Emergence Flashcards
What does it mean to ‘model’?
To create simulations of how we think something works, in this case, the mind/brain
In the context of these lectures, what is the ‘mind’?
An emergent property of the electrochemical interactions between neurons within the brain
What is an experiment?
A situation in which a variable is manipulated and the consequences/ effects of such manipulation are observed and recorded (the effects on things like memory, language, cognition etc).
We can do this, for example, while someone is inside a scanner –> to measure brain activity as a function of manipulation. At a higher level of investigation, we might combine different technologies e.g. PET, fMRI, EEG etc.
What is the major problem with neuroscience as a discipline?
Sometimes drilling down to the most intricate, biological level, is not the most direct way to begin to understand how the mind emerges from the brain.
The technological breakthroughs being made in neuroscience are not in and of themselves a direct way of making progress towards understanding unobservable phenomena like the mind.
Could a neuroscientist understand a microprocessor?
Basically, you could try your best to understand the processor, but you would not be able to understand what the thing is for, and what function it serves…you wouldn’t see the game just from looking at the board…
YOU CAN ONLY DRAW INFERENCES FROM NEURO TECHNIQUES
Emergent properties of complex systems obey laws that cannot be understood at the level of the component, in the same way you can’t look at the DNA of an ant and see the anthill or the brain of a starling and see the murmuration, you can’t look at individual neurons and see the mind as an emergent property of the brain. SO complex systems cannot be understood by reductionist approaches.
What is computational neuroscience?
This is where neuroscience is investigated and described on a more functional level. So instead of investigating individual cells (and their properties), maths is used to abstract the biological detail.
Computer programs are developed to apply equations over and over in loops, simulating a neuron firing, for example, to see if the models are accurate representations of brain activity.
Talk to me about cats bro
Rosenbleuth and Weiner - REDUCTIONIST APPROACH
“The best material model of a cat is another, or preferably the same cat.”
This basically means that, as is the current direction of comp. neuro., by making more and more complex models of systems… if you succeed in making a truly faithful model, in the end, all you’ll have made is an exact replica of the original thing, not necessarily making any progress. From his reductionist perspective, he said you may understand the internal workings and detail better, but you are no closer to understanding the cognitions, feelings or thoughts of the cat.
What is a mechanistic model?
Mechanistic models assume that a complex system can be understood by examining the workings of its individual parts and the manner in which they interact
What does the emergence lecture cover?
- The brain as a complex system
- The emergence of complexity
- Self-organisation
The brain as a complex system
The brain is incredibly complex, the most complex system in the universe, which is daunting in terms of attempting to model it… BUT
Evolution has created the brain using an incredibly limited set of instructions… amounting to about 50 megabytes (less than Microsoft used to code for clippy). Meaning, there is a huge amount of data compression within genetic material.
Does brain complexity = intelligence?
The complexity of the brain might not be able to be as simple to explain in terms of “more complex brain = more going on” (in terms of cognition, thought etc) in an equal positive correlation…
for example, just because humans have an EQ (encephalization quotient) of 7 and hippos of 0.3, it doesn’t mean that humans are exactly 23x cleverer than hippos. An orca has an EQ of 3, this doesn’t mean they are half as intelligent as humans.
EQ/ TRADITIONAL INTELLIGENCE
Are other complex systems intelligent?
ADAPTIVE INTELLIGENCE
Sepkoski - Intelligence is just one adaptation for survival… running in a herd while being dumb as shit is an excellent adaptation for survival for water buffalo.
Intelligence evolved for tetrapods in the same way that herd behaviour did for buffalo, in response to a specific set of environmental pressures, different species evolve to live under a different set of environmental demands… explain this
… don’t need to be able to solve maths problems or learn new languages… they need to mate, find food and escape predators… operating within own niche of adaptive behaviour in order to out compete others…
What is collective intelligence?
COLLECTIVE INTELLIGENCE
Wildebeast -
Each wildebeast is behaving in an individually stupid way, fighting each other to climb up the embankment. But, collectively, this individually selfish and unintelligent behaviour, gives rise to group intelligence, creating an optimal rate of flow up the embankment, ensuring as many of the animals (their genes) get to safety.
In terms of complexity theory…
Why, from a ‘different kinds of intelligence’ perspective, are brains not special?
Brains are complicated systems, and nature is full of complex systems.
What is emergence?
In this context, emergence refers to the emergence of complexity from the brain.
What is complexity?
Complexity theory refers to the emergent properties of systems, of simple interacting components.
Complexity theory is talking about what happens when lots of simple components come together to form a more complex system, and what can occur when they do.
Complexity exists within the mutual dependency of organisms (the way they help each other and are helped by each other).
Steps of explanation:
- system is comprised of simple components that behave/ act in simple ways
- the behaviour of those simple components is selfish
- the interaction of the components individually selfish behaviour collectively gives rise to more sophisticated emergent properties (collective intelligence)
- the simple components are mutually dependent on each other, as alone, no individual component could achieve such feats
Neuron - ants, fish, starlings
Brain - collective interaction of the ants, fish or starlings
mind - anthill, school or murmuration emerging from the interactions between individuals
Explain some worked examples of collective intelligence in terms of complexity theory
Ants - create anthills, underground tunnel systems, branches using their bodies to collect food… at what level is the plan for these structures specified? Not genetically coded for… but by interacting in simple ways, they can function collectively in complex and intelligent ways, achieving things that no individual bug could achieve alone (codependence)
Fish - move in schools to avoid predators, each individual fish moves selfishly to avoid the shark, this selfish movement gives rise to the collective movement of a school to distract the predator and promote both group and individual survival
Starlings - murmuration forms with individual - group movement, they form the complex structures to both avoid predation and to exchange information e.g. good feeding areas
Termites
also… NEURONS (simple interactions give rise to intelligent, complex emergent properties)
Neuron - ants, fish, starlings
Brain - collective interaction of the ants, fish or starlings
mind - anthill, school or murmuration emerging from the interactions between individuals
Is complexity intelligence?
EQ doesn’t necessarily equate intelligence but does complexity?
There are lots of complex systems in nature/ other areas of life…
- Immune system
- Any system at a cellular level e.g. eyes
- Economy
- Intelligence/ consciousness emerges from non-conscious material substrates
Each component of these example systems works chiefly for their own gain in simple ways, but collectively this individually simple and selfish behaviour gives rise to complex emergent properties that could not be achieved by any individual alone.
In terms of intelligence, complexity may not BE a kind of intelligence, but it may serve to explain how intelligence arises
Quotes
- Rosenbleuth and Weiner
- Deacon
- Camazine
- Goodwin 2001
- Carroll 2006
- Kant 1790
“The best material model of a cat is another, or preferably the same cat.” - Rosenbleuth and Weiner
Complex systems show “a discontinuity of properties despite compositional continuity” - (Deacon, 2003)
“A process in which pattern at the global level of a system emerges solely from numerous interactions among the lower-level components of the system. Moreover, the rules specifying interactions among the system’s components are executed using only local information, without reference to the global pattern.” - (Camazine, 2001)
“[self-organising] systems produce something out of nothing… There is no plan, no blueprint, no instructions about the pattern… [the relationships between components have a naturally occurring spatial and temporal pattern]” - (Goodwin 2001)
“One of the most striking truths about animal body patterns is their regularity at all scales, from the overall body plan to the fine details of an individual structure or body part” - (Carroll 2006)
“In an organised being, the parts exist for and by means of the whole, and the whole exists for and by means of the parts.” - (Kant 1790)
Self-organisation
Kind of the process of complexity as it emerges from simple components, referring to the way emergence of complexity is coded for.
the global pattern is coded for by interactions between components, rules come from only local information - (Camazine, 2001)
there is no large-scale blueprint for complex systems, the pattern occurs naturally from interactions - (Goodwin 2001)
There is regularity in biological systems at all scales - (Carroll 2006)
Co-dependence, more than the sum of its parts?
“In an organised being, the parts exist for and by means of the whole, and the whole exists for and by means of the parts.” - (Kant 1790)
Basically, the parts aren’t special, the components of a complex, organised system (ants, starlings, fish, buffalo, neurons etc) aren’t special, but the plan as played out via interactions gives rise to a whole that is special
Take home points from Emergence as a concept…
- The mind is an emergent property of the brain (and it’s interaction with the body and world)
- Complexity refers to emergent properties of systems of simple interacting components e.g. neurons by communicating with each other in simple ways give rise to a complicated system such as the mind.
- Emergent properties obey laws that cannot be understood at the level of components
Something is lost if you try and understand this complex system of mind by reducing it to it’s simple components. If you try and simulate how those simple interactions play out on mass, then we might discover that simulations can take on properties that we otherwise find difficult to understand. These emergent properties (things that emerge from simple interactions) can themselves behave as though they are governed by rules, which are difficult to see at the level of those simple interactions.
–> Emergent properties of simple interaction can obey law-like rules
- Complex systems (by definition) cannot be understood by reductionist approaches.
Maybe there are limits to a reductionist approach, reducing things down e.g. to individual neurons, is not necessarily going to help us answer questions at top levels of thinking - Computational modelling (running simulations and asking the computer to evaluate those models of simple interactions) is a tool for understanding complex systems like the brain. Can reveal to us things that are otherwise very difficult to see (more than just evaluating our mathematics), computer programs can generate behaviours in a way that is not so easily understood by just looking at the equations that we put into the computer – we can give a computer simple instructions and it can simulate those instructions and generate something interesting and new
