I’ve never used Palantir. I’ve never used DCGS-A. When I started as an Analyst you (no-shit) used pencil and paper (and a thing called a guhor stick…but that’s a lewd joke for another day). The kerfuffle over Palatir vs. DCGS-A reminds me of the days when computers started making in-roads in analysis shops, and I hope everyone involved can remember some of those lessons learned.
Now my working world in those early days wasn’t entirely computer-free, but back then computers were where you stored data and recorded activity and typed up reports, you didn’t “link” things together and you certainly didn’t draw, graph or do anything anyone coming up in the business today would recognize as computer-oriented.
If there was a quantum leap in the utility computers gave to analysis it was this application called Analyst Notebook. Analyst Notebook would take in the data you had already entered into some other system (assuming you could get it out of said system), and kick out diagrams and pictures that let you make quick sense of who was talking to whom, what happened when, and identify connections or anomalies you may have missed staring into a green screen at row after row, column after column of letters and numbers.
That’s the key here: Analyst Notebook, Palantir, etc. are Analyst’s tools, they are not analysis tools. Is that a distinction without a difference? I’m not aware of any software application that will think on your behalf. I’m not aware of anyone in the military or IC who would trust answers produced entirely by an algorithm and without human interpretation or enhancement. If you could computerize analysis you wouldn’t have a headcount problem in the IC. Analyst Notebook, Palantir, DCGS-A . . . they’re all tools, and if you’ve been working with hand tools all your life and suddenly someone hands you a Skil saw, of course you’re going to think the Skil saw was sent from heaven.
Now, is the government notorious for producing bloated, expensive, minimally functional software that everyone hates to use (when it works at all)? We don’t have time to go into all the examples, but the answer is ‘yes.’ If I offer you tool A OR tool B when you’ve been using tool C, which are you going to choose? Does that make your other choice crap? Of course not.
It sounds to me like if there is a 800 lb gorilla in the room it’s usability, and if there is one thing that commercial apps excel at its the user experience. Think about the Google interface, and then think about a data retrieval system fielded in the 70s, and you tell me what your average analyst would rather use…
If the ultimate requirement is capability, then the answer is simple: hold a shoot-out and may the best app win. Pretty-but-sub-capable isn’t going to cut it; functional-but-frustrating isn’t either. If DCGS-A is all that, they should be big enough to learn from what Palantir does well; If Palantir is really about saving lives and national defense, they ought to be big enough to implement what GIs need most. Competition raises everyone’s game, but this isn’t about .com vs .gov, it’s about lives.