Been a while since I’ve written. This blog isn’t a safe space for me anymore, for various reasons. Blogs written under one’s own name are not what they used to be. Personal blogs for shooting the shit don’t exist anymore; there are other places you do that now, while blogs exist as a long-form medium, and as a luddite refuge for those abandoning social media and returning to self-hosted. Remember finger .plans?
I’ve wanted to write about ethics in software engineering for a long time. But afore-mentioned safety and familial/work priorities have consumed me. Also, once I started to gather links for writing about this, it was not soon thereafter that the whole Google employee mini-revolt happened, and those people spoke to everything with far more authority and experience than I or anyone I knew could.
Later I wrote a draft of this but didn’t publish it. Now I’m just banging out something quick, with less fucking boring writing flourish.
Let’s just state up front that cohesive industry ethics in software engineering is a non-starter. Doesn’t exist now, won’t demonstrate much power in the future. The idea that engineers can influence the course of industry is hopeless. My having worked with software engineers, most are either too self-centered, weak within the industry, soon to be cast asea by a wave of over-saturation, or demotivated to be significant influences on anything in general.
Though, it might be overstated that engineers are assholes for the most part. Individually the engineers I meet are cool, and they generally want to do no harm, and it’s usually they themselves who get in, well, get in their own way more than anything else.
The main problem I see is that engineers overlap a lot with the type of people who disengage. You know what I mean? Too jaded, too measured to actually join a party or pick a religion or take any sort of affiliation. On the sliding scale of “radicalization”, a Facebook post is the lowest bar to entry, while silence in the light of fretting over NYTimes and The Economist articles is the majority middle, and then marching is the pinnacle of activism. Those who have ambition for more are, usually, sociopaths whether they realize it yet or not.
So you have a bunch of engineers who actively avoid grassroots community, who shun all the major political blocs of power such as unions, religions, local government, etc. And not because they don’t have time (myself being a parent who’s struggling to stay on top of his own life, let alone anyone else’s), but because they lost faith in those institutions.
We’re going to expect these people to defend the rest of our institutions? We expect these people to not take a salary if it means compromising our own privacy, which is being used against us and weaponized by people who don’t give a shit about “good faith”?
I identify with a un-mobilized Gen XY group of people who remember the unbridled optimism for extreme sharing of personal information, when the Twitter API stoked the imaginations of developers everywhere and early adopters wanted to share anything and everything to explore where we could aggregate information for the public good. It also seems so naive now, now that not only everyone and his mother have weaponized that data against weaker people, weaker institutions, and weaker voting districts.
After the revelations about Uber actively building tools to troll and confuse its competitors (e.g. Hell, which spied on Lyft drivers) or even cops (Greyball, which identified law enforcement and avoided them), and Cambridge Analytica sucking out data from Facebook’s API before Facebook began to lock it down more, thus allowing even an intern to work on simple scripts to calculate how best to gaslight specific demographics of voters, it just seems like we need to accept that it just takes a handful of engineers who don’t give a fuck to compromise the way of life of decent folk who just want to aspire to something a little better than what they had before.
Let’s just prepare for the worst.
Higher impact from an extremely small number of engineers means that you can’t hope that organizational friction will prevent people who don’t give a fuck about society and the social good from fucking up your life. At best, engineers want to work on cool problems with a cool company that’s going to net them a lot of money. At worst, engineers who know where their bread is buttered will do whatever it takes, even if the code they write is morally and ethically dubious.
To some degree they will chase whatever the “in” problem is. Right now, the “in” problems are machine learning, deep learning, artificial intelligence, big data. The distribution of easier and easier open tools to handle these problems (like TensorFlow) combined with the ease of using cloud computing dramatically drops the barrier to entry for developers.
Even if most engineers chose not to use these tools to manipulate governments or citizenry, or to allow paramilitary or military to control citizenry more, there will always be enough engineers to cobble together something that can accomplish the goal.
The common good of algorithmic and tooling leveling up by companies such as Google, Uber, etc. are going to be weaponized by authoritarian governments as well as people who never succeeded at anything at life unless it was defrauding other people.
Engineers for the most part don’t even need to willingly write code that can be used against ethical boundaries.
Engineers can easily deceive themselves into seeing code without any humanistic or moral concerns. And it will be easy to justify the widespread abuse of androids once we have them; people are easily cruel, and Spielberg’s A.I. is no exaggeration of our inevitable future, if we can’t even treat refugees, poor, and minorities with human decency. Engineers will help sow the seeds of discontent.
Engineers can solve most of the difficult problems in completely unrelated realms of knowledge; all it takes is someone with the knowhow to cobble that code together, remixing and repurposing it for whatever nefarious task s/he wants.
This is the core inevitability of technological progress: the massive, rapid tearing down of barriers to entry.
So, I am pessimistic about engineers, who are not unionized and who have nothing like the Hippocratic Oath. I doubt that there will be a unified stance against obviously morally and ethically wrong projects, let alone morally ambiguous projects.
On April 12, DefenseOne published an article that landed like a bombshell within Google. Drawing upon conversations with Pentagon officials, it revealed that Project Maven was actually a pilot project for future collaborations between Google and the military. In particular, Project Maven was part of Google’s push to win the Joint Enterprise Defense Infrastructure (JEDI) contract.
JEDI is the military’s next-generation cloud that will network American forces all over the world and integrate them with AI. It’s basically Skynet. And all the big cloud providers want to win the contract because it’s worth $10 billion.
Google engineers from several different parts of the company — cloud, AI, communications, the Google Brain team, and DeepMind, among others — strongly objected to Project Maven, which would have provided machine learning assistance for military drone targeting.
As they can. They have the luxury of a job for life, for their resumes grant them free passage throughout the kingdom. Google on your resume means you’re golden.
Despite the ethical issues, it’s hard to blame companies like Google and Amazon for taking dubious contracts or working in China or whatever. It makes a lot of sense in the long run. Those companies have no obligation to take a stand, even after the CSR movement.
Like most fucking things in life, everything happens according to the easiest plan, except those things that don’t: underdogs, the exceptions to the rule, the times when people said fuck it.
Those are the things I want to care about and focus on. Who fucking cares if it’s ultimately successful? It’s drawing your red line, and sticking to it. Those are the moments worth remembering.
But Google will eventually be over-run. Facebook already has been. Twitter already has been.
What’s worth remembering in that story? Perhaps the morsel that there were Google engineers who became known as the “Group of Nine”. I wasn’t able to find much on who they are, but perhaps Tyler Breisacher and Liz Fong-Jones were part of the movement.
Let’s also give Rekognition to Amazon employees for their letter, which demanded that Amazon ban Palantir, “the data firm that provides intelligence to U.S. Immigration and Customs Enforcement (ICE) and the Department of Homeland Security (DHS), from using AWS in light of widespread outrage surrounding immigrant detention practices at the border.”
“Microsoft C.E.O. Satya Nadella, on the other hand, has defended his company’s work with ICE, saying that Microsoft assists the agency only with “legacy mail, calendar, messaging and document management workloads.” And even if some Silicon Valley companies step aside, others will inevitably arise to take their place. Palantir, NBC News reported this week, made more than $4 million last month alone from its ICE contract, which began in 2015. In March, Thomson Reuters Special Services reportedly signed a $6.8 million contract with ICE, using data targeting to help “locate, arrest, and remove criminal aliens.” And Hewlett Packard Enterprise, which publicly denounced the Trump administration’s child-separation policy, signed a $75 million contract in September 2017 to run the U.S. Customs and Border Protection’s network operations center.”
I used to work for a sub-contractor for DHS — while there are definitely worthwhile projects going on within DHS and its agencies, there is little in the way of a moral compass guiding projects which could easily slide into abuse as opposed to safety.
Okay, so there’s no middle ground. The industry is doomed to provide its brightest minds to enable the most evil of intentions. Where do we go from here?
Some people think we should pledge to whistleblow and refuse to take part. I think it’s rarely this clear-cut; typically you’re in too deep before you even realize you were a part of it.
I think the solution is to swear to a Hippocratic oath (ACM has a code of ethics, as an example) and build parallel institutions, starting from zero participants and only taking in those who not only want to work for ethically-minded companies, but companies themselves who only pursue ethical businesses and internal practices.
It’s the only way to actively dissuade profiteers, as well as to keep out and identify information operations geared towards bribing those profiteers and wedging them between everyone else.
It allows communities with fewer controls and less rigid oversight to let edgelords and trolls know that they are not welcome even if they’re just joking around lol.
It allows us to break away from trying to convince people like Mark Zuckerberg and Jack Dorsey that their complacency with their platforms are enabling the most evil of society while harming the weakest of society.
It helps create a foundation for markets which don’t need to rely on inherently morally dubious business to succeed.
It helps engineers and companies refuse to hire assholes, destructive opportunists, and mercenaries.
It forces people who want to take part in a better society to actually stake their claim to it, and take a stand.
It gives individuals some leverage to begin saying to more powerful people, “You fucked with me. Now you’re cut off. For life.”
Where do I fucking sign?