The Limitations of Deep Learning

Re:The Limitations of Deep Learning by jonobelotti on Monday July 17, 2017 @07:07PM (#14793044)

Brooks' 'Intelligence Without Representation' (http://people.csail.mit.edu/brooks/papers/representation.pdf) starts with a pretty strong argument imo against the story of 'stick-together' AGI you're describing.

[ Parent | on HN ]

Re:The Limitations of Deep Learning by randcraw on Tuesday July 18, 2017 @04:56PM (#14800004)

I think Brooks' Cog initiative was an attempt to 'ground' the robot's perceptions of the physical world into forming a rich scalable representation model. But it looks like that line of investigation ended ~2003 with Brooks' retirement. Too bad, given the seeming suitability of using deep nets to implement it.

http://www.ai.mit.edu/projects/humanoid-robotics-group/cog/o...

[ Parent | on HN ]

Re:The Limitations of Deep Learning by webmaven on Monday July 17, 2017 @08:17PM (#14793397)

Thanks for the link to this interesting paper.

I think we're seeing some recapitulation of those arguments WRT 'ensembles of DL models' approaches.

[ Parent | on HN ]

Re:The Limitations of Deep Learning by thundergolfer on Monday July 17, 2017 @10:03PM (#14793826)

I agree. Google has come out with some papers that are, to put it harshly, basic gluing together of DL models followed by loads of training on their compute resources.

[ Parent | on HN ]

Re:The Limitations of Deep Learning by webmaven on Monday July 17, 2017 @11:27PM (#14794084)

Not just Google. The FractalNet paper comes to mind.

[ Parent | on HN ]