My understanding of Palantir's actual, technical offering is profoundly boring: a hosted platform that connects to existing diverse sources of data and organizes them according to well-defined (by Palantir's FDEs) useful schemas. I have developed this impression through actually building a product on the Foundry as well as several rounds of interviews. Frankly that is profoundly boring. The anti-Palantir propaganda, portraying them as this all-powerful Skynet software, is as much a part of their marketing as anything else.
On the other hand, their effectiveness appears to be less in question: the article above claims that Scotland Yard found hundreds of police officers to have been abusing their posts in various ways through use of the Palantir system. I am not a fan of corrupt cops, so I think this is good. Similar stories exist elsewhere, like a 68% reduction in 48-hour mortality at a Tampa hospital through deployment of Palantir's anti-sepsis monitoring tech.
Thus I arrive at the conclusion that this decision is ultimately a loss. Khan's legal standing appears to rely on them not investigating other potential suppliers—I'm not sure that there are any, and "develop these simple data systems in-house" is a bad option because if they could have they would. I suppose ultimately I don't think that Palantir's "bad vibes" among constituents should impact governments' desire to be effective in the programs they purport implement.
This is far more than just "bad vibes", and just a handful of many examples. The vibes also tend to be pretty bad around things that are used to enable spying, a secret police force, or bombing children.
This isn't even touching on the name of the company itself and the origin of that name, or the fact that Peter Thiel founded it, or many of the other things that give it "bad vibes".
Palantir revolutionized the enterprise software playbook (more government than enterprise, but I digress) by investing heavily in Forward Deployed Engineers, Palantir engineers deployed at customer sites and working hand in hand with engineers at the customers to make it happen. Most software companies pay only lip service to customer success, and seldom provide any engineering after pre-sales.
You don't have to like the company to respect the hustle. I deem them utterly despicable, on par with IBM who sold the Nazis the tools to round up and exterminate Jews during the Holocaust, and indeed their UK division is run by the grandson of Oswald Mosley, founder of the British Union of Fascists.
Yes, the law of conservation of social capital applies in the UK. There is this inexplicable fascination with the despicable Mitford Sisters (one of which was Oswald's wife). The Germans don't pine for the glamor of Leni Riefenstahl.
I discovered Radicle back in 2020 (when their website looked incredible: https://web.archive.org/web/20201201030505/https://radicle.x...). I bounced off of it, in part due to being unable to effectively delete repositories. They used to have an FAQ about that—looks like it's gone now, though the public-private repository area is much more fleshed out (you can make a repo private, in which case no new updates will be publicized but the history will still exist). In truth, it's just profoundly difficult to effectively "delete" things in a decentralized system (see: Matrix, BitTorrent, et. al.). But definitely something to consider; people accidentally upload secrets, and want to have some recourse when that happens.
Still, time has passed and I have become more interested in GitHub alternatives (https://figbert.com/posts/ideating-tragit/). Will likely end up moving to Tangled. But first I need to add support over there for pushing over HTTPS...
In general, whatever has been made public, is hard to unmake public. There can always remain a copy.
It is acutely true for peer-to-peer distributed systems like Radicle, but is still true for the more centralized GitHub, and for the Web in general. If you want to be able to walk something back, better don't publush it.
Publishing a token or an ssh key should not be a big deal, such a token or key should be immediately revoked, which is as good as deleted. The problem occurs when the fact is not noticed immediately. A git hook can help avoid such mishaps.
It’s a fair point about the difficulty in deleting data in a decentralised system but that’s also true in more centralised systems like GitHub or any other website. Once some data is out there, you have no control over whether it can be removed. Other parties could have copied it and may re-share it. “All” removing data from a centralised system does is slow down this spread, sometimes to the point that it is effectively deleted but that can’t be guaranteed.
The best response I’m aware of is to invalidate any secrets that have been accidentally shared. Sometimes that is easier said than done though. And of course there are plenty of other reasons someone might want to delete data.
It's not a technical issue but a UX one. It would be a pain to set up pushing via SSH to a knot container on my server because I am use SSH for the host already. And specifying a port is ugly. Doing this over the internet negates all these.
Location: Palo Alto, CA, USA
Remote: Much prefer in-person but if compelling I can be flexible
Willing to relocate: Yes, anywhere in the world
Technologies: Software (Go, Swift, Gleam, Ruby, Python, WebDev, Deployment), Hardware (CAD, rapid prototyping, DfM, drafting, mechanical engineering), Systems Thinking
Résumé/CV: https://figbert.com/files/cv.pdf
Email: figbert [at] figbert [dot] com
I'm an industrial design junior at Stanford University looking for summer work at the intersection of software and hardware. Want to have a tangible impact on the creation of meaningful products. Love working on diverse problems, diving in on areas where progress has stalled and getting things going again. Feel free to reach out.
I was tricked into trying Gleam earlier this summer, and really liked it! I anticipate that I will at least use it for all future web projects (one example: https://github.com/FIGBERT/bdab) due to my serious JavaScript allergy.
Gleam Decoders are something I haven’t fully wrapped my head around, but are supposedly very powerful and do exactly what this article is focused on (parsing input data into Gleam types) in a more(?) idiomatic way.
I increasingly have come to believe that it is the screen itself that lies at the root of the ills of technology. It brings so much benefit—and so much convenience, from its flexibility—but it is in its fundamental glow-y rectangular nature that sucks us in, crushing our attention, posture, and so much else. Was incredibly fun to experiment with something radically different.
If this is your website, as a heads up it doesn't work well on my browser. Firefox on Android, I believe I have a dark mode and the text is still black but on a very dark background.
ChatGPT?
Seriously though, this is such a weird reply and doesn't fit at all with the account's previous comments. Also has that kind of not-quote-right feel that a lot of ai generated content has
On the other hand, their effectiveness appears to be less in question: the article above claims that Scotland Yard found hundreds of police officers to have been abusing their posts in various ways through use of the Palantir system. I am not a fan of corrupt cops, so I think this is good. Similar stories exist elsewhere, like a 68% reduction in 48-hour mortality at a Tampa hospital through deployment of Palantir's anti-sepsis monitoring tech.
Thus I arrive at the conclusion that this decision is ultimately a loss. Khan's legal standing appears to rely on them not investigating other potential suppliers—I'm not sure that there are any, and "develop these simple data systems in-house" is a bad option because if they could have they would. I suppose ultimately I don't think that Palantir's "bad vibes" among constituents should impact governments' desire to be effective in the programs they purport implement.