Bringing powerful people together to deal arms
Content Warning: Surprisingly painful-to-write topic. Awful attempt at journalism.
tldraw is a performant canvas drawing web application. I use it often.
Our story starts with me looking at tldraw’s devlog and finding out it is funded by a venture capital firm called Lux Capital.
Looking at the Github issue tracker of tldraw, the team is acting increasingly like a business, and less like free software.
I got curious as to what Lux Capital is, so I looked into Lux Capital’s website. The navigation header features a weirdly named link: Riskgaming. I clicked on the link. I felt a new emotion on the border of disgust.
There comes the title of this article. I thought I should annoy you with what I’ve seen, and see what emotions you will have when you learn about the same thing yourself.
In short, Riskgaming (a board game / roleplaying game) is designed to captured the behavior of politicians and military officials (see the game rules). It also brings those people together with arms manufacturers to the same table. On the game description page, you can download the game rules in PDF version and watch a YouTube video. The beginning of the video contains a typo right after the intro animation sequence. Then, it’s someone who talks in a way that I find bizarre even for a US patriot. The beginning of the game rulebook contains an introduction of the writer — a blatant appeal to authority. Reading through both reading material, I was amazed (and horrified) by how well they filter their audience. At the same time, I wondered: is this to open up a new market, or is this to pave the way for existing companies?
I found the answer on their portfolio page. Anduril. The company’s products range from border surveilance to flying bricks. Of course, they have flying missles and guns too; for more products, see the “Products” section at the footer of their website.
My thoughts on this
Having thought about this for a week, it turns out that I don’t have much thoughts on the matter.
About the invitation. When I first read about how they invite people to join a game session, my first reaction was, “wait you can do that?”. This is my first time seeing sales tactics being used in reconciliation, and I now think of fear of missing out being used in transitional justice to “encourage” participation. To think that I had learned something from sales…
About the game. The game features 6 actors who can control the game flow: the city mayor, the state defense politician, the tech company CEO, the press, the work union president, the naval admiral. Since ordinary people aren’t included in the simulation, I doubt the game’s advertised historical accuracy.
Overall, I have to applaud Lux for designing this plot to bring people from different background together.
Now, onto the weapons company.
I previously thought that developing automated weapons was outright banned by international treaties. As it turns out, it is not banned as of time of writing.
Here is what their system is currently (as advertised) capable of: sensor systems pick on ground or air targets. With a press of a button on a web UI, drones launch and fire missles and an assortment of hurtful objects at the targets. The system itself can already recognize vehicles and people (from the advertisement footage they provide, I think it’s using YOLO. So, the system is fully automated and delegates the responsibility of killing to the one pressing the button. Insofar as the current system (advertised) does not differentiate between people in different uniforms, such a feature can be implemented in a day by reusing the model’s understanding of pictures.
What kind of human behavior is considered acceptable? This is the human version of the AI safety question. The answer is subjective, and my personal opinion is complex enough that I can’t stuff it inside this article without bloating the article. In any case, developing automated flying killing machines is not something I see myself doing. My confidence is in shambles.
The silverlining here? Bodged-together military solutions cost way less than for-profit ones. Ideally though, I would not let my hypothesis be tested.
I think I’ll read some EFF to clear my mind. Then, some history.