Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

This was interesting.

In the microcosm of this experiment, the choice of the free candy bar is irrational. The kids are actually choosing between two outcomes: outcome A which is three Hershey's kisses and small Snickers bar, and outcome B which is two Hershey's kisses and a large Snickers bar. Outcome B is definitely better.

Interestingly, in the grand scheme of things this isn't so illogical. If, for example, this was more like a storefront, where the kids could come and at any time get either free small Snickers bars or large Snickers bars for one kiss, then the former strategy is actually better.

Take this another step further and start to consider the cost of decision making itself. Think of the mental cost of constantly processing every outcome in search of maximum benefit and minimum cost. In the long run, applying the heuristic of FREE = GOOD is probably the best strategy from a cost/benefit perspective, since in most cases it does actually work.

This is a long way of saying that I think opting for free is not always rational in the immediate context but probably makes sense as a sort of general strategy for individuals in life.

Maybe I'm just restating the obvious but I thought I would throw this out there.




  Think of the mental cost of constantly processing every outcome in search 
  of maximum benefit and minimum cost.
That's an interesting observation... It would be extremely exciting if we could somehow process real-world information and create programs that calculated and displayed this kind of information to us. Humans seem to be lousy at this, and hence there is great potential for making machines that allow us to make more efficient decisions.

It is the same idea as applying fuzzy/probabilistic logic and rules of interference to the real world; the main technological difficulty isn't the logic per se, but translating information about the real world to a format we can apply the rules of logic to. I wonder what scientific insights we need to make before it becomes possible to do these things...it's got to be possible, right?


Says the human populace: "You can bury my cold dead body before I let some computer program make my decisions for me!"

Say businesses: "Please, we'll do anything to increase profits and ROI!"




Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: