Symbolic AI Debate

The cluster discusses the resurgence and merits of symbolic AI (GOFAI) versus modern deep learning, debates on combining them into neurosymbolic approaches, and criticisms from figures like Gary Marcus and Chomsky.

➡️ Stable 0.7x AI & Machine Learning
3,013
Comments
20
Years Active
5
Top Authors
#9377
Topic ID

Activity Over Time

2007
11
2008
11
2009
28
2010
34
2011
24
2012
44
2013
53
2014
63
2015
112
2016
191
2017
197
2018
246
2019
247
2020
184
2021
208
2022
245
2023
358
2024
351
2025
387
2026
19

Keywords

IMHO LLM SAT AIXI GPU STRIPS AI HN GOFAI lesswrong.com symbolic ai deep learning learning deep machine learning reasoning neural neural networks marcus

Sample Comments

chillacy Dec 3, 2018 View on HN

How might symbolic AI become fashionable?

Workaccount2 Aug 22, 2025 View on HN

My favorite "paper" on AI pretty accurately describes this line of thinkinghttps://ai.vixra.org/pdf/2506.0065v1.pdf

cs702 Jan 6, 2023 View on HN

Marcus has been on HN many times before, always criticizing other people's work as "machines that manipulate data but aren't really intelligent, because they have no understanding of the world:" https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que... . For a good summary of th

colordrops Mar 29, 2021 View on HN

Why is there a pendulum? Aren't both necessary and two sides of the same coin? I know little of modern AI, but I've seen work where both low level raw power for NNs is combined with a symbolic organization of many NNs.

msla Oct 1, 2020 View on HN

Did he say that about GOFAI or Deep Learning?

musha68k Dec 17, 2022 View on HN

Chomsky and Gary Marcus predict that the current approach will not be sufficient; is and will increasingly be detrimental to society because of that “almost right at best” reason. Chomsky only has hopes in a combined approach that integrates “old AI” with the engineering / data / GPU driven one. See also discussion here for anyone interested: https://news.ycombinator.com/item?id=3385754

DeathArrow Feb 13, 2024 View on HN

What if we mix GOFAI with machine learning? Wouldn't the result surpass the capacity of either?

nxobject Feb 8, 2024 View on HN

I think they meant rule engines/symbolic AI loses to LLMs/machine learning again.

mindgam3 Dec 9, 2018 View on HN

The author, Gary Marcus, makes a cogent argument for why we need both deep learning and symbol manipulation to build "true" AI. what I found convincing was the argument that deep learning models currently can't even generalize to novel instances outside the training space in the field of object recognition, which is their strong suit. Why would any rational person believe that they have the capability to generalize to novel instances for other types of problems outside of p

jedharris Dec 16, 2020 View on HN

This paper drives a stake through the heart of claims that explicit symbolic reasoning is needed for intelligent systems. It builds extra-symbolic (deep learning) systems that substantially outperform symbolic reasoning on tests specifically designed to require symbolic reasoning.Since today's programming is 1000% explicitly symbolic, the potential shift to better methods that are extra-symbolic will be pretty disruptive!