Название: The Challenge to AI: Consciousness and Ecological General Intelligence
Автор: Stephen E. Robbins
Издательство: Mercury Learning and Information
Серия: MLI Generative AI Series
Год: 2024
Страниц: 365
Язык: английский
Формат: pdf, epub (true)
Размер: 10.1 MB

This book invites readers to explore the intricate world where Artificial Intelligence (AI), consciousness, and human cognition intersect. This groundbreaking book considers the profound differences between man and machine, challenging existing notions in AI and cognitive science. It argues that the key to understanding intelligence lies not in software, but in the hardware of our brain – a complex biochemical system far removed from current AI architectures. Through a deep examination of time, perception, language, and the nature of thought, the book presents a compelling case for the indispensability of biology and consciousness in cognition. To achieve this, to engineer this, will indeed be a challenge for AI.

Artificial general intelligence (AGI), is the gleam in the eye of the AI community. This gleam was there “at the beginning,” the beginning being in the 1950s. However, in the author’s experience (circa 1972) the computer metaphor of mind swept into a then rather sleepy cognitive psychology just barely emerging from the grips of Skinner’s rats, a field that, given this injection of the computational framework, ultimately became cognitive science. This beginning was with the publication of Newell and Simon’s Human Problem Solving, a work dedicated to showing that a fundamental algorithm, means–ends analysis, could be applied to tasks like theorem proving, cryptarithmetic problems, even chess. It was dedicated to a human-simulation quest, that is, to showing that human solution processes for the problem closely emulated the steps taken by Newell and Simon’s computer algorithm (which they termed the General Problem Solver (GPS)). This algorithmic scheme, it turned out, would eventually evolve, becoming the industry of “expert systems.”

The generative pretrained transformers (GPTs) of generative AI, we shall see, choose a different route, in one way doing an end-around for the entire problem, in another way, simply following the same approach, but essentially relying on a form of a premade solution to the problem.

Features:
Presents an alternative to the conventional computer metaphor, offering a new framework for understanding the mind
Introduces a fresh perspective on the role of time in AI and philosophy, highlighting its critical importance
Explores the biochemical basis of cognition, challenging the traditional focus on symbol manipulation and neural networks




ОТСУТСТВУЕТ ССЫЛКА/ НЕ РАБОЧАЯ ССЫЛКА ЕСТЬ РЕШЕНИЕ, ПИШИМ СЮДА!