#4 Social Neuroscientist Dr Redmond O' Connell - How we make Decisions, Free Will and How Intelligent is AI really?

Dr Redmond O’ Connell is a cognitive neuroscientist from Trinity
College Dublin, the college where I received my degree.


He has spent decades studying our decision making systems in the
brain, how we build evidence in a manner shaped by our experience
and perceptions and ultimately how we make heavily biased
decisions.


Dr Redmond has won many awards and is a brilliant researcher and
scientific communicator. Subscribe on youtube and follow on
Spotify so you don’t miss the next neuroscience podcast, I hope
you enjoy!





00:00 Intro03:23 The Brain's Decision-Making Process04:56 The
Evidence Accumulation Process11:58 Overcoming Biases through
Education and Critical Thinking15:15 Effective Communication to
Address Biases39:52 Introduction and Free Will55:50 Limitations
of Large Language Models01:02:42 Mimicking the Human Brain in
AI01:10:31 Understanding Consciousness in AI01:16:29 Exciting
Advances and Concerns in the Next Decade of Science





The brain's decision-making process is still not fully
understood, but there are some guiding principles that have been
established. One principle is the evidence accumulation process,
where the brain samples and weighs different pieces of
information before making a decision. Expectations and biases
also play a role in decision making, as they shape the thresholds
for accepting evidence. Biases can be difficult to overcome, but
education and critical thinking can help. Effective communication
is crucial in addressing biases and promoting a better
understanding of complex issues. The concept of free will is
subjective and depends on individual definitions and
perspectives.


In this conversation, Redmond O'Connell discusses the concept of
free will and its relationship to neural activity and
decision-making. He challenges the idea that neural activity
preceding conscious decisions negates free will, arguing that it
is a matter of quantity rather than a binary concept. O'Connell
also explores the limitations of large language models like
ChatGPT, highlighting their inability to reason and their
reliance on statistical associations. He suggests that mimicking
the human brain and incorporating embodied cognition may be key
to developing more intelligent AI.


O'Connell emphasizes the importance of understanding
consciousness and the potential ethical concerns surrounding AI
and misinformation.