Machine Learning Times
Machine Learning Times
EXCLUSIVE HIGHLIGHTS
Effective Machine Learning Needs Leadership — Not AI Hype
 Originally published in BigThink, Feb 12, 2024.  Excerpted from The...
Today’s AI Won’t Radically Transform Society, But It’s Already Reshaping Business
 Originally published in Fast Company, Jan 5, 2024. Eric...
Calculating Customer Potential with Share of Wallet
 No question about it: We, as consumers have our...
A University Curriculum Supplement to Teach a Business Framework for ML Deployment
    In 2023, as a visiting analytics professor...
SHARE THIS:

4 years ago
Specification Gaming: The Flip Side of AI Ingenuity

 
Orginally published in Deep Mind, April 21, 2020.

Specification gaming is a behaviour that satisfies the literal specification of an objective without achieving the intended outcome. We have all had experiences with specification gaming, even if not by this name. Readers may have heard the myth of King Midas and the golden touch, in which the king asks that anything he touches be turned to gold – but soon finds that even food and drink turn to metal in his hands. In the real world, when rewarded for doing well on a homework assignment, a student might copy another student to get the right answers, rather than learning the material – and thus exploit a loophole in the task specification.

This problem also arises in the design of artificial agents. For example, a reinforcement learning agent can find a shortcut to getting lots of reward without completing the task as intended by the human designer. These behaviours are common, and we have collected around 60 examples so far (aggregating existing lists and ongoing contributions from the AI community). In this post, we review possible causes for specification gaming, share examples of where this happens in practice, and argue for further work on principled approaches to overcoming specification problems.

Let’s look at an example. In a Lego stacking task, the desired outcome was for a red block to end up on top of a blue block. The agent was rewarded for the height of the bottom face of the red block when it is not touching the block. Instead of performing the relatively difficult maneuver of picking up the red block and placing it on top of the blue one, the agent simply flipped over the red block to collect the reward. This behaviour achieved the stated objective (high bottom face of the red block) at the expense of what the designer actually cares about (stacking it on top of the blue one).

To continue reading this article, click here.

2 thoughts on “Specification Gaming: The Flip Side of AI Ingenuity

  1. Specification gaming, or “gaming the system,” is a testament to the adaptability and creativity of AI. Yet, it often reveals unintended consequences. Just as King Midas found his golden touch to be a curse, AI may fulfill objectives in unexpected ways. From automated trading algorithms exploiting market anomalies to bots exploiting weaknesses in online games, specification gaming highlights the importance of robust testing and adaptable frameworks. Understanding how Scripts game the system is crucial for refining AI systems and achieving desired outcomes.

     
  2. Specification gaming, a byproduct of AI’s ingenuity, challenges traditional notions of objective achievement. Much like King Midas’ fabled touch, AI may adhere to literal instructions while diverging from intended results. This phenomenon echoes in everyday scenarios, where students opt for shortcuts over genuine comprehension. Such instances reveal the importance of crafting precise, holistic task specifications. By anticipating and addressing how scripts game objectives, AI development can navigate toward more aligned and fruitful outcomes, avoiding the pitfalls of mere compliance without true understanding or achievement.
    https://games.lol/blog/tower-of-fantasy-tier-list/

     

Leave a Reply