Scribd
Upload
12 views2 pages

Subtitle

Uploaded by

chefabordovideo
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as TXT, PDF, TXT or read online on Scribd
12 views2 pages

Subtitle

Uploaded by

chefabordovideo
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as TXT, PDF, TXT or read online on Scribd
You are on page 1/ 2

Barbocle, back again here.

I had mentioned that generative


AI has more problems, of course, than
just hallucination. One of the biggest problems from our perspective
as professors, is that it makes it
really easy to cheat. This is indeed a problem. But I should say that
years back, let's say, when people were first
starting to learn to write, they were imprinting on clay
tablets with cuneiform. When they were
doing that, people were complaining
because they said, "You, Hey, you're going
to lose your memory. You're not going to remember things if you write
things down." There were a lot of well placed
criticisms of learning to write because people wouldn't rely so much on their
memory. It's rather similar now. People know that there are
bad aspects of generative AI. But the trade off is such
that it can be worth it. But for us as professors
and teachers, generative AI has really pulled the rug out
from under us. It has changed many
of the ways that we have used in the past
to help students learn, for example, writing an essay or about how to work out
their homework problems. For all of these things, learners went off and they
had to do them on their own. You don't need to do that necessarily now
with generative AI. It means that not only
is it easier to cheat, but you can also end up
learning more lightly. Why learn to write an
essay, for example, if you can just go to generative AI and ask it
to write essays for you? Well, that is a problem. As we're going along in the next
decade and
the decades to come, people will be figuring out the solutions to some
of these challenges. But there's another aspect or another way of
looking at this that can help you understand why it's still important to learn
some of these ideas, whatever you're
studying, to learn the rigor of mathematical
thinking, for example, and to become very
familiar with some of the key aspects of programming or to learn to speak
another language. What I'm really
driving at is related to something called
the Flynn effect. In the Flynn effect, researchers found that from
the 1930s to the 1970s, IQ scores rose dramatically. Well, why was that?
Finally, the conclusion was it was due to education. People had improved access to
good ways of learning and, of course, improved education. As a consequence,
IQ scores went up. People got smarter. But here's the funny thing. From the 1970s
onwards, there's plenty of
research showing that IQ scores are declining. This is even within a family. You
can see that kids
who are younger, born after the latter
part of the 1970s, have a lower IQ than older
kids within the same family. Why would this be? In the 1970s,
calculators came out. When calculators came out, suddenly it became, well, you
don't need to
remember things anymore. You can always just
look things up. As you'll see later
in this course, that's a real problem. Students stopped
practicing within their own mind and
remembering things. What this really
tells us is that with the revolution
of generative AI, we don't want to make the same mistake we made
when calculators came out. We want to keep learning
and getting practice with some of the essential ideas that are really important,
for example, learning the
multiplication tables or learning how to put
together an essay. We don't want to be
like the 1970s where everybody got all excited
about a new technology, and then they started saying, you don't need to
actually learn some of these important ideas because you can always go look it up.
This idea that generative
AI is problematic is, I think, an important one
for us to keep in mind. But I also want to talk
about something that's a little bit in defense
of generative AI. I wasn't good at math and
science when I was growing up. I enlisted in the army
out of high school, studied a language, I
picked Russian at random, and became a translator. But then I realized that
I'd followed my passion, just as everyone
had told me to do, but I hadn't
broadened my passion and started learning
about broader fields. That meant when it
came time to get jobs, that was pretty tough for
me to do because few people were interested in
my Slavic languages and literature skills. At age 26, I went to the
university and decided to try to retrain my brain and
learn in math and science. I started at the lowest
possible level of remedial high school algebra and slowly began climbing
my way upwards. Now, of course, I'm a
professor of engineering, and I love math and science. But one challenge that I had
when I was trying to retrain my brain was I found that professors and teachers
would
often keep things hidden. In other words, you might get
a lot of homework problems, but your teacher didn't
necessarily want to give you the solutions to the
homework problems because then you'd know
what the answers were, and some people would
use that to cheat. You couldn't get a lot of
problems to solve and to work with something that
related to exactly what the professor was
teaching in the course. It was just difficult
to find lots of practice problems
because teachers certainly weren't
going to be willing to share their testing
or practice materials. But nowadays, it's
very different. If you're really
interested in learning, you can get lots and lots of practice materials from
ChatGPT and other engines. You don't have to worry so much that the knowledge is
something reserved only for those who worked really hard to create the test
materials and such like. The world is more open for learners who truly
want to learn, and you want to be someone
who truly wants to learn. Again, I refer to
Fei Fei Li's quote, AI won't replace humans, but humans that use AI will
replace humans that don't.

You might also like