Please wait while the transcript is being prepared...
0:00
Hello, everybody. I'm pleased to
be invited to give you a talk on
a topic which is so
relevant today and so
talked about today;
Generative AI in
drug development.
I'm Sarfaraz Niazi.
I'm adjunct professor at
University of Illinois.
Been teaching all
my life in Chicago.
I have diversified
into many areas,
written several books
on this subject,
and I'm very pleased
to provide you
literally a summary of where do
we stand with generative AI.
0:42
Generative AI is a word
that has become literally
a household word especially
in the media and in the
scientific literature,
but one has to understand
what this technology does.
How it affects or how it
changes the development
of drugs is one of the most
significant application.
It takes billions of dollars
to develop a new drug,
and generative AI can not
only cut the cost down,
but more importantly,
cut the time down so
the drugs can reach
to patients sooner.
Generative AI, if
you look at it,
it's the last part of the train.
First you start with
artificial intelligence.
By the way, in reality
this is not artificial.
The intelligence is real,
but it's developed by a machine.
A better word maybe
adaptive intelligence,
but I'm not arguing
with that because
that's accepted terminology.
Machine learning.
How do you teach a machine?
In the case of
today's computers,
machine learning is simply
the memory kept there
and memory accessed.
Think of this way. You have
read every book in the universe,
you remember every word,
and at any single point,
you can gather all the
information together.
That's the intelligence that
we're talking about here.
That was never possible before
until the modern range
of computers came in.
Neural network is
almost like your brain.
When you are thinking
of your neurons,
you split the message across
many areas and then bring
back the information.
Think of this way.
The computers now are
working as if they
have a brain cells.
Deep learning is a machine
learning combined with
neural networks to learn
and extract the features.
It's not just a paragraph
that you have read.
But if you have
read 10 paragraphs,
can you consolidate them
which leads to the
generative AI,
which is basically
the term says,
you are creating a new content.
It could be images. It could
be an article you write.
It could be a theory,
whatever it is.
That is the base definition.
I wanted to share with you how
