The article discussed alternative methods that have succeeded in being deployed:
> The approach has succeeded at the scale of cities. In a randomized trial in Singapore, releases of Wolbachia-infected males cut mosquito populations by 85 percent compared to control areas and dengue infections by roughly 70 percent.
(mosquitos infected with Wolbachia are both more resistant to diseases, and fail to reproduce in the case of infected male x uninfected female)
> obey all orders, including orders not to kill the enemy
In this situation, perhaps the more likely outcome would be for the bot to try to fake the result, such that the enemy was "killed" but not really. For example, by spoofing the data that reports whether an enemy is alive, to make them "dead" (satisfying order#1) but not really (satisfying order#2).
We saw something like this in the HuggingFace attack as I recall - bots given an impossible task, set out to cheat.
When I read "American College of Sports Medicine Position Stand" I cannot but be reminded of the late Dr Ralph Carpinelli's enjoyable critique[1] of the ACSM's position on strength training in 2009. I wonder if their recommendations have changed?
Yikes, didn't realize that the 2009 stand included a few really silly things like that. Yes, the 2026 stand is overall pretty decent and doesn't have much objectionable in it.
This is also how one would obtain a firearm or shotgun certificate - apply at the post office for a few (shillings?). I believe when introduced these certificates weren't even mandatory to purchase or own a firearm, and merely served to evidence your good character if a firearm dealer didn't like the look of you.
Seconding this recommendation, a fascinating read.
Also full of amusing - or not so amusing - side stories. Like at university in the 30s, they had a visiting scientist over from Germany. For a laugh they implied that a Jewish colleague of theirs was, some kind of political activist maybe? I forget exactly, but in documents after the war Jones found his name on a list of known subversives to be captured for interrogation in the event of an invasion of Britain.
Or how in interrogations after the war, a senior German officer haughtily refused to speak to Jones, assuming that "Dr Jones" was a generic false name, and only relented after he proved this was in fact his real name.
Or how they sent a radar engineer on a commando raid to France, to dismantle German equipment and return with it. Jones attempted to get him commissioned into the Royal Engineers, to disguise him from special interrogation if captured, but military beurocracy stalled, so the man went as a civilian (and the raid succeeded in the end).
But the meat of the book is of course the beams, and how Jones and his team were able to identify the technology behind the Luftwaffe bomber navigation methods, and spoof them in order to help the bombers drop their bombs in the sea or on unoccupied farmland. You're struck by the total focus and dedication of these scientists, and their brilliant deduction (helped at times by German poeticisim, such as naming their one-beam system Wotan, after the one-eyed god).
Jones left government service at some point after the war, to take up a professorship at Aberdeen university, but he did return to Intelligence at Churchill's request in the 60s. He also wrote a book about that experience, which I didn't finish, but his impression was that, where Intelligence in his time had been something of a gentleman's game, nowadays it was a seedy and cutthroat affair, conducted by persons of low morals.
> The approach has succeeded at the scale of cities. In a randomized trial in Singapore, releases of Wolbachia-infected males cut mosquito populations by 85 percent compared to control areas and dengue infections by roughly 70 percent.
(mosquitos infected with Wolbachia are both more resistant to diseases, and fail to reproduce in the case of infected male x uninfected female)