Comment by parsimo2010
3 months ago
Agreed- they write as if being overwhelmed 3% of the time is a victory. A good system would have people feeling overwhelmed 0% of the time.
3 months ago
Agreed- they write as if being overwhelmed 3% of the time is a victory. A good system would have people feeling overwhelmed 0% of the time.
>A good system would have people feeling overwhelmed 0% of the time.
There are benefits to being pushed past your limits from time to time. Also, there's just no such thing as 0. When you're designing limits you don't say "this never happens", you're saying "this event happens less than this rate for this cohort".
I'd agree that it is worth pushing your limits during training, but the best-case scenario during actual conflict is to be as close to 0% overwhelmed as you can be.
How does that follow?
That would mean leaving some performance on the table the rest of the time.
It doesn't seem clear at all whether one outweighs the other.
6 replies →
> they write as if being overwhelmed 3% of the time is a victory
We’re talking about a soldier. Commanding a company’s worth of firepower single-handedly from relative safety. 3% would be an exceptional improvement over the status quo.
The real question is what happens in that 3%. If they are still able to control the drones that is very different from they set the drones to kill your own people. (This is DARPA so we can assume killing people is a goal in some form). There is a lot in between too.
This is a common error, if not outright fallacy. The correct amount of <negative event> is rarely zero due to diminishing returns -- it is where the cost curves intersect.
E.g. to decrease 3pct to 0.3pct might require operating only half the drones -- not a good trade.
This is peacetime thinking. If you've got a whole army trying to kill you, you're going to get overwhelmed sometimes.
Compare it to a control group - I feel overwhelmed at least 5% of the time and I’m not even controlling any robots.
Yeah I really don't like that phrasing. Take off and landing is the most dangerous part of flying but only makes up a tiny percentage of the total flight. If that 3% of the time referenced is the most dangerous or most critical 3% of time then it hardly matters how easy the rest of it is.
This is about the army. Depending on the case, it's acceptable that 30% of people die if it serves strategic goals. That's how "a good system" is defined by those who have the power to enact it .