Potato potato. Palantir explicitly provides tools for data analysis and operations for government and enforcement agencies. Postgres is a database used for virtually anything that needs a database. The two aren't the same.
For the record - Palantir isn't 100% culpable of a government resource abusing its capabilities, but it sure unlocked a whole bunch of capabilities that were either too expensive or too difficult to do previously (for example, storing records in an RDMS).
> Palantir, the tools came with a robust ACL solution.
ACLs require humans to configure them. Uber also has a robust ACL. It doesn't stop someone in the org from using and abusing its God-mode.
Yes, absolutely. I think it's very common for people to make ethical decisions in how they make their money and how they spend their money.
If you're asking if every citizen living in Germany from 1933-1945 did only evil things, the answer is obviously no. If you're asking if every person who was ever in the nazi party was a uniquely evil human being, also no.
Actually there are us who would consider what impacts our employers have. E.g. I avoid working for adtech, and sure as hell would not want to work for a company providing government surveillance software.
For example, your job at the ad tech company could be anonymizing data and protecting people’s privacy. Your job at the children’s charity could be scamming old ladies.
This only happens because companies need to adhere to regulations, not because they're doing it out of respect for people.
And by simply not working in adtech, I don't need to go through some mental gymnastics to justify what I spend 40 hours a week building. The beauty of being a programmer is there's a bunch of work out there that doesn't involve crappifying the internet.
Nobody involved in that decision is motivated by respect? You're sounding pretty pessimistic for such a big emphasis on morals.
Go ahead and share what industry you work in then. I guarantee it is not unambiguously good.
“And when the scribes and Pharisees saw him eat with publicans and sinners, they said unto his disciples, How is it that he eateth and drinketh with publicans and sinners?”
You are not a moral person for working at an elementary school instead of a car manufacturer. The question is actually what you do, what talents you have, and how that affects other people.
Computer ads aren’t evil, surveillance, spam, and scams are evil.
I honestly don't have a clue how what I said could be interpreted as arguing that all buying and selling is immoral, so I wouldn't even know how to begin addressing that line of thinking.
I am not. Palantir not only brands themselves as providing data analysis for governments but they also actually sell those tools too.
Do you call your gas company to make sure they didn’t buy from OPEC?
The impact is exactly what is pointed out in the article - someone in the local police department could use it against me.
> Do you call your gas company to make sure they didn’t buy from OPEC?
No, because that doesn't affect me.
See, you keep clinging to some morality claim, and I am explicitly not. I am saying that there is potentially a direct impact on abuse against me personally.
“Moral” in quotes is a tell, too. You’re arguing that morality is illegitimate unless someone has done a full “utilitarian” breakdown of everything they do in their lives. And of course you will only accept that breakdown if you deem it sufficient.
This shallow label analysis (oil bad, palantir bad, non profit good) is not “morality”, it’s politics.
Most people espouse a utilitarian ethics, but rely on media and social cues rather than harm to determine their choices.
If your job is to pour barrels of oil into the ocean, I think that’s worth evaluating. If your job is to do accounting for the oil company, I’m not sure you’re causing harm relative to other accounting employment.
It’s also valid to decide image is important to you, but don’t tell me it’s because you are a Good Samaritan.
The name of the company itself is another dead giveaway.
1. Or carbon monoxide mask, if the distress of asphyxiation is throwing you off the from point I'm trying to make
Sure, if PostgreSQL were specifically selling their tools to organizations known to commit those sort of abuses.
Palantir for example works directly with the government on State and Federal levels, and know damn well what they are doing, what their tools are used for, and answer directly to the requests of the government in regards to contracted work (all at the cost of the taxpayer mind you). These are mass surveillance tools, they have one purpose.
I also don't think its insignificant to recognize the backroom deals here that have created this vicious cycle of:
working as a govt official and giving kickbacks and heavily inflated contracts to contractors, and forming laws favorable to those contractors -> then getting a consulting job at those same defense contractors or lobbying groups -> and then moving back into politics
we can't just strip away the context here, a database has a purpose, what purpose does mass surveillance tools have? I'd argue that there is no proper use case for these tools against Americans, other than authoritarianism.
That said, I do find your argument interesting as it parallels the "guns don't kill people" argument. An I am a gun rights advocate yet am against Palantir's product usage on US citizens. Good food for thought.
The general form of both arguments is that you have a tool that can be used for good or bad things, so should you ban the tool (guns) or just the bad uses of the tool (murders)? If you had perfect enforcement then it would obviously be the latter, because you'd arrest the murderers without harassing people who buy guns for hunting or self defense etc.
But law enforcement isn't perfect, so there is going to be some spillover where some murderers wouldn't have been caught. Then you have to get into debatable questions like, if it was illegal for them to buy a gun, would they have just bought a gun illegally, or used some other kind of weapon? If there are a lot of murders happening, how many of them could be prevented by using more resources to catch them or e.g. legalizing drugs to reduce gang activity, without banning guns? The pro-gun argument is that the positives outweigh the negatives and/or we're better off doing the other things to prevent murders instead.
You can ask similar questions about Palantir and mass surveillance, but that doesn't mean you have to come to the same conclusion, because it has different positives and negatives and ability to be mitigated if you do it.
In particular, preventing abuses by governments is much harder because you'd be relying on the government to prosecute itself, which is notoriously ineffective, especially with programs that are insulated from public accountability through secrecy. Meanwhile the negatives are much more dangerous because of the speed, scale and severity at which an already-existent mass surveillance system can be converted into a tool of mass oppression when there is a change in administrations. And the positives are muted because the legitimate goals of mass surveillance can also be achieved -- often at lower expense, given how much it costs to sift through a firehose of >99% false positives and innocent behavior -- using traditional targeted investigative methods.
I disagree with your framing. The bill of rights specifically guarantees both the right to bear arms and the right to not be unreasonably searched.
Suppose you have a private mass surveillance system, where people "voluntarily" (i.e. because the market is consolidated and every company is doing it) give up their data to a private company that keeps a big database. Then the company "voluntarily" informs on citizens to the government (because the company wants to be on the government's good side), or imposes penalties for crimethink all on its own. Is that okay then? Should we be satisfied with it because it's not the government? Nope. Still a big problem.
The reason isn't that it's the government, it's that it's a bad trade off.
Suppose you're the "good guys" and trying to catch terrorists. There are 330 million people and 50 terrorists. If you have a mass surveillance system with only a 0.1% false positive rate, the system is totally useless. You have 330,000 false positives and investigating all those dead ends would be a huge waste of resources that would be better spent using traditional investigative methods on traditional leads.
Now suppose you're the bad guys trying to catch resistance fighters, with the same mass surveillance system. 0.1% false positive rate? Okay, round up the 330,000 people and execute them all without trial. Way cheaper than having to sift through them. Very effective system when you don't have to care about that guilt or innocence stuff.
And that's what makes the system so dangerous. It's only useful in proportion to how little you care about innocent people, and the extent to which an entity does is a thing that can change over time, so it's massively dangerous to leave a system like that sitting around without vigorous efforts to dismantle it.
In the US we just use people who didn't think about it too deeply.
If I had to choose to be surveilled 24/7, but my data stuck, unlabeled in there, or 20% of my actions randomly sampled, but fed to a tool that can interpret it, I choose the first, hands down
I'm not advocating for the company, but the statement in the comment is shallow. I was a̶b̶u̶s̶e̶d̶ assaulted while employed there, which is why I left, but my comment stands.