Can anyone using Palantir explain what it does? I've never heard or seen anything concrete - I had one redditor tell me his company looked into it and they just put data into a database and make "connections".
You must have seen the same demo that I did... when they were recruiting at a university jobs fair... ten years ago...
Something along the lines of: okay so you have this guy who's a criminal right? (expand node) here are his phone numbers (expand node) here are the phone numbers who called it (expand node) here are the phone numbers who called those numbers (expand node) here are who they're registered to (expand node) here are their addresses (expand node) here's who else lives there.
I still assume their special sauce isn't their ability to build a node expanding UI, but the data sources they have access to. If they're used for so many government contracts, they're definitely using the data from, say, one police department to inform another. So when another department wants data about criminals, they sign up as a client for the company where that data already is.
And don't discount that the government would pay a huge sum of money to build a surveillance state, even if that surveillance state turned out to be pretty useless.
Pretty, but above all convenient , UI matters much more than programmers think. It's the difference between seeing a tree of comments like on HN, and a raw Usenet stream.
So it's like observability metrics but for ships and missles and such that people care about in the real world. Very cool. I did find the ending where the commander literally selects from a dialogue tree to resolve the situation a little ridiculous. Sorry commander, that choice of action is not ethical, is there something else I can help you with?
Back in 2011, I joined a software startup working on a product which aimed to prevent stuff like the Deepwater Horizon explosion. The founders talked about how throughout the industry and the world, tons more data was being generated than ever before (of course), but we (humanity) didn't have a good handle on making use of all of that data, so a lot of it was not being interpreted, reacted to, etc.
Not long after that, I found out that Palantir got their first energy industry customer. Some employees in Palantir thought it wouldn't be worthwhile to go after oil & gas companies because they viewed them as old school and slow to adapt to new ways of doing things. However, in those days Palantir was focused on acquiring no more than one customer per industry, because they were trying to boost their valuation as quickly as possible and wanted to demonstrate how many industries they could serve. And then, due to a contrarian within the company, they did get an O&G customer, and it was the biggest single deal so far in the company's history.
What I don't know is exactly how well they served that customer. What I do know is that they were selling the customer on the idea that they have sooooo much data and they're leaving tons of low-hanging fruit that could easily boost savings in the company. Stick all this data into Palantir's product and we'll help you on a million fronts and therefore boost your profits significantly. One example is the longevity of machine parts and all the data you can get about historical wear and tear, etc. Now, this is what they were selling, and I don't know which areas ended up being the most helpful, but yeah, as someone else in here said, they were insisting on customers putting in all the data they have and the product would make connections, or reveal insights, etc.
Even if it's a crude product by Star Trek standards, you've gotta admit there is probably a ton of low-hanging fruit like that in most companies, and a lot of the time they don't go after all that because they're making money, but just don't have extra manpower or expertise to look into all that, so if Palantir delivers on the promise to boost the bottom line, then I guess it's worth it. It also means there's room to compete, although a big problem is even if you're a sharp cookie, Palantir probably started with way more business connections than most nerds will ever have.
Plenty of HN discussions (like this: https://news.ycombinator.com/item?id=41855006). Pretty much a consulting company, they don't own their dev stack. People see 80% gross margins and think it's enterprise SaaS, but they are just great at selling their "FDEs" at 5x cost.
I find it interesting the number of responses that had success after gathering e-mails. If your 'Show HN' asks for an e-mail here, everyone gets incredibly angry and lets you know they stopped looking at your product right away.
I think too though for some submissions there's a discrepancy between readers vs commenters, I think in part because HN as a site is attractive to certain types of users (text only, more web 1.0-y and refreshing vs gamified forum experiences), which leads to some skew in the demographic of who choose to register as users and engage.
See for example the disproportionate number of comments on HN from users who disable Javascript and or (all) cookies and remark about a site experience based on it (can't read article, some aspect not working, etc), yet from general statistics represents a niche minority.
(Which isn't to say that any such critiques are invalid either, just an observation on perceived audience)
I just realized two implications of this HN demographic:
* Good: That side of HN is within the target for a little niche Web site I plan to launch in a few weeks.
* Bad: I used a Web framework that requires 10x the hosting resources that it should (resume-driven-development), so I might have to upgrade from free-tier hosting before HN mention, just to not be embarrassed by hug of death.
What framework? I'm getting into full stack and I was under the impression that all the shiny new stuff is focused on being more efficient, rather than less?
Asking for emails allows you to target people who are more likely to engage with your marketing, while people who find this annoying would probably be more critical of the product and in general are more picky. In the end they still give you engagement in a form of comments, so it's a win-win. Similar strategy is leaving easily spottable typos in your post so that nitpickers can't help but comment on it.
I still have my original copy of "The MAD Adventures of Captain Klutz", probably bought around 1970ish. Such a singular talent. Died pretty young (68), which is sad.
Seems like a pretty common gesture when you are on the phone, someone is talking to you on the other end, and you want to get another person's attention in the room who is not looking at you. You are reading way too much into it.
reply