01 Sep 2018
Hey hey! Tableau! Your ICE contract has got to go! Yesterday we were out at Tableau’s Seattle office with @MobToMeltICE and the IWW, demanding that the company end its collaboration with ICE.
On Tuesday, we were invited to speak at the United Nations in Geneva, Switzerland. We joined our friends from the Campaign to Stop Killer Robots in amplifying the call for machines to never be allowed to authorize an action that would harm or kill a human being.
Imperfect data + imperfect programmers = imperfect AI. AI will not magically get smarter; it will only get better at serving the interests of the institutions who build it or pay for it to be built. So as long as those institutions — military, police, or otherwise — violently oppress others, AI will be a force of oppression.
Read the full statement from Tech Workers Coalition at the end of the newsletter or watch the livestream (start at 17:40) here.
We’re encouraged and invigorated as more voices speak up to hold tech companies accountable and take steps towarda more ethical tech industry — we’re excited to see this work through to the end.
Labor Day Action with Unite Here! Local 2
Monday, 9/3 10AM meet at Market Street between the 4 Seasons & Marriott hotels in San Francisco
Follow #1Job on Twitter
Seattle TWC Monthly General Meeting
Thursday, 9/13 6:30PM at Seattle Labor Temple Association in Seattle
Bay Area Anarchist Bookfair
Saturday, 9/15 10AM at Omni Commons in Oakland
Learning Club: The Revolution Will Not Be Funded
Saturday, 9/22 2PM at Omni Commons in Oakland
The Code of Conduct is in effect at all TWC events.
In The News
Some employees at Facebook think their company has a problem with “political diversity.” For these employees, “political diversity” means being able to freely demean, silence, and harass co-workers on the basis of their race, religion, and LGBTQ status. As tech workers, it’s our responsibility to respond by sharing another vision of diversity, real diversity, which is grounded in empathy and respect.
See the latest statement from individuals joining the national #PrisonStrikeand spread the word. Read Kevin Rashid Johnson’s account of the brutality of prison authorities and tactics used to silence imprisoned people and perpetuate exploitation. Prison labor is slavery.
Recruiters are feeling the pain of boss-worker clashes over ethics and workplace conditions. “For Geiduschek, the recruiting email from Amazon felt like a powerful “personal leverage point,” an opportunity to share her disapproval of what Amazon is doing in a way that might actually get people to listen. “If it’s just me, it’s not going to change,” she said. “But if they felt like 10% of their recruiting pipeline was turning away when they would have otherwise taken a job, then I think it really would become on a lot of executives’ radars.”
Amazon is bribing employees to write positive tweets, in an attempt to cover up the company’s unrelenting exploitation and atrocious working conditions. Also, a rare public response from Amazon to the recent outpour of criticism.
Fun fact: Human Relations and Organizational Psychology were created by Gilded Age industrialists, with the aim of “creating psychiatric techniques to keep workers from radicalizing.”
TWC Addresses The United Nations
Hello, my name is Amr Gaber. I am a very low level Google software engineer and I’m here speaking on behalf of myself and not my employer. I am also a volunteer with the Tech Workers Coalition. We’re a labor organization of tech workers, community members and other groups that are focused on bringing tech workers together from across industries, companies and backgrounds in order to make technology and the tech industry more equitable and just. I’m here to talk to you today because I took action with several of my co-workers at Google to stop Google from pursuing artificial intelligence in military applications. So there is a contract called Project Maven which is also a larger Department of Defense project and Google’s involvement was to use artificial intelligence to enhance drone surveillance and targeting through image recognition. So I heard about this project from my co-workers internally through whistleblowers who were sharing it in the company forums. I co-wrote an open letter with several other Google employees to the CEO Sundar Pichai asking Google to cancel the project. Over 4,000 of my co-workers signed it, but that alone was not enough.
It was actually a combination of workers, academics, students and outside groups that eventually got Google to step back from the contract. Organizations like the Tech Workers Coalition, Coworker.org, the Campaign To Stop Killer Robots, and the International Committee for Robot Arms Control which wrote a petition that over 1,000 academics signed including Noam Chomsky supporting me and my co-workers and our call to cancel the contract. After all that there was a concession from Google. In May they said that they would not renew the contract when it was due to expire and that they would not further pursue this line of work. It also accelerated the timeline for the publishing of the AI ethics principles that came out a week later. This isn’t an issue that is specific to Google. It’s actually something that concerns the entire tech industry as a whole and any workers who are working on technology.
After we found out about Project Maven we found out about another contract called the Joint Enterprise Defense Infrastructure which is a United States military program that for all intents and purposes has artificial intelligence as one of the core tenets of future United States military infrastructure. And this is a contract that a lot of tech companies are bidding on — Amazon, Microsoft, IBM and it was supposed to be Google, but maybe that’s less certain now because of the actions that me and my coworkers took. This leads into applications of these technologies in other areas as well. The action we took at Google inspired tech workers at other companies to also take action. Amazon workers asked the company and its CEO to stop selling visual recognition technology to police and to kick Palantir out of Amazon Web Services cloud. Palantir is a company that provides software directly to United States Immigration and Customs Enforcement which helps them separate children from parents and put them in cages.
It also inspired workers at Microsoft to ask their CEO to cancel a contract with ICE, and Salesforce workers to ask their company to cancel its contract with US Customs and Border Protection for similar reasons. And workers in other industries as well: workers at Deloitte also calling on their company to cancel their contract with ICE and McKinsey & Co. — they successfully got their employer to cancel the contract with Immigration and Customs Enforcement. This actually brings up another facet of this technology that’s very important. It’s not just military application and use but also domestic policing, surveillance, and violence towards oppressed groups through automated weapon systems and surveillance systems.
But we’ve got Google’s AI principles, right? So everything’s going to be fine. They’re not going to commit any harm. If other companies get onboard then we’ll automatically only build technology for a better world.
That’s not going to be enough. I would actually point to AI Now’s accountability framework as a stronger example of ways that can make this kind of technology more ethical and more accountable. One of the things that we’ve learned recently after the AI principles came out was about a project called Dragonfly happening at Google. And again, we only learned about this project because of whistleblowers - this time telling the public. One of the biggest problems with this project is that some of my co-workers who were working on the project didn’t know what they were working on. The company hid the true nature of the project from its own employees. This kind of thing is only going to increase as technology gets more opaque through layers of automation, the cloud, and the artificial intelligence and machine learning technologies. I’ll give you a couple of examples.
So a not so recent example was that in 2015, Google’s own image recognition technology was classifying black people as gorillas. At the time the fix wasn’t to actually change the algorithm, it was to remove gorillas as a possible result of the tagging. A more recent example is this year, Amazon’s facial recognition classified 28 members of the US Congress as criminal suspects. The problem is that this technology is mostly based on the training data and the imaginations and skill of the engineers and creators. The data is always going to be imperfect because we have an imperfect understanding of world. The programming and engineering will be imperfect because we’re human beings and we’re not infallible and therefore if you have imperfect data and imperfect creators you’re going to have imperfect outcomes. There is nothing that will change this fact. A program should never authorize an action to end the life of a human being. We don’t perfectly understand the world and we will never be able to make machines understand it perfectly either.
So how do we actually ensure that we build equitable and just technology going forward? Tech companies are massive private organizations that enforce decisions and outcomes through code. Not laws, not deliberation, not police, not militaries. These decisions and outcomes affect billions of lives but they are not beholden to human rights or democratic processes. They’re beholden to shareholders and the decisions of executives. However when these decisions fail to protect human lives it is governments and regulatory bodies who will also be held accountable and responsible. This is where the public interest, employee interests, state, and UN interests align in my opinion.
Tech workers are already taking the lead on these issues but we cannot do it alone. The international community and individual countries have an opportunity to work with us to ensure that future technology is centered on lives and not violence, oppression, or shareholder value. The first step is for you to publicly defend the rights of whistleblowers and protect workers who speak out on ethical grounds. There is a toxic culture in the tech industry right now to attack leakers and that kind of risk taking and speaking up.
In the service of human rights and giving voice to the voiceless, Google employees including myself have gone one step further than simply asking for the next unethical project to be cancelled. We are instead demanding transparency, public oversight, and employee representation and decision-making power. You should openly support us and other employees in the tech industry who are standing up for the rights of all people and to give those on the receiving end of these technologies control over their own lives.
So I don’t usually get to come to the UN very often, and I definitely don’t get to speak. So I’m going to end on one other thing that might be off topic but I think is worth saying. So I’ve heard a lot about curbing the harm of technology and following international humanitarian law in these discussions. But I haven’t heard anything about justice. Banning lethal autonomous weapon systems is an elementary first step in curbing the harm of unaccountable technology. But there is so much more work to do to make sure that machines serve the needs of human beings and not the other way around. Like I said tech workers and other workers are already taking a stand on these issues. But we are tired. Even the most privileged of us. I am a living example of this.
We are tired of weak, selfish and dishonest leadership, bottomless greed and exploitation and we are tired of artificial conflict. We are tired of hearing that these issues are too hard to solve and we have to wait and we have to continue to suffer while the rich and powerful eat the world and make it harder for us to live decent lives and ensure the same for our children. We are standing up now for justice, for freedom, homes for the unhoused, food for the hungry, dignity for all work, keeping families together, and a better future for all people around the world regardless of race, gender, sexual orientation, religion, national origin, immigration status, age, or ability. Join us. Thank you.