It isn't like governments aren't aware of these tactics. They can and do work to substitute out vendors once they get too expensive. In fact I've heard whispers of several customers of Palantir doing just that.
(Disclosure: I work for a competitor of Palantir.)
This is one of the more over-used quotes. In this case, it's not particularly relevant since US fiscal and monetary policy is almost entirely designed around incentivizing long-term investment over short-term gains.
If Palantir is indeed a good "long run" bet, then every pension fund, 401(k), mutual fund, or ETF (read: most institutional investors) should in theory eat it up after the IPO.
To borrow another pithy quote:
"Show me the incentive and I will show you the outcome."
- Charlie Munger
or maybe not in theory but in practise, the entire stock market performance has shifted so strongly towards technology companies due to the general low growth environment we're in that the whole thing comes crashing down in five to ten years because there's actually no business model
This is common to consulting companies, Palantir is more like McKinsey & Company/Accenture/The Boston Consulting Group. The recent deals with the UK gov for £1 are classic get your foot in the door then bleed them dry once the switching cost is high.
Anyone want to throw me $1B? I'll just burn it over 20 years but maybe in 2070 I'll make a profit.
Government contracts take a long time to close, but once closed they generally stay closed and often expand. Defense, in particular, is the most enterprise-y of all enterprise clients. So it's reasonable for Palantir to outlay a large fixed cost (in the form of sales, consulting, and integration) in exchange for a perpetual, increasing annuity. Their quarterly losses are a result of them closing many new deals that are in the integration phase and haven't yet started to monetize fully. Notably, Palantir's per-deal long-term margins (70-90%) are closer to those of a SaaS than to those of a Big Four consultancy (~30%).
The "lose money now, make money later" playbook is common to all SaaS companies and works phenomenally well. Palantir's twist is that they do significant integration and custom work up front, at the expense of less sales staff. If you dig into their numbers, it's hard to argue they haven't made a strong case that this model works.
So while I understand the skepticism at their high-level metrics, the details paint a more optimistic picture than "low interest rates made this bad company viable" (an argument I've seen elsewhere). No IPO investor wants to subsidize losses indefinitely, after all.