Archive

Archive for August, 2016

No QuEST Meeting this week, 26 Aug

August 25, 2016 Leave a comment

Due to Capt Amerika’s travel schedule, there will be no meeting this week.  We will plan on resuming our schedule next week.  Hope everyone has a safe and fun weekend.

Advertisements
Categories: Uncategorized

Weekly QuEST Discussion Topics and News, 19 Aug

August 19, 2016 Leave a comment

We want to start this week with an example of ‘reward hacking’ in nature.  While watching a YouTube Ted talk at the link (provided by sean a quest colleague) below from Don Hoffmann

https://www.ted.com/playlists/384/how_your_brain_constructs_real

Below are my notes on the video:

What is the relationship between your conscious experience and reality – just like Alladin appearing from the bottle the magic appearance of consciousness is just as mysterious – brain activity are correlated with conscious experience but not know why – we still don’t know – we lack the necessary constructs to explain –

Made false assumption – that assumption is do we see reality as it is? Open eyes and have experience a red tomato a meter away – I believe there is a red tomato a meter away – I close my eyes I think there still is a red tomato a meter away – so logic seems to dictate that our conscious experience is reality BUT

Used to think world was flat – senses seem to imply – we found false – similar thought earth  was center of universe – we were wrong – taste, odors, colors are in the mind of the agents not in  the world –

Neuro sci say a third of cortex used for vision – think vision is a camera – an objective reality – there is part that is like a camera – the eye – 130million photo receptors in the eye – but the billions of neurons and trillions of synapses involved in vision – we construct what we see – we don’t construct the whole world only what we need in the moment – example is visual illusions – we construct a 3 d cube Necker cube – we create motion when flipping colors of dots on a page

Neuro scientist think we reconstruct reality – accurate reconstruction of real red tomato that really exists – why would we reconstruct reality – those that do right better likely to have offspring with more reality … – vision useful because it is so accurate is in common text –

Accurate perceptions are better  – implications –

Now the example that I want to talk about:

This is the Jewell Beetle – it’s purpose in life is to find other Jewell beetles and mate – but then there are human men – their main purpose in life is to drink beer – and throw the bottles into the outback – bottles in outback caused beetles to attempt to mate – had to change bottles – evolution had told beetle that big bumpy glossy things have sex with – the beetle was going extinct– the male couldn’t make this mistake – even moose make the mistake – does natural selection really favor reality as it is – there is reward hacking —- ***** beetle is great example – they were going extinct –

Jewell beetle

 

Jewell beetle attempting to mate with a beer bottle

The fitness function being used did for millennia accomplish having the beetles reproduce – but then something changed – beer bottles – all autonomous systems look to solve an objective function as efficiently as possible – BUT –

Steak – fitness of animal – for a well fed lion – not the same thing as reality as it is – fitness is the key part of the evolution equation –

Modeling and sim – some see part of reality – who wins – perception of reality goes extinct – organizations that tuned to fitness not reality – perception does NOT favor accurate perception of reality –

How can not seeing the world accurately be better off – we don’t see reality as it is – we are shaped by tricks / hacks / reward hacking that keeps us alive

Metaphor – desktop – of computer – icon of ted talk – is the text file blue rectangular an in corner of screen – the purpose of the interface is not to show reality of the computer – it is there to hide reality – it is to be useful – evolution has given us an interface that hides reality and consciousness is icons on the desktop – train coming down the track – so step in front of it Prof Hoffman – this is the structural coherence tenet

Evolution has shaped us with perceptual signals that keep us alive – they keep us safe – doesn’t mean we should take them literally – metal of train mostly space – physics has taught – know reality of computer – see pixels of computer with magnifying glass – we all see the train – so none of us construct the train – the necker cube see we construct – we all see the cube cause we all construct the cube – all physical objects we construct perception – perception is not a window on reality  – reality is like a 3d desktop that hides complexity of reality – we believe space time and objects are the nature of reality as it is

There is something that exists where we look – but it isn’t what we are perceiving –

We have advantage over the jewell beetle –

Donald Hoffman –

Interacting with reality is not reality – lions perceived is not what a lion is –

Reality whatever it is – brains and neurons – species specific sets of symbols –

Consciousness – perhaps reality is a machine – vast interactive agents

Give up false assumption on perception of reality –

Given this set up let’s return to our investigation of hypnosis – since the ideas being acted upon by the person that is hypnotized are not reality – how does this occur? –

  • seems to me this is the ideal case of hypnosis, you have an agent that you have become almost hyper-aligned with and thus when they introduce context into your representation you take it for granted, no filtering it out regardless of how badly it might fit with your current representation.  still looks to me like this is an explanation for hypnosis impacting the sys2 representation?
  • i would argue that a basic agent with no two system representation CANNOT be hypnotized. Way to demonstrate animals are consciousness!  they cannot experience a disconnect between something they know to be real and something that they think SHOULD be real.

for the purposes of this discussion assume that in the hypnotic state(and that such a state really exists) you are using qualia not unlike sleepwalking, the hypnotist can insert into your dream state suggestions that manipulate your ‘dream’ – so there is still a sys1 set of calculations that go on below the level of the sys2 qualia and the qualia of sys2 are those aspects of the hypnotic state that in that state you use qualia for you are ‘conscious’ of ***

 

specifically what I am interested in for this discussion is optimizing performance under stress for an agent:

 

Psychophysiology, 49 (2012), 1417–1425. Wiley Periodicals, Inc. Printed in the USA.

Copyright © 2012 Society for Psychophysiological Research

DOI: 10.1111/j.1469-8986.2012.01449.x

 

The effect of challenge and threat states on performance:

An examination of potential mechanisms

 

LEE J. MOORE, SAMUEL J. VINE, MARK R. WILSON, and PAUL FREEMAN

College of Life and Environmental Sciences, University of Exeter, Exeter, UK

 

Abstract

Challenge and threat states predict future performance; however, no research has examined their immediate effect on motor task performance. The present study examined the effect of challenge and threat states on golf putting performance

and several possible mechanisms. One hundred twenty-seven participants were assigned to a challenge or threat group and performed six putts during which emotions, gaze, putting kinematics, muscle activity, and performance were

recorded. Challenge and threat states were successively manipulated via task instructions. The challenge group performed more accurately, reported more favorable emotions, and displayed more effective gaze, putting kinematics, and

muscle activity than the threat group. Multiple putting kinematic variables mediated the relationship between group and performance, suggesting that challenge and threat states impact performance at a predominately kinematic level.

 

In my mind I want to tie this to the issues emotional intelligence – and back to our discussion on what is intelligence – and how do qualia contribute to intelligence – and given that position how can we give the skills to our agents to facilitate acceptable responses by improving their emotional intelligence

 

  • Emotional intelligence (EI) or emotional quotient (EQ) is the capacity of individuals to recognize their own, and other people’s emotions, to discriminate between different feelings and label them appropriately, and to use emotional information to guide thinking and behavior
  • I’ve struggled with the ‘label’ part of the definition – I’ve always felt that it is the experience not the word – see our computing with words discussion – the intelligence is a function of the discretization of the experience space not being able to articulate labels – only that you can experience the distinct range

So the question is how do we provide an improvement of the qualia discretization over the stimuli space to improve the intelligence of a human or computer agent – approve the agents ability to acceptably respond to a wider range of stimuli

 

news summary (24)

Categories: Uncategorized

Weekly QuEST Discussion Topics and News, 12 Aug

August 11, 2016 Leave a comment

QuEST 12 Aug 2016

We have several topics that have captured my attention this week.  The first is to recast the use of AI to solve really Big Problems – making the world a better place.  How to adapt the approaches to complex and abstract problems like the ones we’ve discussed with our colleague on inner city violence / drugs.

http://www.technologyreview.com/news/545416/could-ai-solve-the-worlds-biggest-problems/

Could AI Solve the World’s Biggest Problems?

Advances in machine-learning techniques have opened up a wealth of promising opportunities for AI applications, but some tech executives are thinking about ways it can make the world a better place.

Demis Hassabis, CEO of Google Deepmind, a division within Google doing groundbreaking work in machine learning, and which aims to bring about an “artificial general intelligence” (see “Google’s Intelligence Designer”), said the goal of this effort was to harness AI for grand challenges. “If we can solve intelligence in a general enough way, then we can apply it to all sorts of things to make the world a better place,” he said.

And the chief technology officer of Facebook, Mike Schroepfer, expressed similar hope. “The power of AI technology is it can solve problems that scale to the whole planet,” he said.

https://www.technologyreview.com/s/601139/how-google-deepmind-plans-to-solve-intelligence/

How Google DeepMind Plans to Solve Intelligence

Mastering Go is just the beginning for Google DeepMind, which hopes to create human-like artificial intelligence.

Sponsored by

Padded walls, gloomy lighting, and a ceiling with floral wallpaper. It doesn’t look like a place to make groundbreaking discoveries that change the trajectory of society. But in these simulated, claustrophobic corridors, Demis Hassabis thinks he can lay the foundations for software that’s smart enough to solve humanity’s biggest problems.

Another example problem that fits into this pocket in my mind is how to take on the issue of ISIS online either controlling or inspiring terrorist activity.

from http://science.sciencemag.org/ on June 20, 2016

 

New online ecology of adversarial

aggregates: ISIS and beyond

  1. F. Johnson,1 M. Zheng,1 Y. Vorobyeva,2 A. Gabriel,1 H. Qi,1 N. Velasquez,2
  2. Manrique,1 D. Johnson,3 E. Restrepo,4 C. Song,1 S. Wuchty5,6*

 

Support for an extremist entity such as Islamic State (ISIS) somehow manages to survive globally online despite considerable external pressure and may ultimately inspire acts by individuals having no history of extremism, membership in a terrorist faction, or direct links

to leadership. Examining longitudinal records of online activity, we uncovered an ecology evolving on a daily time scale that drives online support, and we provide a mathematical theory that describes it. The ecology features self-organized aggregates (ad hoc groups formed via linkage to a Facebook page or analog) that proliferate preceding the onset of recent real-world campaigns and adopt novel adaptive mechanisms to enhance their survival. One of the predictions is that development of large, potentially potent pro-ISIS aggregates can be thwarted by targeting smaller ones.

 

 

The next topic is associated with group intelligence.

 

http://www.theatlantic.com/business/archive/2015/01/the-secret-to-smart-groups-isnt-smart-people/384625/

The Secret to Smart Groups: It’s women

A fleet of MIT studies finds that women are much better at knowing what their colleagues are really thinking. It’s another reason to expect the gender wage gap to eventually flip.

The concept of “general intelligence”—the idea that people who are good at one mental task tend to be good at many others—was considered radical in 1904, when Charles Spearman proposed the theory of a “g factor.” Today, however, it is among the most replicated findings in psychology. But whereas in 1904 the U.S. economy was a network of farms, mills, and artisans, today’s economy is an office-based affair, where the most important g for many companies doesn’t stand for general intelligence, but, rather, groups.

So, what makes groups smart? Is there any such thing as a “smart” group, or are groups just, well, clumps of smart people?

Downloaded from www.sciencemag.org on October 31, 2010

 

Evidence for a Collective Intelligence

Factor in the Performance of

Human Groups

Anita Williams Woolley,1* Christopher F. Chabris,2,3 Alex Pentland,3,4

Nada Hashmi,3,5 Thomas W. Malone3,5

 

Psychologists have repeatedly shown that a single statistical factor—often called “general intelligence”—emerges from the correlations among people’s performance on awide variety of cognitive tasks. But no one has systematically examined whether a similar kind of “collective intelligence” exists for groups of people. In two studies with 699 people, working in groups of two to five, we find converging evidence of a general collective intelligence factor that explains a group’s performance on a wide variety

of tasks. This “c factor” is not strongly correlated with the average or maximum individual intelligence of group members but is correlated with the average social sensitivity of group members, the equality in distribution of conversational turn-taking, and the proportion of females in the group.

 

 

This leads to a related article (at least in my mind) – on the effect of challenge and threat states on performance …

 

Psychophysiology, 49 (2012), 1417–1425. Wiley Periodicals, Inc. Printed in the USA.

Copyright © 2012 Society for Psychophysiological Research

DOI: 10.1111/j.1469-8986.2012.01449.x

 

The effect of challenge and threat states on performance:

An examination of potential mechanisms

 

LEE J. MOORE, SAMUEL J. VINE, MARK R. WILSON, and PAUL FREEMAN

College of Life and Environmental Sciences, University of Exeter, Exeter, UK

 

Abstract

Challenge and threat states predict future performance; however, no research has examined their immediate effect on motor task performance. The present study examined the effect of challenge and threat states on golf putting performance

and several possible mechanisms. One hundred twenty-seven participants were assigned to a challenge or threat group and performed six putts during which emotions, gaze, putting kinematics, muscle activity, and performance were

recorded. Challenge and threat states were successively manipulated via task instructions. The challenge group performed more accurately, reported more favorable emotions, and displayed more effective gaze, putting kinematics, and

muscle activity than the threat group. Multiple putting kinematic variables mediated the relationship between group and performance, suggesting that challenge and threat states impact performance at a predominately kinematic level.

news summary (23)

Categories: Uncategorized

Weekly QuEST Discussion Topics and News, 5 Aug

August 4, 2016 Leave a comment

This week we will have our colleague Andy M provide us an update on the state of the art in brain stimulation – both internal AFRL work and work that commercial companies are pursuing like Halo Technologies:

http://www.dodbuzz.com/2016/07/26/pentagon-taps-tech-firm-for-headset-to-improve-combat-skills/?ESRC=dod-bz.nl

Posted By: Richard Sisk July 26, 2016

The Pentagon’s new innovation unit will be testing a high-tech headset to see whether it can make special operators even better fighters, Defense Secretary Ashton Carter said Tuesday.

Carter said the controversial theory behind the experimental headset is that it “uses non-invasive electrical stimulation” to improve the brain’s learning skills, which could translate into more powerful operators

The “enhanced human operations” headset project, developed by Halo Neuroscience to improve the brain’s ability to adapt, was funded by the Defense Innovation Unit (Experimental), or DIUx, the Carter initiative intended to connect the Pentagon with cutting-edge technology and keep the U.S. ahead of competitor nations.

The company claims its new Halo Sport headset builds on research already conducted by the armed forces.

“The U.S. military accelerated pilot and sniper training by 50 percent with neurotechnology similar to Halo Sport,” its website states. “We’re bringing these gains to athletics.”

Carter spoke in Boston, where he opened the first East Coast DIUx branch and also announced that Jeff Bezos, the Amazon chief executive and owner of The Washington Post, and Neil deGrasse Tyson, the astrophysicist and director of the Hayden Planetarium in Manhattan, would be joining the Defense Innovation Advisory Board.

The board already includes Eric Schmidt, the executive chairman of Alphabet Inc.; Reid Hoffman, co-founder of LinkedIn; and retired Adm. William McRaven, the former commander of U.S. Special Operations Command who organized the raid that killed Osama Bin Laden.

Carter launched DIUx last year at Moffett Airfield in Mountain View, California, to improve outreach to Silicon Valley in what the Pentagon called a “ground-breaking effort to strengthen connections to the American innovation economy and speed technologies into the hands of the warfighter.”

The effort got off to a slow start and Carter reorganized it in May, putting it under his personal control and naming Rajiv Shah, an F-16 pilot in the Air Force National Guard who most recently was the senior director of strategy at computer firewall maker Palo Alto Networks, as head of DIUx.

“I am proud to announce that in its first 75 days, the new DIUx has made tremendous progress in rebuilding bridges to the technology community,” Shah said. “We’ve demonstrated that DoD can be just as nimble and innovative as the companies we want to do business with.”

Carter has asked for $30 million in the defense budget for DIUx, which he said will now be organized into three teams.

A Venture Team will identify emerging commercial technologies and explore their potential impact on the battlefield; aFoundry Team will identify technologies that aren’t yet fully developed for military applications; and an Engagement Team will introduce innovators to military problems and the military to entrepreneurs, Carter said.

The new East Coast office in Cambridge, Massachusetts, will put DIUx in a city that is “home to a tremendous legacy of service — one that will continue in a new way with DIUx,” Carter said. “It’s a testament to the fact that Boston has always been a place where great minds and great ideas come together to help advance the safety and security of our country.”

 

http://www.mrn.org/files/news/Zap_your_brain_into_the_zone__Fast_track_to_pure_focus_-_life_-_06_February_2012_-_New_Scientist.pdf

http://gizmodo.com/can-you-supercharge-your-brain-with-electricty-1585278622

http://www.scientificamerican.com/article/amping-up-brain-function/

http://www.bbc.com/future/story/20140603-brain-zapping-the-future-of-war

news summary (22)

Categories: Uncategorized