Made-Up Minds: A Constructivist Approach to Artificial Intelligence

By G. Drescher. 1991

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Active learning of inverse models with intrinsically motivated goal exploration in robots (2013, positive): “…(α, ∆ α, ∆ x ) in its memory and creates a database Data which contains elements (α i , ∆ α i , ∆ y i ) representing the discovered change ∆ y i corresponding to a given ∆ α i in the configuration α i (this learning entity can be called a schema according to the terminology of Drescher [111]).”
Microsoft (2019, unknown): “The computational models of developmental learning proposed by Gary Drescher (1991) seem to me to be a good place to start for the tower-building aspect of the problem.”
When Waiting is not an Option : Learning Options with a Deliberation Cost (2017, background): “Temporal abstraction has a rich history in AI (Minsky 1961; Fikes et al. 1972; Kuipers 1979; Korf 1983; Iba 1989; Drescher 1991; Dayan and Hinton 1992; Kaelbling 1993; Thrun and Schwartz 1995; Parr and Russell 1998; Dietterich 1998) and has been presented as a useful mechanism for a variety of…”
Learning to Use Selective Attention and Short-Term Memory in Sequential Tasks (1996, background): “U-Tree is also related to Drescher's Schema Mechanism [Drescher, 1991], in that both build complex representations from simpler primitives using covariance statistics.”
Horde: a scalable real-time architecture for learning knowledge from unsupervised sensorimotor interaction (2011, background): “Drescher (1991) considered a simulated robot baby learning conditional probability tables for boolean events.”
CHILD: A First Step Towards Continual Learning (1997, methodology): “The units created by the TTH network resemble the “synthetic items” of Drescher (1991), created to determine the causes of an event (the cause is determined through training, after which the item represents that cause).”
From Skills to Symbols: Learning Symbolic Representations for Abstract High-Level Planning (2018, background): “Several researchers have learned symbolic models of the preconditions and effects of preexisting controllers for later use in planning (Drescher, 1991; Schmill, Oates, & Cohen, 2000; Pasula, Zettlemoyer, & Kaelbling, 2007; Amir & Chang, 2008; Kruger, Geib, Piater, Petrick, Steedman, Wörgötter, Ude, ”
A Foundational Architecture for Artificial General Intelligence (2007, methodology): “Aside from learning, which we’ll discuss later, the major recipient of the global broadcast is Procedural Memory (PM), which we implement as a scheme net modeled after the schema mechanism [46].”
Map Learning with Uninterpreted Sensors and Effectors (1995, methodology): “14This approach is analogous to Drescher's marginal attribution [7].”
Options of Interest: Temporal Abstraction with Interest Functions (2020, background): “Temporal abstraction methods aim to solve the ﬁrst problem, and a lot of recent literature has also be found in later work on macro-operators (Korf 1983) or the Schema model (Drescher 1991).”
DREAM Architecture: a Developmental Approach to Open-Ended Learning in Robotics (2020, methodology): “Pioneering works on this topic drew inspiration from Piaget’s developmental Psychology work and applied it to simplified environments with predefined representations [133, 134].”
Controlling Recurrent Neural Networks by Conceptors (2014, background): “Such findings led to a diversity of enriched models of human concepts and their organization (my favourites: [62, 22, 64]), and many fundamental questions remain controversial.”