AI IS CHANGING ACADEMIA

Episode Transcript

Speaker 1 (00:01):
Science, elastics and thoughts and pins from the hands will
make the weird science. Welcome to the Weird Science Facts Podcast.
We provide a five minute or less daily podcast on
weird science, debunking myths and old sayings. So joined doctor

(00:23):
Carlos in this daily diet of fun, strange and weird
world the Facts.

Speaker 2 (00:32):
Have you ever wondered what your furry friend is really thinking, Well,
let me introduce to you my new book, Dog Psychology
by doctor Carlos Vasquez. As a psychology professor with over
ten years experience, I've unlocked the secret so the Canine Psyche.
I will teach you how to understand your dog's body language,
to code their barks and wines, sold behavioral puzzles, and

(00:53):
strengthen your pond like never before. With Dog Psychology, you'll
see the world through your dog's eye. Don't just be
a dog owner, become a dog whisperer. So go get
Dog Psychology by doctor Carlos Vasquez, available now on kend
or paperback on Amazon. They go back and visit the

(01:16):
world of AI and let's see what's going on now.
In the academic field, this seemed to be having a
lot of problems with AI. There have been some surveys
that have been done over the last couple of years.
One survey launched just a few months after chat GBT
was launched, a thousand college students found the nearly ninety
percent of them use chat schiptit to help with homework assignments.

(01:39):
Now they're saying anywhere between fifty to sixty percent who
at least admit to it. In his first year of existence,
chat GPT told monthly visits steadily increased month over month
until June when schools lent out. Oddly enough, this wasn't
an anomaly. Traffic dipped again over the summer. In twenty
twenty four, Professors and ta's increasingly found themselves reading essays

(02:00):
with kind of robotic phrasing and clunkiness, but grammatically flawless,
and they didn't sound like it's college student. I am
even wondering if this little article snippet is written by AI,
because it has a lot of these dashes, which is
commonly known, and a lot of students don't delete them,
so you'll have a sentence. We normally people would put
a comma to separate a little bit of a transition

(02:20):
and come back, or an elaboration of a point and
they actually put dashes. And in this article they actually
have dashes too, so it's kind of funny. Two and
a half years later, students at large state schools, the ivs,
liberal art schools, in New England, universities abroad, I'm saying
it's over nationwide are relying on AI to ease their
way through every facet of their education. Now there's different

(02:42):
types of AI chat bots, Google's Gemini Claude, Microsoft's Copilot Grock,
and they take their notes during class divides, their study
guys and practice tests, summarize novels and textbooks, and brainstorm,
outline and draft their essays. Some of these things are
actually great for the students, such as developing outlines, creating
their own practice tests. I think that's wonderful. It's when

(03:05):
you actually have a whole paper written by AI for
you that you have a problem. STEM students are also
using AI to automate their research and data analysis and
to sale through dense coding and debugging assignments. Well, and
this one's kind of a mixed bag. I'm not a
huge research guy, so I don't know how much of
this is really going to be a huge issue. I mean,

(03:26):
there's so many calculators and things over the years that
have made it easier and easier to calculate research. Could
this be helpful for a lot of people when they're
doing their own research and doing data analysis and coding
and asking challenging questions to AI. One thing I have
learned is that AI is limited only by your knowledge base.
In other words, if you don't know a lot about

(03:47):
a particular topic, you only ask generalized basic questions, and
sometimes they can expand. And one recommendation is always to
give you ask it to give you both sides of
the story. So I would assume even in research data analysis,
you can do a lot with it. And the more
you know about research, you can ask what types of
tests should you run? Why? The pos and cons of

(04:08):
each of these tests? There's a lot to offer there.
College is just again how well or anybody really you
can use GPT. At this point, it was a student
in Utah. They recently captured a video of herself copy
and pasting a chapter from her Genocide and mass Atrocity
textbook into chat GPT. Another freshman at a university in

(04:30):
Ontario says she first used chat GPT to cheat. During
the spring semester of her final year of high school,
after getting acquainted with the chap bought, the girl used
it for all her classes in digital studies, law, English,
and a hippie farming class called Green Industries. The grades
were amazing. It changed her life. Sarah continued to use

(04:52):
AI when she started college. In the past rarely did
she sit in class and that's the other students laptops
open to chat GPT. Towards the end of the semester,
she began to think she might be dependent on that website.
She already considered herself addicted to TikTok and Instagram. With
ch at GBT, she says she can write an essay
in two hours and normally takes twelve hours. This is

(05:14):
something too. Is another new app now that creates power
points called Gamma, and you can create a PowerPoint literally
in less than two minutes. You can get your notes,
copy and paste it in Wailah. You got yourself a PowerPoint.
It's really crazy. But teachers have also now been accused
of trying AI proofing, or actually teachers have also been
accused now of using AI to grade papers. So you

(05:36):
have AI writing and AI grading. The only human interaction
is controlling the computer itself. Teachers have tried AI proofing assignments,
returning to blue books, or switching to oral exams, which
is probably the best way to do it. You want
to create exams in class, exams that are written and
do oral exams oral presentations. If the student writes the paper,

(05:58):
then they can have notes, but the have the paper
and they have to make sure they can answer your
questions in regards to the paper. This is what I've
seen a lot of professors. Some professors starting to do
now is they'll actually give the last the student to
give them the paper, or the professor will read the
paper before they present, and then they'll have some notes

(06:19):
to give and ask questions to the student. But that's
not it. There's a lot of other things that AI
is being used for teaching. A course called Ethics and
Artificial Intelligence is a low stakes reading reflection would be safe.
Surely no one would use chat schubeti to write something personal.
But one of the students for a professor in Santa
Clara University turn in a reflection with robotic language and

(06:41):
awkward phrasing that the professor knew was AI. Generated at
the University of Arkansas Little Rock. It caught students and
Ethics and Technology class using AI to respond to the
prompt briefly introduce yourself and say what you're hoping to
get out of this class. It isn't as if cheating
is new, but now as one student, But if the
ceiling has blown off, who can resist a tool that

(07:02):
makes everything easier? And it's true, massive numbers of students
are going to emerge from university with degrees and into
the workforce who could be essentially illiterate. I don't completely
agree with that statement from one of the professors of
cal State Chico, because I do some of the students
are actually reading it and gaining a lot of interaction
with AI. But I do agree with him that most

(07:23):
of students are not. They're just copying and pasting. Look,
if you're going to go that route, you should really
read the material that AI is printing out and compare it. Obviously,
you should read your textbooks that you're being given instead,
but if anything, at least read what you're being given,
both in the literal sense and in the sense of
being historically illiterate and having no knowledge of their own culture.
Must letch anyone else has said the professor that future

(07:46):
may arrive sooner than expected when you consider what a
short window college really is. So we'll see in the
next three to five years what kind of college graduates
we're going to get. Before I open AI released, cheating
had reached sort of a zenith at the time. I'm
many college students and finished high school remotely largely on
supervised with access to tool like Check and course Hero.

(08:06):
These companies advertise themselves as vast online libraries of textbooks
and course materials, when reality, we're cheating multi tools. For instance,
sixteen bucks a month, Check promised answers to homewerk questions
in less than thirty minutes twenty four to seven from
one hundred and fifty thousand experts. When Chat GPT launch,
students were prying for a tool that was faster and

(08:27):
more capable. There's no way to enforce an all out
chat GPT band, so most adopted an adholic approach, leaving
up to professor to decide whether it allows students to
use it or not. But again, the billdy depends on
how you test them. I believe if you test them
in the oral fashion, the old way. You'll get a
lot out of this, so we'll stop there. But I

(08:49):
think this is a really interesting podcast, and I think
professors can counter this by having them go back to
oral presentations written inside in class, written exams, and if
you write research papers and things of that nature, then
they should actually also have to present in class on
those papers. You could do discussion questions, but they should
be video only, which is a bummer for the professor

(09:13):
because they're going to have to sit there for five
minutes when they could have read something in a minute
or two. But it also forces the student to do
it now, I guess they could always read a prompt
about their discussion questions. It's definitely a big problem in
today's world in academia.