Loading presentation...

Present Remotely

Send the link below via email or IM

Copy

Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.

DeleteCancel

Make your likes visible on Facebook?

Connect your Facebook account to Prezi and let your likes appear on your timeline.
You can change this under Settings & Account at any time.

No, thanks

Radical post-cognitivism: new approaches to Intelligence and Mind

Inaugural lecture, Goldsmiths, 13th March
by

Mark Bishop

on 13 November 2015

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of Radical post-cognitivism: new approaches to Intelligence and Mind

As A.I. continues to move ever further away from its cognitivist roots -
at the price of losing claim to genuine machine understanding and/or offering insights on the mind
- the application of new A.I. techniques {e.g. big-data, statistics and machine-learning} is finally enabling engineers to make machines that behave intelligently; en-route making progress on resolving 'the frame problem' [SIRI], 'common-sense reasoning' [JEOPARDY] & alleviating 'combinatorial explosion' [DEEP BLUE].

Conversely, from early roots in the philosophy of Maurice Merleau-Ponty and Gibsonian 'ecological psychology', through the work of Maturana, Varela, Brooks, Damasio, Nunez, Thompson, Noe, O'Regan and Torrance, there is now a well trodden path in Cognitive Science leading away from
mere computations
towards building an embodied understanding of cognition as the interplay of 'a brain' in 'the body' in 'our world'.
Since coming online in 1997, Cleverbot has engaged in about 65 million conversations with Internet users around the world, who chat with it for fun via the Cleverbot website. Like a human learning appropriate behavior by studying the actions of members of his or her social group, Cleverbot "learns" from these conversations. It stores them all in a huge database, and in every future conversation, its responses to questions and comments mimic past human responses to those same questions and comments.
On the relationship between mind and body,
"... it is not sufficient that it [rational soul; mind] be lodged in the human body exactly like a pilot in a ship, unless perhaps to move its members, but that it is necessary for it to be joined and united more closely to the body, in order to have sensations and appetites similar to ours, and thus constitute a real man"

Descartes, Discourse on Method V (1618).
In 2011, 60+ years after Alan Turing first published his seminal paper ‘
Computing Machinery and Intelligence
’, at the ‘Techniche festival’ in Guwahati India,
Cleverbot
[a computer program written by Rollo Carpenter] 'passed' a version of the Turing Test:

... in that just under 60% (59.3%) of an audience of 1334 people assessed Cleverbot’s responses to be 'human';

... for comparison just over 60% (63.3%) of the audience assessed the real human responses as coming from a human.
In contrast to theories of naive (or direct) realism, representational explanations of mind postulate the existence of mental representations which act as intermediaries between the observing subject and the perceived objects (processes or other entities) in the external world. These intermediaries stand for (i.e. represent) to the mind, the objects of the world.

In this 'snap-shot' view of cognition it is the job of the cognitive system to form accurate representations of the 'out-there' world.
Computations and representations
Physical Symbol System Hypothesis
: a physical symbol system [e.g. a computer] has the necessary and sufficient means for general intelligent action (Newell & Simon).
Artificial Intelligence
: the science and engineering of making intelligent computer programs and machines (John McCarthy)
The Turing Test
: a human judge engages in text-based conversations with another human and a machine programmed to respond in a manner indistinguishable from that of a human (Alan Turing).
Cognition
: the mental process(es) by which knowledge is acquired and used: cognition encapsulates perception; reasoning, consciousness etc.
Cognitivism
: the view that cognition consists of internal representations (states; symbols) whose manipulation can be described in terms of computations (rules).
The tribulations of computations ..
Combinatorial explosion (James Lighthill's Report)
Frame problems: what remains unchanged as a result of an action; how to compute the consequences of one's action without also evaluating possible non-actions (Dennett).
How to represent and compute "common sense reasoning" [in propositional logic] (Dreyfus)
Engineering autonomy: 'the machine can only do what we tell it to do' (Ada Lovelace)
[Complete] mathematical reasoning cannot be encapsulated in a consistent set of formal mathematical rules (Godel, Lucas, Penrose)
Computation is
OBSERVER RELATIVE
: 'the meaning of a computation lies in its use';

Dancing with pixies
: the execution of a mere computer program is insufficient to generate consciousness (Bishop)
2011:
the year of the intelligent machine ?
"I believe that at the end of the century the use of words and general educated opinion will have altered so much that one will be able to speak of machines thinking without expecting to be contradicted" (Alan Turing, Computing Machinery & Intelligence, 1950).
Robotics
GOFAI robots: Shakey and the sense-plan-act cycle
The 'subsumption' architecture; the world as its best representation, (Brooks);
Kevin Warwick and the Seven Dwarfs (above); as conscious as a slug ?
Evolutionary robotics - Sussex room centering robots
Cognitive Robotics: Rolf Pfeifer; moving the cogitive load to materials
Epigenetic robotics: the robot child where the goal is to model the development of cognition in natural and artificial systems.
Animats and Cyborgs
John Searle's Chinese room thought experiment
TO PLAY 'THE RESTAURANT GAME'

EACH FRIEND:
1.
Selects a random restaurant hypothesis
: open ‘Yellow Pages’ and select a restaurant to visit at random.

2.
Performs a 'partial hypothesis evaluation
': at dinner that night select a meal from the menu at random, eat meal and decide if it was ‘GOOD’ or ‘BAD’.

3.
Communicates with friends:
at breakfast the next morning -
IF <
last night’s meal was ‘GOOD’
>
THEN
maintain your current restaurant hypothesis
and GOTO (2)
ELSE IF <
last night’s meal was ‘BAD’
> THEN
communicate with a randomly selected friend
:
IF <
colleague’s meal was 'GOOD'
>
THEN
adopt your friend’s restaurant hypothesis
and GOTO (2)
ELSE GOTO (1).
A way forward: 4e's
laying down a path towards a new Cognitive Science
-
Ecological, Embodied, Embedded and Enactive
-
Autopoiesis
(Maturana & Varela): potentially supports 'genuine-autonomy' and 'no external engineer'.
Under the
Stochastic Diffusion
framework (Bishop, Nasuto et al) we are investigating an over-arching unification of the above deeply embodied models, now extended to support 'computation' in both discrete and dynamic systems (SDPs; coupled oscillators; neural-assemblies).
Interactivism
(Mark Bickhard): suggests a way to ground representations in autonomous systems - no more 'empty symbols'.
Another approach to designing intelligent systems seeks inspiration from nature.
Swarm Intelligence
systems are typically made up of
a population of simple agents interacting locally with one another and with their environment
. Swarm Intelligence agents typically follow very simple rules:

There is
no centralized control structure
dictating how individual agents should react and behave;
instead local interactions between agents lead to the
emergence of [seemingly] intelligent global behavior
.

Natural examples of Swarm Intelligence include:
bird flocking, ant colonies, animal herding and fish schooling
.


TO PLAY 'THE RESTAURANT GAME' EACH DELEGATE:
1. Opens ‘Yellow Pages’ and selects a restaurant to visit at random, so defining the agent’s restaurant hypothesis.
2. Partial hypothesis evaluation: at dinner that night the delegate selects a meal from the menu at random and subsequently decides if it was ‘good’ or ‘bad’.
3. Diffusion/communication: the next morning at breakfast …

IF <last night’s meal was ‘good’>
THEN maintain restaurant hypothesis and GOTO (2)
ELSE IF <last night’s meal was ‘bad’> THEN communicate with a random colleague:
IF <colleague’s meal was good>
THEN adopt colleague’s restaurant hypothesis and GOTO (2)
ELSE GOTO (1).
SWARMS: agents, interactions, and decentralised systems
[[I suggest if the restaurant game is broken down into two squares (slides) that you can zoom in. and same for Dancing with pixies]]
A group of friends arrive in New Cross for a series of lectures on 'Radical post-cognitivism' and most importantly need to find a place to eat that they can all enjoy together. Hence a ‘good’ place to eat is the restaurant where most friends are likely to choose a meal they subsequently deem ‘GOOD’.

An individual friend’s response to a randomly selected meal from a restaurant menu {GOOD or BAD} is termed a ‘
partial hypothesis evaluation
’ as it provides
partial evidence
on this restaurant’s overall quality.

The ‘search space’ is the set of all restaurants in New Cross. A naive 'exhaustive search' - where every dish is tried in every restaurant in town - is impractical; there are simply too many great (restaurant : dish) combinations to evaluate over the duration of the lecture series.
The restaurant game: stochastic diffusion search
In 2011 Apple Inc. launched Siri on the iPhone4S platform.

Siri is an integrated intelligent personal assistant and knowledge navigator which works as an application for the iPhone4S.

The application uses a natural language user interface to answer questions, make recommendations, and perform actions by delegating requests to a set of web services.
Evolutionary Robotics, (Harvey)
Animats, (Warwick, Nasuto et al)
With the advent of cognitivism a ubiquitous computational metaphor insidiously begain to pervade Cognitive Science:

Explicitly
: cognition IS ‘computations on symbols’; GOFAI.

Implicitly
: cognition AS ‘computations on [real] numbers’; neural-networks, connectionism.

Descriptively
: cognitive-modelling VIA 'computational simulations'; the behaviours of neurons [or populations of neurons] explained in terms of their performing 'Hodgkin-Huxley computations'.

And over time this led to the confusion of metaphor and reality (epistemology and ontology) ...

... although we can model neural events with computations, neural events are not constituted by computations.
Cognition as Computation
Content vehicle confusion
Jack Gallant's lab (Berkeley)
The conceptual distinction between representational vehicles and representational contents is very old, probably older than Plato ..
Evolutionary robotics - Sussex room centering robots
Animats and Cyborgs
Mark Bishop (Chair AISB)
the UK society for the study of Artificial Intelligence (AI) and the Simulation of Behaviour
new approaches to intelligence and mind
Radical post-cognitivism:
Centralisation,
representation and the
homunculus fallacy
What is cognition if not computation on representation?
Second Order Cybernetics and Enactive Perception
, (Bishop & Nasuto): conceptual unification of Varela's & O'Regan's Enactivism, Dynamic System theory, and Cybernetics [Norbert Wiener Prize 2005]
The themes and ideas outlined today emerged through countless discussions over many years with my grad-students, post-docs, colleagues, friends and family; in particular: Andrew Martin, Chryssa Sdrolia, Mohammad Majid Al-Rifaie, Darren Myatt; Kris de Meyer, Leo Kiernan, Jeff Hannan, Andrew Cohen, Mike Bushnell, Dominic Aitken, Steve Pepper, Paul Beattie, Etienne Roesch, Kevin Warwick, Phil Torr, Steve Westland, my PhD supervisor 'Mike Usher' and my tutor 'Alex Andrews'; Guy Ruddock, Katerina Koutsantoni and my parents..

And - last, but not least - a man to whom i owe such a lot and blame most for many of these seemingly mad (but really quite sensible) ideas; initially one of my grad-students, subsequently colleague and good friend all rolled into one, Slawek Nasuto.

How do I think ?
Why do I feel ?
Wherefore I see?
My brain,
in my body,
in our world.
COPYRIGHT DISCLAIMER
: Texts, marks, logos, names, graphics, images, photographs, illustrations, artwork, audio clips, video clips, and software copyrighted by their respective owners are used on these slides for non-commercial, educational and personal purposes only. Use of any copyrighted material is not authorized without the written consent of the copyright holder. Every effort has been made to respect the copyrights of other parties. If you believe that your copyright has been misused, please direct your correspondence to:
m.bishop@gold.ac.uk
stating your position and I shall endeavour to correct any misuse as early as possible.

Stevan Harnad suggests that a better test than The Turing Test for intelligence will be one that requires responses to all of our inputs, and not merely to text based questions. I.e. the appropriate goal for research in AI has to be to construct a robot with something like human sensorimotor capabilities.
Cognitive science
: the interdisciplinary scientific study of the mind and its processes.
Intelligence as Computation
2012: A.I. and Cognitive Science: a final parting of ways?
Full transcript