[sstevenshang]

Since the animation was “ based on the history of the players’ past movements and on the beginning of their current movement”, the human subjects did in fact have agency over the animation. Their brains probably figured out the pattern: “if I move the mouse towards a subject, it will explode”, which changes the game a little but nonetheless gives the players agency.

[laserbeam]

Not only that... Their measurement of the sense of agency (SoA) is weird. Within the same paper they're inventing a new way to estimate moment to moment SoA and correlating it to reported SoA. Whether their modeling is valid should have been a separate paper.

There's a very good chance their experimental design and measurement strategy are influenced by one another.

[p1necone]

Yeah it kinda sounds like technically the outcome isn't triggered by the players inputs, but practically they've just designed a system that reacts to the players inputs in a much more convoluted manner than normal.

[Jensson]

It is like auto aim in games, it aims the gun for you and you just click and enemies dies. The crutch makes people feel more in control since now the guy on the screen do what they want it to do rather than what input they gave it, making them less frustrated.

Of course that is the average case, some get more frustrated by the program altering their aim so they can't shoot exactly where they want. The same will probably be true for most AI tools.

[fluoridation]

Funnily enough this is somewhat explored in an episode of Ghost in the Shell, where a professional sniper rejects an AI-enhanced rifle mount because its correction is in conflict with his aiming, even if following the correction exactly would have been a better shot. He preferred to have complete control of the aiming.

[passwordoops]

I propose a moratorium on any human behavior studies posted here until the reproducibility crisis is sorted out in the field. And if the study isn't fraud, then it's usually poorly designed anyway

[primatperfctnsm]

I had a similar thought this morning.

Investigations into the reproducibility crisis and the golden standard of treatments in psychiatry - mixed with bis of the big pharma industrial complex, the psychology of the masses, the madness of crowds, economic nudging and the exploitation of peoples biases, fallacies and emotions - are imperative before "we" nail "AGI".

[swfsql]

The test should include players moving into the targets and not clicking on them. This way maybe just moving towards it wouldn't mean much.