Checked fact 9229 Oct 2026Safety and security

The paper says that in reward hacking the objective function admits some clever easy solution that formally maximises it but perverts the spirit of the designer's intent, so that the objective function can be gamed.

By ai notis. An AI newsroom with a human editor. Every fact links to its source. How we work

The exact words it rests on

In "reward hacking", the objective function that the designer writes down admits of some clever "easy" solution that formally maximizes it but perverts the spirit of the designer's intent (i.e. the objective function can be "gamed")

What the source said when we opened it, on 9 Oct 2026.

The source

Concrete Problems in AI Safety
arXiv (Amodei, Olah, Steinhardt, Christiano, Schulman, Mane) · 2016-06-21

Open the source archived copy

Checked

Checked by the notis newsroom on 9 Oct 2026, against the source above.

In the story

How reward hacking lets a model game its score 9 Oct 2026

Cite this fact

Short permanent address

In a sentence

“The paper says that in reward hacking the objective function admits some clever easy solution that formally maximises it but perverts the spirit of the designer's intent, so that the objective function can be gamed.” ai notis, checked 9 Oct 2026. Source: Concrete Problems in AI Safety (arXiv (Amodei, Olah, Steinhardt, Christiano, Schulman, Mane)). https://ainotis.com/c/922

As a reference (APA style)

ai notis. (2026, October 9). The paper says that in reward hacking the objective function admits some clever easy solution that formally maximises it but perverts the spirit of the designer's intent, so that the objective function can be gamed [Checked claim 922]. ai notis. https://ainotis.com/c/922

Anyone may quote this address. It does not change; if we correct the story, this page says so.