[Wikimedia-l] DARPA FOSSS programs of interest

classic Classic list List threaded Threaded
10 messages Options
Reply | Threaded
Open this post in threaded view
|

[Wikimedia-l] DARPA FOSSS programs of interest

James Salsman-2
Re https://lists.wikimedia.org/pipermail/publicpolicy/2016-April/001335.html

Should the Foundation mount a campaign to rescue BOLT from whomever
took it down from the DARPA site?

"The Broad Operational Language Translation (BOLT) program is aimed at
enabling communication with non-English-speaking populations and
identifying important information in foreign-language sources by: 1)
allowing English-speakers to understand foreign-language sources of
all genres, including chat, messaging and informal conversation; 2)
providing English-speakers the ability to quickly identify targeted
information in foreign-language sources using natural-language
queries; and 3) enabling multi-turn communication in text and speech
with non-English speakers. If successful, BOLT would deliver all
capabilities free from domain or genre limitations."

Program Manager: Dr. Boyan Onyshkevych

Contact: [hidden email]

Here is another one that we need to rescue:

"The general goal of the Social Media in Strategic Communication
(SMISC) program is to develop a new science of social networks built
on an emerging technology base. Through the program, DARPA seeks to
develop tools to support the efforts of human operators to counter
misinformation or deception campaigns with truthful information."

Program Manager: Dr. Rand Waltzman

Contact: [hidden email]

Are there any reasons that trying to do this might be a bad idea?

_______________________________________________
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
New messages to: [hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:[hidden email]?subject=unsubscribe>
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] DARPA FOSSS programs of interest

Oliver Keyes-5
Off the top of my head:

1. Because we have absolutely no idea, from the briefs given, about
the technological stack and how well it meshes with our existing
expertise as a movement, or the WMF's existing expertise as a
technical entity;
2. Because we have absolutely no idea, from the briefs given, how much
progress was made and thus how much work is needed or even if the
projects were a failure;
3. Because the policy you're linking to is a policy for 18F, and it's
totally unclear as to whether DARPA contractors or employees are
covered by that, or whether the source code would be released or fall
under the restrictions section given its origin in a military research
arena;
3. Because "hey, we took a load of propaganda software from the US
military and now we're spending donor dollars on it" looks
*incredibly* bad and ominous to pretty much anyone outside the US,
including most of Africa, the Middle East, and Asia, areas we are
trying to expand into as their populations come online more and more
and areas the US military that wrote this software is continuously
bombing the shit out of.

Basically it's far too early to make any useful determination about
this, and even if it wasn't, the optics are atrocious.

On Tue, Apr 12, 2016 at 1:54 PM, James Salsman <[hidden email]> wrote:

> Re https://lists.wikimedia.org/pipermail/publicpolicy/2016-April/001335.html
>
> Should the Foundation mount a campaign to rescue BOLT from whomever
> took it down from the DARPA site?
>
> "The Broad Operational Language Translation (BOLT) program is aimed at
> enabling communication with non-English-speaking populations and
> identifying important information in foreign-language sources by: 1)
> allowing English-speakers to understand foreign-language sources of
> all genres, including chat, messaging and informal conversation; 2)
> providing English-speakers the ability to quickly identify targeted
> information in foreign-language sources using natural-language
> queries; and 3) enabling multi-turn communication in text and speech
> with non-English speakers. If successful, BOLT would deliver all
> capabilities free from domain or genre limitations."
>
> Program Manager: Dr. Boyan Onyshkevych
>
> Contact: [hidden email]
>
> Here is another one that we need to rescue:
>
> "The general goal of the Social Media in Strategic Communication
> (SMISC) program is to develop a new science of social networks built
> on an emerging technology base. Through the program, DARPA seeks to
> develop tools to support the efforts of human operators to counter
> misinformation or deception campaigns with truthful information."
>
> Program Manager: Dr. Rand Waltzman
>
> Contact: [hidden email]
>
> Are there any reasons that trying to do this might be a bad idea?
>
> _______________________________________________
> Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
> New messages to: [hidden email]
> Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:[hidden email]?subject=unsubscribe>

_______________________________________________
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
New messages to: [hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:[hidden email]?subject=unsubscribe>
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] DARPA FOSSS programs of interest

Nathan Awrich
In reply to this post by James Salsman-2
On Tue, Apr 12, 2016 at 1:54 PM, James Salsman <[hidden email]> wrote:

> Re
> https://lists.wikimedia.org/pipermail/publicpolicy/2016-April/001335.html
>
> <snip>
>
> Are there any reasons that trying to do this might be a bad idea?
>
> ______


Because the WMF is not, at least as far as we know, in the business of
covert HUMINT or surveillance technology.
_______________________________________________
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
New messages to: [hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:[hidden email]?subject=unsubscribe>
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] DARPA FOSSS programs of interest

Andy Mabbett-2
In reply to this post by Oliver Keyes-5
> DARPA

Yeah, what have DARPA ever done for us..?

--
Andy Mabbett
@pigsonthewing
http://pigsonthewing.org.uk
_______________________________________________
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
New messages to: [hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:[hidden email]?subject=unsubscribe>
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] DARPA FOSSS programs of interest

Marc-Andre
On 16-04-12 06:25 PM, Andy Mabbett wrote:
> Yeah, what have DARPA ever done for us..?

For the benefit of those who do not sport beards[1], one of the most
relevant things that is a legacy of DARPA - and certainly the one Andy
is alluding to - is that of Internet itself (née Arpanet).

-- Coren / Marc

[1] Which beards may be virtual, as befits those of us dinosaurs who
couldn't grow one if their life depended on it[2].

[2] Or whose marriage depends on not having one.  :-)


_______________________________________________
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
New messages to: [hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:[hidden email]?subject=unsubscribe>
Reply | Threaded
Open this post in threaded view
|

[Wikimedia-l] DARPA FOSSS programs of interest

James Salsman-2
In reply to this post by James Salsman-2
> we have absolutely no idea ... about the technological
> stack [or] how much progress was made....

Can anyone think of another way to find out?

> covert HUMINT or surveillance technology

If we publish the code, it's not covert anymore. We all deserve to see the
mentions of Wikipedia which occurred in the SMISC program and
project archives, if we want to protect our readers from whichever
intelligence agencies have hacked Foundation servers.

I selected BOLT and SMISC
from
https://web.archive.org/web/20150529033655/http://www.darpa.mil/opencatalog/index.html

because they appeared compatible with Asimov's three laws of robotics, and
did not appear to be harmful. There is one project in there, CSFV which
could be actively harming the Foundation's ability to attract and retain
volunteer editors:

"Crowd Sourced Formal Verification (CSFV) is a DARPA program that aims to
investigate whether large numbers of non-experts can perform formal
verification faster and more cost-effectively than conventional processes.
The goal is to transform verification into a more accessible task by
creating fun, intuitive games that reflect formal verification problems.
Playing the games would effectively help software verification tools
complete corresponding formal verification proofs."

Doesn't that mean that the Foundation has the legal standing to see all
three of those projects published?
_______________________________________________
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
New messages to: [hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:[hidden email]?subject=unsubscribe>
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] DARPA FOSSS programs of interest

Nathan Awrich
On Wed, Apr 13, 2016 at 9:53 AM, James Salsman <[hidden email]> wrote:
>
>
> Doesn't that mean that the Foundation has the legal standing to see all
> three of those projects published?


Where do you see legal standing being a factor here, and how would the WMF
have it?
_______________________________________________
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
New messages to: [hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:[hidden email]?subject=unsubscribe>
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] DARPA FOSSS programs of interest

James Salsman-2
In reply to this post by James Salsman-2
> Where do you see legal standing being a factor...?

On further reflection, it would certainly be better to simply ask the
DARPA Crowdsourced Formal Verification (CSVF) Program Manager Daniel
Ragsdale, who has left DARPA and is now a Professor at Texas A&M
University, about the extent to which enhancing games with logic
puzzles produced crowdsourced assistance. I wonder whether they were
using first person shooters, flying birds, modified tic-tac-toe, or
what.

Similarly for Rand Waltzman, who is now an Associate Director of
Research at Carnegie Mellon's Software Engineering Institute, about
the more interesting Social Media in Strategic Communication (SMISC)
program and the extent to which it and related programs in other
agencies have already involved Wikipedia. He might want to talk about
that because the program was not intended to be covert. As an open
program, it's very similar to multiple proposals from the community
we've seen recently. As a covert program, it's likely discoverable and
certainly referenceable in the lawsuit against the NSA as a means to
measure the extent to which such programs across the government have
resulted in law enforcement prosecutions from parallel construction
affecting people because of the Wikipedia articles readers have chosen
to access.

The director of the Broad Operational Language Translation (BOLT)
program, which was originally intended to be an open source and far
more fully-featured alternative to Google Translate, is still at DARPA
and the program is ongoing with software development partnerships at
ten universities and three private companies:
 https://slator.com/technology/darpa-doles-out-millions-to-academia-and-vendors-to-translate-any-language-by-2019/

There is absolutely no question that the Foundation would directly
benefit tremendously if the BOLT program were returned to unclassified
free open source. Clearly that would not be in Google's interest at
all.

As far as I can tell from
https://www.fbo.gov/utils/view?id=047042419d79fc12b8f6a12e41af570c
the only reason the BOLT program requires Top Secret clearance is
because the identities of wartime human translators are secret. I am
no expert on classification and declassification, other to have
noticed that even classification advocates say that there is far too
much of it.

Therefore I think it would be worth writing a letter asking that the
BOLT, SMISC, and CSFV be returned to open source to the extent
possible. This is the sort of thing that I imagine would take a few
hours at most by the people working on the NSA lawsuit asking for a
Mandatory Declassification Review per
http://www.dtic.mil/whs/directives/corres/pdf/523030m.pdf

If I am mistaken or if anyone thinks it is not a good idea to ask for
this, please let me know.

On Wed, Apr 13, 2016 at 7:53 AM, James Salsman <[hidden email]> wrote:

>
>> we have absolutely no idea ... about the technological
>> stack [or] how much progress was made....
>
> Can anyone think of another way to find out?
>
>> covert HUMINT or surveillance technology
>
> If we publish the code, it's not covert anymore. We all deserve to see the
> mentions of Wikipedia which occurred in the SMISC program and project
> archives, if we want to protect our readers from whichever intelligence
> agencies have hacked Foundation servers.
>
> I selected BOLT and SMISC from
> https://web.archive.org/web/20150529033655/http://www.darpa.mil/opencatalog/index.html
> because they appeared compatible with Asimov's three laws of robotics, and
> did not appear to be harmful. There is one project in there, CSFV which
> could be actively harming the Foundation's ability to attract and retain
> volunteer editors:
>
> "Crowd Sourced Formal Verification (CSFV) is a DARPA program that aims to
> investigate whether large numbers of non-experts can perform formal
> verification faster and more cost-effectively than conventional processes.
> The goal is to transform verification into a more accessible task by
> creating fun, intuitive games that reflect formal verification problems.
> Playing the games would effectively help software verification tools
> complete corresponding formal verification proofs."
>
> Doesn't that mean that the Foundation has the legal standing to see all
> three of those projects published?

the other two being:

"The Broad Operational Language Translation (BOLT) program is aimed at
enabling communication with non-English-speaking populations and
identifying important information in foreign-language sources by: 1)
allowing English-speakers to understand foreign-language sources of
all genres, including chat, messaging and informal conversation; 2)
providing English-speakers the ability to quickly identify targeted
information in foreign-language sources using natural-language
queries; and 3) enabling multi-turn communication in text and speech
with non-English speakers. If successful, BOLT would deliver all
capabilities free from domain or genre limitations."

and

"The general goal of the Social Media in Strategic Communication
(SMISC) program is to develop a new science of social networks built
on an emerging technology base. Through the program, DARPA seeks to
develop tools to support the efforts of human operators to counter
misinformation or deception campaigns with truthful information."

_______________________________________________
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
New messages to: [hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:[hidden email]?subject=unsubscribe>
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] DARPA FOSSS programs of interest

Oliver Keyes-5
That didn't really identify any of the questions. You're suggesting
that counsel spend their time writing to agencies to ask about the
copyright status of programs with the intent of considering taking
them over, when we know almost nothing about them.

Since you've identified the people responsible and how to contact
them, if you care about this so much, you do it - and then come back
when you have information as basic as their copyright status or what
language they're even written in or whether they were even a success.
Legal should not be going on snipe hunts.

On Mon, Apr 18, 2016 at 10:10 AM, James Salsman <[hidden email]> wrote:

>> Where do you see legal standing being a factor...?
>
> On further reflection, it would certainly be better to simply ask the
> DARPA Crowdsourced Formal Verification (CSVF) Program Manager Daniel
> Ragsdale, who has left DARPA and is now a Professor at Texas A&M
> University, about the extent to which enhancing games with logic
> puzzles produced crowdsourced assistance. I wonder whether they were
> using first person shooters, flying birds, modified tic-tac-toe, or
> what.
>
> Similarly for Rand Waltzman, who is now an Associate Director of
> Research at Carnegie Mellon's Software Engineering Institute, about
> the more interesting Social Media in Strategic Communication (SMISC)
> program and the extent to which it and related programs in other
> agencies have already involved Wikipedia. He might want to talk about
> that because the program was not intended to be covert. As an open
> program, it's very similar to multiple proposals from the community
> we've seen recently. As a covert program, it's likely discoverable and
> certainly referenceable in the lawsuit against the NSA as a means to
> measure the extent to which such programs across the government have
> resulted in law enforcement prosecutions from parallel construction
> affecting people because of the Wikipedia articles readers have chosen
> to access.
>
> The director of the Broad Operational Language Translation (BOLT)
> program, which was originally intended to be an open source and far
> more fully-featured alternative to Google Translate, is still at DARPA
> and the program is ongoing with software development partnerships at
> ten universities and three private companies:
>  https://slator.com/technology/darpa-doles-out-millions-to-academia-and-vendors-to-translate-any-language-by-2019/
>
> There is absolutely no question that the Foundation would directly
> benefit tremendously if the BOLT program were returned to unclassified
> free open source. Clearly that would not be in Google's interest at
> all.
>
> As far as I can tell from
> https://www.fbo.gov/utils/view?id=047042419d79fc12b8f6a12e41af570c
> the only reason the BOLT program requires Top Secret clearance is
> because the identities of wartime human translators are secret. I am
> no expert on classification and declassification, other to have
> noticed that even classification advocates say that there is far too
> much of it.
>
> Therefore I think it would be worth writing a letter asking that the
> BOLT, SMISC, and CSFV be returned to open source to the extent
> possible. This is the sort of thing that I imagine would take a few
> hours at most by the people working on the NSA lawsuit asking for a
> Mandatory Declassification Review per
> http://www.dtic.mil/whs/directives/corres/pdf/523030m.pdf
>
> If I am mistaken or if anyone thinks it is not a good idea to ask for
> this, please let me know.
>
> On Wed, Apr 13, 2016 at 7:53 AM, James Salsman <[hidden email]> wrote:
>>
>>> we have absolutely no idea ... about the technological
>>> stack [or] how much progress was made....
>>
>> Can anyone think of another way to find out?
>>
>>> covert HUMINT or surveillance technology
>>
>> If we publish the code, it's not covert anymore. We all deserve to see the
>> mentions of Wikipedia which occurred in the SMISC program and project
>> archives, if we want to protect our readers from whichever intelligence
>> agencies have hacked Foundation servers.
>>
>> I selected BOLT and SMISC from
>> https://web.archive.org/web/20150529033655/http://www.darpa.mil/opencatalog/index.html
>> because they appeared compatible with Asimov's three laws of robotics, and
>> did not appear to be harmful. There is one project in there, CSFV which
>> could be actively harming the Foundation's ability to attract and retain
>> volunteer editors:
>>
>> "Crowd Sourced Formal Verification (CSFV) is a DARPA program that aims to
>> investigate whether large numbers of non-experts can perform formal
>> verification faster and more cost-effectively than conventional processes.
>> The goal is to transform verification into a more accessible task by
>> creating fun, intuitive games that reflect formal verification problems.
>> Playing the games would effectively help software verification tools
>> complete corresponding formal verification proofs."
>>
>> Doesn't that mean that the Foundation has the legal standing to see all
>> three of those projects published?
>
> the other two being:
>
> "The Broad Operational Language Translation (BOLT) program is aimed at
> enabling communication with non-English-speaking populations and
> identifying important information in foreign-language sources by: 1)
> allowing English-speakers to understand foreign-language sources of
> all genres, including chat, messaging and informal conversation; 2)
> providing English-speakers the ability to quickly identify targeted
> information in foreign-language sources using natural-language
> queries; and 3) enabling multi-turn communication in text and speech
> with non-English speakers. If successful, BOLT would deliver all
> capabilities free from domain or genre limitations."
>
> and
>
> "The general goal of the Social Media in Strategic Communication
> (SMISC) program is to develop a new science of social networks built
> on an emerging technology base. Through the program, DARPA seeks to
> develop tools to support the efforts of human operators to counter
> misinformation or deception campaigns with truthful information."
>
> _______________________________________________
> Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
> New messages to: [hidden email]
> Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:[hidden email]?subject=unsubscribe>

_______________________________________________
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
New messages to: [hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:[hidden email]?subject=unsubscribe>
Reply | Threaded
Open this post in threaded view
|

Re: [Wikimedia-l] DARPA FOSSS programs of interest

James Salsman-2
In reply to this post by James Salsman-2
> You're suggesting that counsel spend their time writing to
> agencies ... when we know almost nothing about them.

Again, if there is another way to find out, I'd like to learn it.

>> Therefore I think it would be worth writing a letter asking that the
>> BOLT, SMISC, and CSFV be returned to open source to the extent
>> possible. This is the sort of thing that I imagine would take a few
>> hours at most by the people working on the NSA lawsuit asking for a
>> Mandatory Declassification Review per
>> http://www.dtic.mil/whs/directives/corres/pdf/523030m.pdf
>...
> Since you've identified the people responsible and how to
> contact them, if you care about this so much, you do it....

My most recent FOIAR to DARPA was simply ignored, and
when I followed up on it, the same personnel with whom I had
great success corresponding with previously just stopped
communicating, saying they were too busy. Maybe I triggered
a declassification review for something that was legitimately
secret. In any case, I doubt I would have anywhere near
the Foundation's chance of success at this. I am sure that
the Foundation would be taken seriously and that the likely
benefits outweigh the staff time it would take to ask. Even
if the copyright status is unhelpful, finding that out may
encourage Congress to allow taxpayers access to the fruits
of their labor in the future.

_______________________________________________
Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
New messages to: [hidden email]
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:[hidden email]?subject=unsubscribe>