Stephen E. Arnold: Palantir Raises More Money — Comment by Robert Steele on Palantir’s Short-Falls

IO Impotency
Stephen E. Arnold
Stephen E. Arnold

Palantir Raises More Money

Sometimes it seems like Palantir is constantly enticing investors. Forbes reports, “Palantir Aiming to Raise $400 Million in New Round.” Writer Ryan Mac tells us that, according to Securities and Exchange Commission documents, a planned round of additional funding could rake in that sum if all shares sell. Mac continues:

“If the round is completed in full, Palantir’s total funding could swell to about $1.2 billion. It’s currently unclear what the company would be valued at if it were to raise the $400 million round….

“Palantir previously raised about $450 million at about a $9 billion valuation in a round that began in December of last year. That round finally closed in September, after the company continued to add additional shares and investors.

“Even at $9 billion, Palantir was already among Silicon Valley’s most valuable private technology companies, some of which have seen massive bumps in valuations recently.”

Current Panantir investors include In-Q-Tel and Tiger Global Management, and co-founder Peter Thiel remains its largest shareholder. It seems unlikely that the growing company will go public any time soon; last year, CEO Alex Karp noted that Palantir’s existing relationships with government agencies and global corporations make the company ill-suited for IPO status.

Palantir’s massive-scale data platforms allow even the largest organizations to integrate, manage, and secure all sorts of data. Its founding members include PayPal alumni and Stanford computer science grads. The company is based in Palo Alto, California, and has offices around the world. It should come as no surprise that they are currently hiring.

Cynthia Murrell, January 06, 2015

Sponsored by, developer of Augmentext

Robert David Steele Vivas
Robert David Steele Vivas

ROBERT STEELE: There was a time when I thought Palantir had potential. Their current strategy reminds me of derivatives banking — keep building up the pyramid scheme and eventually sell out to someone with more billions than they have brains — IBM as an HP – Autonomy redux, or Facebook, both come to mind. What they may be over-estimating is the degree to which mega-billion clients are now much smarter about the severe deficiencies of what Brother Stephen calls Next Generation Information Access (NGIA) offerings that combine search, “analysis,” and visualization — never mind the degree to which they can or cannot access substantive data. Below are some comments I developed a year ago, when I was contemplating Palantir’s possibilities if it could learn to listen.

01 CEO properly focuses on the secret world that helped start Palantir and that empowers him and the company, but that market space is over-crowded, inattentive, non-collaborative, and unlikely to yield the gains needed to justify the magnificent investments received. Palantir needs a process and procedures and contacts that can move Palantir into organizations across all countries within the eight tribes as separate market spaces that can also be brought together by issue and by location: academic, civil society, commerce, government, law enforcement, media, military, non-government/non-profit. Information costs money, intelligence makes money – organizations are weapons systems.

02 Palantir’s developers, with my experience and oversight, can learn what it means to be intelligence professional and practice the craft of intelligence from requirements definition to collection management to source validation and multi-source fusion to time and space visualization to compelling action-enabling reports. Palantir is not optimized to multiple ingestion of sources as found; to create collaborative human networks; to rapidly discover and translate foreign language materials; to create products in each of the four key domains: strategic, policy, acquisition, and operations. The current combination of low-paid youth and selected Special Forces personnel is not cutting it. Palantir needs a major make-over to get past Analyst Notebook and SILOBREAKER (each a relatively low standard in relation to the art of the possible now). In foreign language Palantir needs to invest in providing its customers with expert-level human searchers able to define language strings unique to each mission and target type, creating a global dictionary of search strings that can be easily selected on the basis of English titles, and the results processed once ingested.

03 Palantir is not tuned in to open source everything (OSE) and its information-sharing and sense-making counterpart, M4IS2 (multinational, multiagency, multidisciplinary, multidomain information-sharing and sense-making). There are some specific international communities that could be approached as development partners. There is an open source geospatial play that could be huge. A very low-cost version of Palantir will open many doors, provide access to priceless data as well as human interfaces, and change the game.

04 Palantir needs to expand in tools for identifying human experts in varied languages and mediums, and rapidly combining exploitation of those experts’ published work with selective direct outreach. Apart from language experts as mentioned above, there is an opening for an international conference, to create a Palantir Collaboration Working Group or a series of commercially-valuable Global, National, and Provincial Intelligence Councils that share information and do sense-making together using Palantir.

05 Palantir is largely lacking in the combination of desk-top analytic tools (including anomaly detection and pattern analysis not only in English but in other languages) and back office machine speed processing of both the shallow web (Google) and the deep as well as the dark web (not the same thing). Much more needs to be done in the way of creating an integrated suite of tools such as proposed by CATALYST (Computer-Aided Tools for the Analysis of Science & Technology).

06 Palantir has failed to make the most of its access to M3. Some simple changes there would make Palantir the starting point for field analysts rather than an afterthought. M3 is still “it.” This is its single biggest failure in  terms of being relevant and valuation from within the belly of the beast.

07 Palantir has not learned to co-develop that which machines can do well, and that which humans can do vastly better than machines, both with and without machine assistance. Palantir is lacking in a strategy for leveraging existing big data (Swiss cheese in grid-lock) and being the go-to procurement authority for filling the holes and collecting new data from open sources in all relevant languages and mediums. Palantir also has a reputation as a “hit and run” outfit that puts a lot of emphasis on selling into the top ranks and then leaves the lower ranks unsupported, with field representatives spread too thin, changes not properly communicated, and generally an environment in which Palantir becomes too much of a hassle to use and hence gets ignored at the working level. A collaboration application could be offered both for Palantir charges – notifications, methodologies, feed status, and work flows – and for any instance, e.g. all analysts across all tribes working on Afghan Electoral Reform, for example. It is essential for Palantir to find a way to become the standard for secure but not secret information-sharing and sense-making. Palantir would benefit — but only if it undertook a complete makeover from within — from a new multinational open source handbook that could be used as both a sales and training and collaboration outreach document and living network. What should be the number one selling point among all potential clients is “Palantir can do what others cannot” – right now that is “sold” in powerpoint briefings but not real.

See Also:

2014 Robert Steele: Appraisal of Analytic Foundations – Email Provided, Feedback Solicited – UPDATED

2012 Robert Steele: The Human Factor & The Human Environment: Concepts & Doctrine? Implications for Human & Open Source Intelligence 2.0

1976-2013: Analytic, Methodological, & Technical Models 2.1

Robert Steele & Anonymous: Most Analysis Software Sucks — And Story of How Steele Correctly Called BSA Not Being Signed in Afghanistan

Worth a Look: 1989 All-Source Fusion Analytic Workstation–The Four Requirements Documents