Meeting #2
Friday, 2/18/2022
View the meeting recording here. Note that the audio quality is poor at times.
Table of contents
Agenda
- Take a team photo!
- Presentation on project guidelines, timeline, and member expectations;
- Polling what members are most interested in learning about;
- Break into groups by topic, assign presentations; propose project ideas that could incorporate that project;
- Brainstorm principles and requirements of a generally intelligent system;
- Discuss possibility of a journal club;
- Solicit personal information to put on the website
Resources
- Meeting Slides: “Interactive Intelligence”
- Janna’s Neuroscience Images
–
Notes
Characteristics of Intelligence
- Creativity (what is creativity)?
- Procedural
- Randomness as creativity (e.g. random vector inputs to GANs or VAEs)
- Novelty of strategy
- Abstractions
- Meta-knowledge, self-identity, self actualization
- Organisms do not necessarily take immediate action for their survival
- Long term planning as an indication of intelligence
- Anticipating or forecasting consequences or rewards
- Trauma and intelligence - trauma has been shown to impede brain structure.
- Is intelligence static? Can smart people be dumb in the moment?
- Is intelligence a scale, a ball? How many dimensions of intelligence are relevant?
- Transferability across environments
- Parasites and hijacking
- Get into hosts and change the structure of the brain or the body to benefit themselves
- Who is smarter: the parasite or the person / host?
- Replika
- Example of people in real life falling in love with code – is this code intelligent? What does it understand?
- Concepts of slavery and servitude
- Robot - originating from Slavic for slave
- Race After Technology book recommendation
- Using the environment as a metric or indication of intelligence
- Communication and intelligence
- Empathy could be useful for intelligence
- Teaching as one metric or indication of intelligence
- Is it possible to be intelligent without action? Yes – paraplegics for example. Is it possible to become intelligent without ever taking actions?
- Role of central nervous system in intelligence
- Organisms without CNS can still make decisions, ie fungi choose which route to take for food
- Changes the time scale of action due to electrical properties
- Are trees intelligent
- We can’t know certain things, can never experience the perspective of the tree, can never know the experience of others
- Standpoint epistemology as it relates to understanding intelligence
- Is the neocortex necessary or helpful for intelligence?
- Is emotional intelligence necessary or helpful – socialization
- Rate of information transfer between individuals as a hallmark of human intelligence
Projects
- We’re a project-based club!
- Goals - learning, presenting about topics we know about, designing, principles that govern what we want to make
- Member Expectations
- Let’s actually do stuff!
- Projects and tasks - if you are assigned to do a task and you have difficulties, be communicative - inform project leaders and communicate!
- We still have a lot to learn - we don’t know exactly where to go with a project.
- We will identify topics we want to identify and research more deeply in groups of 2 to 3 people
- After presenting on these topics and suggesting several projects for each, we will narrow down which we will work on
Project Suggestions
- Libraries available for us to use
- For example, TF, Keras, NumPy
- Biologically plausible neural networks
- e.g. Spiking neural networks, Hebbian learning
- Biologically/neurologically informed neural networks/computational structures
- Eusociality
- Meta-modeling, meta cognitive networks
- Swarm intelligence of robotics or coded agents
- Simulated training for robotics/reinforcement learning settings - translation into the physical world.
- Cybernetic enhancement - using neural networks to enhance the properties of biological entities
- e.g. Neuralink