Google Removes Pledge To Not Use AI For Weapons From Website 58
Google has updated its public AI principles page to remove a pledge to not build AI for weapons or surveillance. TechCrunch reports: Asked for comment, the company pointed TechCrunch to a new blog post on "responsible AI." It notes, in part, "we believe that companies, governments, and organizations sharing these values should work together to create AI that protects people, promotes global growth, and supports national security." Google's newly updated AI principles note the company will work to "mitigate unintended or harmful outcomes and avoid unfair bias," as well as align the company with "widely accepted principles of international law and human rights." Further reading: Google Removes 'Don't Be Evil' Clause From Its Code of Conduct
Google plans to use it's AI (Score:1)
At this point nobody would be surprised... (Score:3)
Re: (Score:2)
Oaths, vows, and pledges (Score:5, Insightful)
What good is it for some corporation to make a pledge that can be changed in the future? None whatsoever.
Re: (Score:3)
What good is it for some corporation to make a pledge that can be changed in the future? None whatsoever.
I mean, that is not different from any election pledge, right? Maybe they could call it their "New Year's resolution", just so everybody understands how dependable it will be.
Re: (Score:2)
What good is it for some corporation to make a pledge that can be changed in the future? None whatsoever.
I mean, that is not different from any election pledge, right?
Exactly. For example, Trump didn't say he'd bring down the price of *all* eggs ... :-)
Re: Oaths, vows, and pledges (Score:4, Interesting)
What good are they? You have never heard of a warrant canary, have you?
Re: (Score:2)
Is it actually a warrant canary? I thought most of the anti-DoD folks got pushed out of Google years ago.
Re: (Score:2)
It's not exactly a warrant canary, but if you know what that is, then you should know what this is.
Re: (Score:2)
What good is it for some corporation to make a pledge that can be changed in the future? None whatsoever.
That's overly simplistic. Times change. Circumstances change. The people controlling corporations change. I once had a close friend once tell me, "Don't quote me to me!" She might have been talking about her stated policy on penises in her mouth... but I digress! My own life has more than a few moments that ran contrary to the declared positions of my younger self.
Re: (Score:2)
That's why you don't make pledges.
Re: (Score:2)
That's why you don't make pledges.
That's the spirit! You're already a step ahead of most politicians!
Re: (Score:2)
So what you are saying is that ple
Re: (Score:2)
Only saying that corporate pledges shouldn't be considered eternal. It's unreasonable. The longer a company lasts, the more circumstances diverge.
My requirement is that they don't be two faced about it. Don't maintain the premise while behaving in a way that's contrary to it. If you aren't going to walk the talk, just say so.
No oath or pledge is broken (Score:2)
Re: (Score:2)
A way of signalling to the world that something has happened.
I mean, there is a wikipedia article [wikipedia.org] about it.
Re: (Score:3)
As long as you remember to have an exit strategy, it's a useful indicator of when you need to stop doing business with them.
And? (Score:1)
AI will be used for weapons. The only question is by who. Not if.
A pledge like that is like pledging not to use metal tubes for weapons.
Re: (Score:2)
Already has been. The only question is why do people suddenly care.
Re: And? (Score:2)
Because someone has the nerve to not hide it anymore.
Re: (Score:2)
The question is nto if metal tubes can be used for weapons, it is whether your particular company will manufacture metal tubes for weapons. Some companies don't want to touch defence for investor/funding reasons. If a company is active in defence, this might change its eligibility to certain public funding schemes, or it might deter some investors, or it might attract/deter a different employee demographics. (My workplace has made such a decision to not touch defence.)
Re: (Score:2)
It's also going to be interesting for Google to be supplying weapons to the US military while also operating as a multinational. I think the US has maintains a list of companies that are sanctioned because they alledgedly supply arms to a certain foreign power's military.
Re: (Score:2)
A pledge like that is like pledging not to use metal tubes for weapons.
Make love, not war.
Re: (Score:2)
A pledge like that is like pledging not to use metal tubes for weapons.
Make love, not war.
I don't think metal tubes are particularly good for that application... plastic tubes seem to be the clear choice among both manufacturers and users.
Re: (Score:2)
Right, and that could be a good thing. Monopoly of Violence theory of government says government's job is to be the strongest, and to use force when necessary to prevent force. (That is very broad, including things like demanding taxes. But in this case it's literal.)
Re: (Score:2)
The point of it is that they used this type of pledges to attract top level researchers to work on their AI systems. Many smart people don't necessarily want to use their knowledge to raise the state of the art in autonomous killbots.
It's like hiring a vegan to design some grinder, and they agree if you pledge not to use it to grind animals to death. And then after they are finished designing, you just 'change the pledge' and immediately use it in your animal slaughter facility.
Illegal? Probably not. But it
Re: (Score:2)
The point of it is that they used this type of pledges to attract top level researchers to work on their AI systems. Many smart people don't necessarily want to use their knowledge to raise the state of the art in autonomous killbots.
It's like hiring a vegan to design some grinder, and they agree if you pledge not to use it to grind animals to death. And then after they are finished designing, you just 'change the pledge' and immediately use it in your animal slaughter facility.
Illegal? Probably not. But it's highly unethical behavior and Google should be called out for it. Basically google has announced that their word is worth nothing. So hopefully governments worldwide pay attention and start making sure that nobody ever trusts google again.
Are you kidding? Governments the world over are probably lining up to buy Google killbots as we speak. We're in a world where breaking your word to make an extra dollar is applauded as brilliant strategy. No government that I'm aware of would balk at this move. Hells, some of them may have been prompting them to drop the pledge to begin with.
As durable as the paper it is written on (Score:1)
Motto... (Score:4, Funny)
Re: Motto... (Score:2)
Making money trumps doing no evil.
What good is a pledge if you change it (Score:2)
Re: (Score:2)
"I am altering the deal. Pray I do not alter it any further."
Google Ethical Workers Leave (Score:3)
Google will do anything for money and market position regardless of ethics or morals, capitalism at its finest.
Corporate charters are no longer valid for furthering the Public Good as was their original intent.
Read the past stories below from Google's Own Search Engine:
https://www.google.com/search?... [google.com]
Military Killer AI Arms Race (Score:2)
And perhaps they didn't make this clear. But AI is a military and killer technology arms race since whoever gets to the point where they can manufacture and deploy vast quantities of non-human devices to attack and invade other countries and take control of their land and resources, then that power will likely use that technology very quickly to grab control of strategic places where resources are scarce and important .
Unfortunately, the reality is all new technologies get sponsored by the military for kill
Re: (Score:2)
(...) In 1994 — the same year the Highlands Forum was founded under the stewardship of the Office of the Secretary of Defense, the ONA, and DARPA — two young PhD students at Stanford University, Sergey Brin and Larry Page, made their breakthrough on the first automated web crawling and page ranking application. That application remains the core component of what eventually became Google’s search service. Brin and Page had performed their work with funding from the Digital Library Initiative (DLI), a multi-agency programme of the National Science Foundation (NSF), NASA and DARPA.
But that’s just one side of the story.
Throughout the development of the search engine, Sergey Brin reported regularly and directly to two people who were not Stanford faculty at all: Dr. Bhavani Thuraisingham and Dr. Rick Steinheiser. Both were representatives of a sensitive US intelligence community research programme on information security and data-mining.
Thuraisingham is currently the Louis A. Beecherl distinguished professor and executive director of the Cyber Security Research Institute at the University of Texas, Dallas, and a sought-after expert on data-mining, data management and information security issues. But in the 1990s, she worked for the MITRE Corp., a leading US defense contractor, where she managed the Massive Digital Data Systems initiative, a project sponsored by the NSA, CIA, and the Director of Central Intelligence, to foster innovative research in information technology.
“We funded Stanford University through the computer scientist Jeffrey Ullman, who had several promising graduate students working on many exciting areas,” Prof. Thuraisingham told me. “One of them was Sergey Brin, the founder of Google. The intelligence community’s MDDS program essentially provided Brin seed-funding, which was supplemented by many other sources, including the private sector.”
This sort of funding is certainly not unusual, and Sergey Brin’s being able to receive it by being a graduate student at Stanford appears to have been incidental. The Pentagon was all over computer science research at this time. But it illustrates how deeply entrenched the culture of Silicon Valley is in the values of the US intelligence community.
In an extraordinary document hosted by the website of the University of Texas, Thuraisingham recounts that from 1993 to 1999, “the Intelligence Community [IC] started a program called Massive Digital Data Systems (MDDS) that I was managing for the Intelligence Community when I was at the MITRE Corporation.” The program funded 15 research efforts at various universities, including Stanford. Its goal was developing “data management technologies to manage several terabytes to petabytes of data,” including for “query processing, transaction management, metadata management, storage management, and data integration.”
At the time, Thuraisingham was chief scientist for data and information management at MITRE, where she led team research and development efforts for the NSA, CIA, US Air Force Research Laboratory, as well as the US Navy’s Space and Naval Warfare Systems Command (SPAWAR) and Communications and Electronic Command (CECOM). She went on to teach courses for US government officials and defense contractors on data-mining in counter-terrorism.
In her University of Texas article, she attaches the copy of an abstract of the US intelligence community’s MDDS program that had been presented to the “Annual Intelligence Community Symposium” in 1995. The abstract reveals that the primary sponsors of the MDDS programme were three agencies: the NSA, the CIA’s Office of Research & Development, and the intelligence community’s Community Management Staff (CMS) which operates under the Director of Central Intelligence. Administrators of the program, which provided funding of around 3–4 million dollars per year for 3–4 years, were identified as Hal Curran (NSA), Robert Kluttz (CMS), Dr. Claudia Pierce (NSA), Dr. Rick Steinheiser (ORD — standing for the CIA’s Office of Research and Devepment), and Dr. Thuraisingham herself.
Thuraisingham goes on in her article to reiterate that this joint CIA-NSA program partly funded Sergey Brin to develop the core of Google, through a grant to Stanford managed by Brin’s supervisor Prof. Jeffrey D. Ullman:
“In fact, the Google founder Mr. Sergey Brin was partly funded by this program while he was a PhD student at Stanford. He together with his advisor Prof. Jeffrey Ullman and my colleague at MITRE, Dr. Chris Clifton [Mitre’s chief scientist in IT], developed the Query Flocks System which produced solutions for mining large amounts of data stored in databases. I remember visiting Stanford with Dr. Rick Steinheiser from the Intelligence Community and Mr. Brin would rush in on roller blades, give his presentation and rush out. In fact the last time we met in September 1998, Mr. Brin demonstrated to us his search engine which became Google soon after.”
Brin and Page officially incorporated Google as a company in September 1998, the very month they last reported to Thuraisingham and Steinheiser. ‘Query Flocks’ was also part of Google’s patented ‘PageRank’ search system, which Brin developed at Stanford under the CIA-NSA-MDDS programme, as well as with funding from the NSF, IBM and Hitachi. That year, MITRE’s Dr. Chris Clifton, who worked under Thuraisingham to develop the ‘Query Flocks’ system, co-authored a paper with Brin’s superviser, Prof. Ullman, and the CIA’s Rick Steinheiser. Titled ‘Knowledge Discovery in Text,’ the paper was presented at an academic conference.
“The MDDS funding that supported Brin was significant as far as seed-funding goes, but it was probably outweighed by the other funding streams,” said Thuraisingham. “The duration of Brin’s funding was around two years or so. In that period, I and my colleagues from the MDDS would visit Stanford to see Brin and monitor his progress every three months or so. We didn’t supervise exactly, but we did want to check progress, point out potential problems and suggest ideas. In those briefings, Brin did present to us on the query flocks research, and also demonstrated to us versions of the Google search engine.”
Brin thus reported to Thuraisingham and Steinheiser regularly about his work developing Google. (...)
the rest of the article: https://medium.com/insurge-int... [medium.com]
Re: (Score:2)
They're all cheering for it because they think it'll never be directed at them. Of course, by the time they're proven wrong, it'll be too late.
Google Captcha = Click in all TANKs! (Score:5, Interesting)
I was wondering if the Google captchas are going to start making the civilian and our militarian ex-military personnel start clicking to identify enemy assets like tanks, airplanes, drowns and infantry formations? When are we going to start seeing combat pictures that we have to identify enemy assets to ensure that the AI is learning correctly so that we can teach it friend from Pho military versus civilian objects?
I think that time is coming and I wonder if Google is going to use its wealth of personal data to identify personnel who can teach its AI. What is good? What is bad or are they already past that point? And they have AI teaching AI to identify military objects and assets away from civilian.
Re: (Score:2)
What a shame (Score:2)
Diversity does not equal white/passive aggressive anymore. They got some mileage out of it before it died though.....
My surprise (Score:2)
Re: (Score:2)
They had trouble hiring smart people. This way they at least got highly intelligent naïve people. I was at one of their hiring events with Page and Brin back when they pretended to take that motto seriously and where still relatively small.
Google motto de-evolution (Score:2)
1. Don't be evil
2. Be only a little bit evil
3. Only be evil on even numbered days
4. Be evil, our competitors are
5. Make evil weapons
6. Lie about making evil weapons
7. Automate lying and being evil
8. Lie about automating evil
9. Blame it on the poor and downtrodden
10. Merge with Satan
Re: (Score:2)
I believe your number 10 is wrong. It should be, "Reveal they were run by Satan all along." Then, 11 would be, "Zuckerbot attacks claiming trademark violation."
Good (Score:2)
Even if a weapon is never deployed, the state of MAD needs to be maintained.
Development of weapons does NOT contradict a do-no-evil mentality.
Such a surprise (Score:2)
I wonder whether AI will be the end of the human race after all.
The abridged commandments (Score:2)
Honestly google makes itself look like the Animal Farm. Changing its slogans and guiding principle all the time.
I remember when they changed "Don't Be Evil" to "Do the right thing."
Now they just delete a pledge that they probably only made a short time ago in the first place. They really would be better off not making all these pledges and pretending. The truth is, the company postured an ideal of nobility that it probably never had a chance to live up to, and now has a whole bunch of disgruntled employee
Legitimate target (Score:2)
If Google makes weapons then Google has made themselves a target, if those weapons are used in a military conflict.
Not a smart move.
Strange journey. (Score:1)
Profit first and always (Score:2)
Re: (Score:2)
I think it's wonderful that they're sticking to (Score:2)
I think it's wonderful that they're sticking to their current alignment, which is Lawful Evil. You could easily see them shifting to Chaotic Evil after the election of Darth Trump but, they do have a sturdy foundation. Well, after the shift from Chaotic Good, I guess they don't. Maybe they've found their way, like Yussuf Islam.
Living up to its new motto (Score:2)
"Don't be Evil" (Score:2)