HOUSE_OVERSIGHT_025964.jpg
2.81 MB
Extraction Summary
2
People
4
Organizations
0
Locations
1
Events
2
Relationships
4
Quotes
Document Information
Type:
Email/correspondence (part of house oversight committee production)
File Size:
2.81 MB
Summary
This document is a page from a correspondence (likely an email found in the House Oversight production) discussing the intersection of neuroscience, artificial intelligence, and linguistics. It references Google's DeepMind acquisition, specific machine learning milestones from 2015, and Noam Chomsky's linguistic theories, while also forwarding a controversial hypothesis linking race, motor development, and IQ. The document includes Bates stamp HOUSE_OVERSIGHT_025964.
People (2)
| Name | Role | Context |
|---|---|---|
| Noam | Linguist/Academic (Implied Noam Chomsky) |
His criticism of machine translation is discussed regarding Latent Semantic Analysis models.
|
| Unknown Sender/Recipient | Correspondents |
Individuals discussing AI, neuroscience, and controversial theories regarding race and IQ.
|
Organizations (4)
| Name | Type | Context |
|---|---|---|
|
Mentioned regarding their photo app, image recognition, and acquisition of DeepMind.
|
||
| DeepMind |
Mentioned regarding their Atari game learning program and acquisition by Google for 500M.
|
|
| WSJ (Wall Street Journal) |
Source of a linked article.
|
|
| CreativeAI |
Source of a linked article regarding deep generator networks.
|
Key Quotes (4)
"In humans, it is reflected for instance by the fact that races with faster motor development have lower IQ."Source
HOUSE_OVERSIGHT_025964.jpg
Quote #1
"A machine learning program that can learn how to play an Atari game without any human supervision... now just takes about 130 lines of Python code."Source
HOUSE_OVERSIGHT_025964.jpg
Quote #2
"Noam's criticism of machine translation mostly applies to the Latent Semantic Analysis models that Google and others have been using for many years."Source
HOUSE_OVERSIGHT_025964.jpg
Quote #3
"Motivation tells the brain what to pay attention to, by giving reward and punishment."Source
HOUSE_OVERSIGHT_025964.jpg
Quote #4
Full Extracted Text
Complete text extracted from the document (3,717 characters)
literally smarter on every level, and because the abilities of the higher levels depend on those of the lower levels, they can perform abstractions that mature gorillas will never learn, no matter how much we try to train them.
The second set of mechanisms is in the motivational system. Motivation tells the brain what to pay attention to, by giving reward and punishment. If a brain does not get much reward for solving puzzles, the individual will find mathematics very boring and won't learn much of it. If a brain gets lots of rewards for discovering other people's intentions, it will learn a lot of social cognition.
Language might be the result of three things that are different in humans:
- extended training periods per layer (after the respective layer is done, it is difficult to learn a new set of phonemes or the first language)
- more layers
- different internal rewards. Perhaps the reward for learning grammatical structure is the same that makes us like music. Our brains may enjoy learning compositional regular structure, and they enjoy making themselves understood, and everything else is something the universal cortical learning figures out on its own.
- extended training periods per layer (after the respective layer is done, it is difficult to learn a new set of phonemes or the first language)
- more layers
- different internal rewards. Perhaps the reward for learning grammatical structure is the same that makes us like music. Our brains may enjoy learning compositional regular structure, and they enjoy making themselves understood, and everything else is something the universal cortical learning figures out on its own.
This is a hypothesis that is shared by a growing number of people these days. In humans, it is reflected for instance by the fact that races with faster motor development have lower IQ. (In individuals of the same group, slower development often indicates defects, of course.)
Another support comes from machine learning: we find that the same learning functions can learn visual and auditory pattern recognition, and even end-to-end-learning. Google has built automatic image recognition into their current photo app:
http://blogs.wsj.com/digits/2015/07/01/google-mistakenly-tags-black-people-as-gorillas-showing-limits-of-algorithms/
http://blogs.wsj.com/digits/2015/07/01/google-mistakenly-tags-black-people-as-gorillas-showing-limits-of-algorithms/
The state of the art in research can do better than that: it can begin to "imagine" things. I.e. when the experimenter asks the system to "dream" what a certain object looks like, the system can produce a somewhat compelling image, which indicates that it is indeed learning visual structure. This stuff is something nobody could do a few months ago:
http://www.creativeai.net/posts/Mv4WG6rdzAerZF7ch/synthesizing-preferred-inputs-via-deep-generator-networks
http://www.creativeai.net/posts/Mv4WG6rdzAerZF7ch/synthesizing-preferred-inputs-via-deep-generator-networks
A machine learning program that can learn how to play an Atari game without any human supervision or hand-crafted engineering (the feat that gave DeepMind 500M from Google) now just takes about 130 lines of Python code.
These models do not have interesting motivational systems, and a relatively simple architecture. They currently seem to mimic some of the stuff that goes on in the first few layers of the cortex. They learn object features, visual styles, lighting and rotation in 3d, and simple action policies. Almost everything else is missing. But there is a lot of enthusiasm that the field might be on the right track, and that we can learn motor simulations and intuitive physics soon. (The majority of the people in AI do not work on this, however. They try to improve the performance for the current benchmarks.)
Noam's criticism of machine translation mostly applies to the Latent Semantic Analysis models that Google and others have been using for many years. These models map linguistic symbols to concepts, and relate concepts to each other, but they do not relate the concepts to "proper" mental representations of what objects and processes look like and how they interact. Concepts are probably one of the top layers of the learning hierarchy, i.e. they are acquired *after* we learn to simulate a mental world, not before. Classical linguists ignored the simulation of a mental world entirely.
HOUSE_OVERSIGHT_025964
Discussion 0
No comments yet
Be the first to share your thoughts on this epstein document