institute' sounds all technical and intelligent (silly stereotypes), but in reality they've not changed. Still the same people, doing the same things.

Gonna have to change the site name, though...

Reply
1private_messaging13y
The Machine Intelligence Research Institute indeed sounds like some older guys, maybe cranky but well accomplished in the field, working on something... That is until you find star trek reference.
[-]wedrifid13y120

And "Machine Intelligence Research Institute" is appropriately descriptive while still being general enough.

MIRI. Machine Intelligence Research Institute.

Adequate. Not especially inspiring but I can't think of anything better either. It is certainly better than "Singularity Institute". (Good change!)

Reply
[-]novalis13y90

Don't forget to redo your credit card design, when you get around to it.

Reply
[-]timtyler13y90

"Machine Intelligence" is my preferred term. "Singularity" seems like pseudoscientific terminology to me.

Reply
[-]buybuydandavis13y70

it's better to have a name that doesn't give a new person all kinds of weird initial associations as their first impression.

Yeah, though I buy into much of what Kurzweil has to say, Singularity has always rubbed me the wrong way - too much of Omega Immanentizing the Eschaton for me.

Is there any other snappy term out there for the idea that we're headed for very big changes through exponential performance improvement in lots of technologies?

Reply
0Eugine_Nier13y
Changing the name doesn't change the fact that they're trying to Immanentize the Eschaton.
[-]A1987dM13y60

Why "Machine" rather than "Artificial"?

Reply
[-]Kaj_Sotala13y120

To me, "Machine Intelligence" sounds less worn than "Artificial Intelligence", and also seems to more strongly imply that they're talking about general intelligence rather than narrow AI. But I don't know whether those were the actual reasons.

Reply
3amacfie13y
I was under the impression that this use of the word "machine" was archaic -- it was used decades ago for naming things like machine learning, machine translation, and the Association for Computing Machinery. I don't immediately see why a more familiar term wasn't used.
0loup-vaillant13y
Possibly for the "M". Imagine "AIRI" instead of "MIRI".
2A1987dM13y
It does, but why “worn” is a bad thing in this context? Wouldn't you want a familiar-sounding phrase? I get the reverse impression, probably because “artificial intelligence” reminds me of science fiction, whereas “machine intelligence” reminds me of Google Translate and self-driving cars.
2FiftyTwo13y
Agreed. Also is narrower, you could plausibly argue that lots of things were 'artificial intelligence' (e.g. bioengineered neural goop) but machine is closer to what we're actually talking about.
2David_Gerard13y
The first hit on "AIRI" isn't as good.
[-]RomeoStevens13y60

Is it an acronym or initialism?

Reply
8Baughn13y
I'm vaguely suspecting a retronym, based on that Star Trek episode.
[-]Tuxedage13y50

I personally dislike the change, but I trust that you guys have changed your names for a reason. I think I may be reacting negatively due to the nostalgia factor.

Reply
[-]Qiaochu_Yuan13y220

Reversal test: if it were always called MIRI and just now decided to change its name to SIAI, how would you feel about it? (This isn't quite the right test because maintaining the same name over time does have some value, but it might help.)

Reply
[-]Alexei13y100

This is the main reason.

Reply
[-]JoshuaFox13y20

Good.

The name "Singularity Institute for Artificial Intelligence" seemed clunky to me from the moment I encountered it.

Too long. Two parts that don't work together, a first part with social acceptability issues, and a second part which has been taboo in leading tech circles since 1992.

It took me about half a year before I stopped typing siai.org when looking for the Institute's site.

Not to mention that SU took over the mindspace, and that SI occasionally strayed from the area implied by its name.

So, good job! I like it.

Reply
[-]Document11y00

I wonder if anyone suggested the Council for Understanding Logic and Technology.

Reply
[-]curiousepic13y00

I just realized that "MIRI" is (perhaps intentionally) evocative of the word "mirror", which is all kinds of suitable.

Reply
[-]bentarm13y00

I notice that http://www.miri.org is very definitely not a placeholder for a new Singularity Institute page. Have you managed to acquire it?

(miri.com seems as though it should be available, but not exactly entirely appropriate. Maybe better than nothing).

Reply
2Kaj_Sotala13y
Nope.
[-]JoshuaFox13y00

Now, all we need is a replacement for "Singularitarian" ... this, time, one that people can get right when they try to repeat it.

Reply
7John_Maxwell13y
What happened to futurist, transhumanist, extropian?
1MichaelAnissimov13y
Transhumanist
5[anonymous]13y
Already exists, and not all transhumanists are singularitarians.
1MichaelAnissimov13y
Transhumanist
2JoshuaFox13y
Singularitarians are, IMHO, a subtype of Tranhumanists who are (1) focused on the Intelligence Explosion rather than nanotech, life extension, cryopreservation, Cyborg art, or any other Transhumanist area, and (2) do something about it. Of course, people canbe Singularitarians and also other things as well, and people can believe something is a good idea without doing anything about it, but the above definition seems significant enough to be worth having a name for. You could also potentially be a Singularitarian but not a Transhumanist, if you focused on AGI safety and even Friendliness, without caring about the entire H+ memeplex. Though there may be nobody like this today, I can imagine a future in which a game theorist or other scientist enters the field out of a combined desire to do good and to boost their career, but is not interested in all the other stuff.
-1The Dao of Bayes13y
Mirian seems easy to say :)
2John_Maxwell13y
That's a terrible idea if they want to maintain credibility as a serious independent bunch of academic types.
-4The Dao of Bayes13y
I had assumed the question was more aimed towards LessWrong, not serious academic usage. I'd expect academic references to be in the form "I work for MIRI", not cutesy shorthand.
2John_Maxwell13y
Still, no one walks around calling themself a CSAILian. If there are people calling themselves Mirians, even if those people aren't directly affiliated with MIRI, that could hurt MIRI's credibility, I suspect, because it would make them seem unusual relative to other serious research groups.
-2CronoDAS13y
"Phyg" member! ;)
-3[anonymous]13y
"phygger"
-6Nornagest13y
[-]Andreas_Giger13y00

Looks like an attempt to get rid of the negative image associated with the name Singularity Institute. I wonder if it isn't already too late to take PR seriously.

Reply
8gwern13y
From OP: I'm not sure this name change is a good idea or worth whatever SU offered (or that there was a real brand issue), but there apparently was some other motivation than 'SI now has embarrassing connotations'.
[-]Kaj_Sotala13y220

SU and SI kept getting confused with each other all the time.

Reply
[-]Paul Crowley13y170

Also, when you're trying to explain that there's a gigantic difference between you and another organisation with a similar-sounding name, you can sound a little like the People's Front of Judea.

Reply
1gwern13y
By whom? I think the only time I personally saw that happen online was in one British newspaper article.
[-]Alex_Altair13y370

My impression is that anyone who has ever heard of Singularity University doesn't even have it in their hypothesis space that you mean something different when you say Singularity Institute.

Reply
4Kevin13y
Yup. Even when they do have it in their hypothesis space, it still gets mangled. I recently got a follow-up email from someone that still thought I was Singularity University. I had briefly explained to him about how SU had acquired the Singularity Summit from us, and his follow-up email said "now that you have acquired the Singularity Summit, you may be interested in my product..."
[-]lukeprog13y220

By whom?

Almost every time I spoke to anyone who wasn't deeply familiar with either SU or SI. Including almost every press person.

Reply
[-]Alicorn13y170

Rudi Hoffman confused them when I sought a quote from him and mentioned Singinst. And him you'd expect to move in the right circles to know the difference.

Reply
6Paul Crowley13y
When I stayed with a friend in the Bay Area, I was confused that he said he knew loads of people in SingInst, but kept naming people I'd never heard of - and guess why!
[-]David_Gerard13y160

There is a real brand issue. I say "Singularity Institute" to people down the pub, the ones who've heard the word go "ah, Kurzweil!" (I was trying to explain this site I like called LessWrong.)

Reply
[-]Rain13y150

I told someone at work, and they said, "Oh, like on that Fringe episode [about Kurzweillian uploading]."

Reply
[-]pleeppleep13y00

Kinda awkward to say aloud. I think Institute for the Research of Machine Intelligence would sound better. Minor nitpick.

Reply
[-]Kaj_Sotala13y110

Really? To me, the extra words in "Institute for the Research of Machine Intelligence" feel redundant, and MIRI is better for being concise and to the point.

Reply
1Kevin13y
IRMI? irm-y? Sounds like squirm. Or the name Erma.
[-]Shmi13y-30

What MIRI really is is the Institute for study of emergent intelligences, whether machine, biological, hybrid or any other kind, but given how Eliezer dislikes the term emergence, I can see why EIRI would be a non-starter. Still, I like the new name better than the old one.

Reply
5hankx778713y
"What exactly do you mean by ‘machine’, such that humans are not machines?" - Eliezer Yudkowsky
1Shmi13y
Good point. The Wikipedia description certainly covers humans.
2Baughn13y
There's a reason he doesn't like it.. I'm not entirely sure what your sentence means. Could you rewrite it to not use "emergence" (or define "emergence")?
[-]Shmi13y120

The reason he does not like the term is that, as pointed out before, "emergence" is not an explanation of anything. However, it is an observational phenomenon: when you get a lot of simple things together, they combine in ways one could not foresee and the resulting entities behave by the rules not constructable from (but reducible to) those of the simple constituents. When you combine a lot of simple molecules, you get a solid, a liquid or a gas with the properties you generally cannot infer without observing them first. When you get a group of people together, they start interacting in apriori unpredictable ways as they form a group. Once you observe the group behavior, you can often reduce it to that of its constituents, but a useful description is generally not in terms of the constituents, but in terms of the collective. For example, in thermodynamics people use gas laws and other macroscopic laws instead of the Newton's laws.

I am guessing that one reason that the (friendly) machine intelligence problem is so hard is that intelligence is an emergent property: once you understand it, you can reduce it to interactions between neurons, but you cannot infer it from suc... (read more)

Reply
4drethelin13y
Emergence is a subset of the word Surprise. It's not meaningless but you can't use it to usefully predict things you want to achieve with something, because it's equivalent to saying "If we put all these things together maybe they'll surprise us in an awesome way!"
1Shmi13y
Sort of. It is not surprising that incremental quantitative changes results in a qualitative change, but the exact nature of what emerges can indeed be quite a surprise. It is nevertheless useful to keep in mind the general pattern in order to not be blindsided by the fact of emergence in each particular case ("But... but.. they are all nice people, I didn't expect them to turn into a mindless murderous mob!"). And to be ready to take action when the emergent entity hits the fan.
2Baughn13y
Or in simpler terms, AI is a crapshoot.
0drethelin13y
Agreed. Like with surprises, you can try to be robust to them or agile enough to adapt.
-1timtyler13y
If something is an emergent property, you can bet on it not being the sum of its parts. That has some use.
0loup-vaillant13y
Aiming the tiny Friendly dot in AI-space is not one of them, though.
2gjm13y
Surely what MIRI would ideally like to do is to find a way of making intelligence not "emergent", so that it's easier to make something intelligent that behaves predictably enough to be classified as Friendly.
-2Shmi13y
I don't believe that MIRI has been consciously paying attention to thwarting undesirable emergence, given that EY refuses to acknowledge it as a real phenomenon.
0gjm13y
I fear we're at cross purposes. I meant not "thwart emergent intelligence" but "find ways of making intelligence that don't rely on it emerging mysteriously from incomprehensible complications".
0Shmi13y
Sure, you cannot rely on spontaneous emergence for anything predictable, as neural network attempts at AGI demonstrate. My point was that if you ignore the chance of something emerging, that something will emerge in a most inopportune moment. I see your original point, though. Not sure if it can be successful. My guess is that the best case is some kind of "controlled emergence", where you at least set the parameter space of what might happen.
-1David_Gerard13y
First hit on EIRI. More appositely, EIRI exists.
2fubarobfusco13y
... and here I was thinking of Masami Eiri from Serial Experiments Lain.
0Shmi13y
Surely this is not your real objection. One can try EII or IEI or...
[+]Thomas13y-180
Moderation Log
More from Kaj_Sotala
View more
Curated and popular this week
96Comments

http://singularity.org/blog/2013/01/30/we-are-now-the-machine-intelligence-research-institute-miri/

As Risto Saarelma pointed out on IRC, "Volcano Lair Doom Institute" would have been cooler, but this is pretty good too. As the word "Singularity" has pretty much lost its meaning, it's better to have a name that doesn't give a new person all kinds of weird initial associations as their first impression. And "Machine Intelligence Research Institute" is appropriately descriptive while still being general enough.