When Secrecy News gained unauthorized access to a restricted U.S. Army manual on visual identification of U.S. and foreign aircraft, we supposed that it was just one more case of unnecessary and inappropriate secrecy.
But it turns out to be something worse than that, since the document (pdf) contains a surprising number of technical errors.
The dimensions given in the Army manual for the Predator unmanned aerial vehicle are wrong, the Entropic Memes blog astutely noted. And the entry for the B-52, among others, is likewise incorrect.
“Please,” Entropic Memes exclaimed. “If they can’t get the details of one of their own systems correct, how much faith can you have that they got the details of anyone else’s systems right?”
In this case, the secrecy of the Army manual was not just an arbitrary barrier to public access. It also “protected” numerous errors that may make the document worse than useless.
Conversely, exposing the document to public scrutiny may now make it possible to correct its errors so as to fulfill its intended purpose.
Since it was posted on the Federation of American Scientists website 48 hours ago, the Visual Aircraft Recognition manual has been downloaded over seventy thousand times, an exceptionally high rate of access.
Update: “This is not a subject I’ve so far spent a lot of time on, but the entry for every aircraft I’ve looked up in the manual thus far contains errors,” adds Entropic Memes in a new post.
These ideas aim to advance the detailed policy solutions needed to foster public trust and implement fairness in the adoption of AI across diverse domains, from healthcare and government benefits to rural access, education, and worker protections.
The evidence is clear: algorithmic pay-setting is established in app-based work, and payroll/timekeeping failures show how software can produce systemic wage harm at scale
While a few states have taken steps to implement decision-making mechanisms for certain AI systems, too many leaders are simply accepting narratives about AI’s purported public benefit at face value – jumping to the “how” of AI implementation before thoroughly vetting potential systems and deciding whether they are appropriate to use at all.
When properly structured — with specific numeric targets, secured financial obligations, independent monitoring, and meaningful enforcement — CBAs transform data center deals into durable community partnerships.