0:20 - DB
The problem is that really nobody knows how to find a job.
0:23 - MM
Right, right. Do you mind if I do my 15-minute presentation
first? This might help guide the discussion. Have any questions, I can assist
with that.
0:36 - DB
Well, is it related to finding a job in AI?
0:40 - MM
Well, that and with career services letting them know how to
utilize that information, too.
0:47 - DB
All right. We can do that. OK. Let's keep it to 15 minutes,
though. If it starts to run over, it'll impact the other discussion.
0:56 - MM
Oh, no. Trust me. It'll be 15 minutes.
0:58 - DB
15 minutes or less. 15 minutes or less. All right, well, now
let's go ahead and get started. So we're going to talk about how to find a job
in AI today, or we're going to have questions about it. First thing is that MM
from the Career Services Office will give a presentation about the services
they provide. And then we'll switch right into the discussion about how to find
a job in AI. Right ahead, M.
1:32 - MM
OK. May I have sharing access?
1:36 - DB
Yeah, you should have it right now.
1:37 - MM
Let me try one more time. OK, perfect.
1:44 - DB
And so again, for those of you who just arrived, so MM from
Career Services will talk for up to 15 minutes, and then we'll talk about the
subject of today's today's meeting.
1:59 - MM
Okay, great. Thank you all for inviting me to this wonderful
discussion. I wanted to provide you first with some information regarding
University Career Services. So I'll start with our actual page. One way to find
out resources about whether it's negotiating your salary or resume writing or
mock interviewing for you need to know the basics first before moving forward,
we have that information on our YouTube channel. I'll also go back to some
career readiness information. We also have, we're part of an organization
called NACE. I recommend this to my students to understand the different
competencies that employers are looking for. Here is a list of those competencies
and along And with that, more details on what those competencies entail. In
addition to this, if you have a CALS library card, you can get LinkedIn
Learning for free. And we curated different types of videos that they provide
that are in relation to the core competencies that employers are looking for. I
definitely recommend using that access.
3:18 - MM
Another piece of information we have in the career readiness
portion is, go back really quick to this, where we have an outline if you needed
to devise a plan for your four-year map out of your career, resume writings,
interview prep, and communication network. We do have a, Communication and
Skills Center that can help even further with how to network or even having a
meeting with me on those particulars. When you go to resume and cover letters,
you'll get different templates. There's a resume review rubric, how to create a
template, and along with presentation slide decks. So now I'm going to move on
to the Handshake profile. This is our career management site. So this is
similar to whether LinkedIn or Indeed, but it's for colleges. To access, it's
the same setup when it comes to job search strategies. You click on jobs, and
you can filter down your different interests. And I'm going to talk about AI in
a minute. If you notice right here, we have a generalist AI trainer. So there
is information on here. Additionally, if you go to Career Center, which is more
focused on hours, Center in general, you will be able to see a couple of things
that will be important. So to set up an appointment, you click Appointments,
then you click Schedule an Appointment, and then University Career Services.
And this will tell you all the different options we have for our services. So
if you need a quick chat, resume cover letter, job searches, or even the Trojan
We provide that you get five garments. That's for a year and you can come back
the next year and it's completely free. All of our services here and you can
set up an appointment. I'm going to go back to the career center piece. And for
our upcoming career fairs or any events related to career services, you will
find this information here. What I tell the students is use this as your power
tool.
5:39 - MM
Find out the employers that are going to be there and map
out the ones you want to reach out to. Another pro tip I heard from another
career service professional, they say, go to the ones that you really don't
want to maybe go to. If you have any nervousness, you can practice on them until
you get to the ones that you really want to go to. So that's just a little bit
of a tip. If you're not able to make it, still reach out to them and say, I
have a time conflict. Is there another time that we can communicate? And so
you're still being engaged and active. So this is our great power tool. We
also, I want to mention that you can find fellow Trojan alumni here. I tell
students to not be discouraged if no one communicates with you, but maybe
they're working at a particular place that you're interested in and you can
reach out and say, Hey, I'm a fellow Trojan. Is there a way for us to connect?
So this is a little bit more, I would say, personal than maybe LinkedIn. And it
does have options. Now, I was excited. Now, Dr. Bernard said that I might not
know anything about some particular AI jobs. But one thing about Handshake that
I did notice is that they have a fellowship, I believe. And I'm going to go to
their actual website. So give me one second to find that.
7:02 - MM
It's taking me back to my account. So if you click on
Handshake, and I think it is the Handshake I believe they have a fellowship
opportunity that's going on. And you can see about participating in their
research. So that's just- It was one click above it said opportunities. Oh, OK.
Do you see it? OK, there it is. So there's actually two setups. There's one
that they have for fellowships, and then they have another one that is mainly
for their career side. So we do have some opportunities that you can see down
here for AI and finding jobs. I think I'm under my 15 minutes. I'm definitely
open for questions as a part of the discussion. But I'm going to end here.
Again, my name is M. I'm the career specialist, and I'm here to be able to
assist you the best way that I can. And M, I've heard so many wonderful things
about you. Nick's office is right next to me. You are the rave. And I hope to
have a more in-depth conversation with you one day, too. That's all I have, Dr.
Bernard. I'll pass it on to you.
8:18 - DB
OK, thank you. Any questions for M or about participants?
Services or anything like that?
8:27 - EG
No questions. I have a question.
8:30 - HA#+#MM
I have a question.
8:31 - HA
You said something about because I am looking for a job in
AI. Thank you. And I'm a student, a graduate student. So good evening. Thank
you for being here.
8:40 - MM
Yes, yes.
8:41 - HA
So you're saying that you found AI jobs and you were a
little bit too fast for me. I was sorry. I'm I'm just starting to log in into
this handshake. That's the first time I heard about So I'm just wondering if
you could just share where you found exactly jobs in AI and that. Okay.
8:58 - MM
So there's two ways that you can look at it. So, okay, if
you type in, you can type in Handshake. So think of it this way, you know, most
companies will have their own career platform. So don't think like on the
actual Handshake site, which you can still find jobs there. But if you go to
Handshake in general, and you hit career I'm sorry, Handshake and then
Opportunities. So this is the opportunities that they're providing through
Handshake, like the Handshake actual company, you would be working with them.
9:36 - HA
Okay. You can work with them.
9:39 - MM
Now, handshake the other and I'm going to see if I can find
that piece. Is it under research?
9:47 - MM
So AI research. I would look into these different areas
because I know I saw information regarding them having a fellowship and let me
see if I can find that again. Let me just type that in handshake fellowship.
10:07 - MM
And what I'll do I'll just add this into the chat and see if
it's still here.
10:19 - MM
If I find more information on it, I'll definitely share it,
but you can go within Handshake itself and find information too. So like if you
were to type AI jobs, just like any search, it will provide you those provides
you that listing?
10:43 - HA
I'm trying to find it here, where I would find that listing.
OK.
10:48 - MM
See, they may not have there. You would think so, but maybe
if you type in Handshake AI, let's do that first. Because I don't want to take.
And you'll see particular positions that are there and the handshake feature.
11:06 - DB
So M, if a student like H or any other student wanted to
have some follow-up questions, can they make an appointment with you? Oh, yes.
11:14 - MM
You can do a quick check, and we can speak to you. Thank
you.
11:17 - HA
You're welcome. Anyone else have a question for M?
11:25 - DB
OK.
11:27 - MM
Well, thanks again, MM of Career Services.
11:33 - DB
And our next item on the agenda then will be to, I'll share
my screen, we'll go to some natural questions that anyone has about how to find
a job in AI. So I do need to share my screen.
11:53 - DB
Yeah, okay, well, any questions? Any questions about how to
find a job in AI? Anybody?
12:07 - EG
I'm going to speak from my experience since I've worked in
AI.
12:11 - Unidentified Speaker
It's different depending upon your education level and your
age.
12:18 - DB
Is a young person's sweet spot.
12:24 - EG
But it also depends upon your education. You have to
basically go through mentorships.
12:40 - EG
I forget what it's co-ops and co-ops where you can actually
get experience doing it before you could really head into the workforce. At a
senior level it's a completely different you have to have a GitHub where people
can look at the history of what you've done to see your approach to problems.
OK.
13:09 - DB
All right.
13:14 - DB
I think, M, you were actually looking for an AI hire at one
point. Yeah, I found one. You found one, OK.
13:24 - MJ
Yeah, so I've been doing work. I still have a company called
Smart Media, and I'm still doing work for some companies. I'm combining AI and
marketing, specifically for e-commerce. And then one of these clients asked me,
since I've been doing all this AI training, if I knew of who could help him do
an integration application development thing. And this is one of many that will
come. So he has many needs, but this is the first kind of lowest hanging fruit.
So yeah, I found a graduate from UA Little Rock, who I know and have known for
a couple of years. And when I was taking my little MIT class coding for
application development. He helped me out a little bit. And he works for
another company, but he had the time to help me work with this company. And so
he's creating, you know, we're about to go over the MVP of what he developed.
And then, and I will say that, you know, it's really about problem solving and
to what extent what they really need is about, isn't necessarily AI. I mean,
it's, it may, what they might need is just stand up code. Exactly.
14:49 - EG#+#MJ
Yeah, it's really about solving problems still.
14:51 - MJ
And the other thing I'll say about AI is, you know, if
you're going to be developing something that's going to go on a manufacturing
production line, you know, you have to have an understanding of software
lifecycle, your regular old software development stuff, you know. SDLC.
15:13 - EG
Yeah, you can't just know about AI.
15:15 - MJ
You can't just vibe code your way through that. I mean,
you're going to have to, because I'll be standing by for maintenance or any
other issues that might come up in the future. And so I have commitment from
him and others to help if it's needed in the future, which it will be. We know
it will be, right? It's going to happen. We're going to need maintenance.
That's a foregone conclusion.
15:41 - EG
In fact, when I mentor developers, I tell them don't code
for yourself. Code for the people who are coming behind you, so that way they
understand it. But yeah, you're right. A lot of people use the term AI. And
what they really mean is a rules engine. They're not looking for AI, they're
just looking for specific things to occur at specific times.
16:03 - MJ
And by the way, we went through this problem that they have,
and AI is is definitely a part of the solution. I mean, it is going to be an
important part of the solution, because what they're doing is pretty
complicated, lots of variables, and they've been trying to code for it and not
being successful. So I think that we're going to be able to knock this one out
of the park. The other thing I'll tell you is that a lot of companies have
these hackathons, and we have a hackathon. So we're coming up on our second
year of the AI Hackathon, Coding for Wellness. And Van and I had a call with
NVIDIA yesterday, which was extremely exciting, because if we can have 150
students at this hackathon, and to give you an idea, we had 40 year for our
inaugural year, then they'll bring luminaries to speak, and they'll have people
from NVIDIA there to help the students with their development and they'll set
up a discord and coach the students. And so I've been Fannie, just so you know,
since then I've been just kind of working nonstop to see, can we actually
handle 150 students for the hackathon? And so far I'm getting all, all my chat,
all my boxes are getting checked positively. I've got more support staff
coming, I've got more space in the going to more money. But I just wanted to
share with you all that a lot of companies use hackathons to find employees.
17:42 - DB
So those winning these hackathons are doing really well.
17:46 - MJ
They use them to get to find. And Nvidia has already told us
they want to have a recruitment booth for Nvidia at our hackathon. So this is a
very big deal. I mean, they're doing them. Everywhere, you know, MIT, obviously
Harvard, Stanford, you know, but also other smaller community colleges, Santa
Clara College, you know, others, you know, if you can get 115 students at a
hackathon, they're going to be there to recruit. And we have interest also from
Google. And then we're, I'm following up on a lead from Anthropic. Thanks. And
I'm also following up on another lead from, from Microsoft. So co-pilot. I
think that we're going to have a lot of great resources, and it'll be a great
recruitment opportunity. And I do agree with you, EG. Unfortunately, I think
they are looking for young people.
18:39 - EG
No, I know that. But the thing is, I've got a little bit of
a name out there. I've won some Kaggle competitions. Kaggle will actually get
your name out there, but the tough part, I I think a lot of it is a lot of
these hackathons are on site.
19:03 - EG
They're for that type of networking. And it's difficult for
people who are remote. Like, I am really, really remote up here.
19:16 - DB
But I mean, I think, I mean, the topic today is how to find
job and it sounds like going to a hackathon is a way to find a job in AI.
19:26 - MJ
It is definitely a way to find a job in AI.
19:30 - DB
Okay, cool. Anyone else have any questions or follow-ups at
this point? I do have a question.
19:38 - MM
So, in the career service field, for us, it used to be a
young person's game, but the issue is, this is the reason why we have the NACE
competencies, because they don't have the soft skills to keep a job. And so I'm
wondering how important that is within this AI field, because yes, you create
this talent, but yet they don't have the skills to come in on time, or they
have these other different things, or they can't completely problem solve. So
how does that, really work because I feel like that's where you know maybe the
non-traditional student comes in because they have stability, maturity that
comes along with the process and they're trying to balance that out. A lot of
students, a lot of young people are being let go. Do you all have that same
experience?
20:26 - EG
Well I did but what we do is we bring them in into a
program. We'll bring in And maybe five or 10 actually make it through the one
year program. 40. And it goes through, they're provided a mentor, and they're
taught all of these skills.
20:47 - MM
When you're just out of college, these don't exist.
20:52 - EG
You're used to getting up at a certain time, doing certain
things, being told exactly what to do, not being given the opportunity to
really think. In the real world, the data is not going to be perfect. You're
going to have ugly data. You're going to have to learn how to do your own EDA.
You're going to have to be able to learn how to fix it. That's not something
that's really taught at the undergraduate or even at the master's level.
21:23 - MM
Well, just a note on our side, one of the things that we do
try, we don't maybe know the tip, technicality to certain things, but like
professionalism in the workplace, things of that nature, we do have career
readiness modules that we're working on to help with the students. And that's
another thing we're trying to incorporate with our smaller internship programs
to help them with that piece. So it hopefully alleviates that in the future.
21:49 - MJ
And M, I can address that a little bit because I work with a
lot of companies and I train their people how to use AI, right? And here's two
trends I'm seeing. One, they want their employees who already understand their
processes, their operations, who already have skills and knowledge in their
field to be able to apply AI effectively in their job role or department. And
the second thing is that the jobs that threatened are the internships.
22:28 - MM
We're already seeing a big giant dip.
22:33 - MJ
What we've been doing historically has been, especially for
big companies like PricewaterhouseCoopers, Accenture, you know, all these
Deloitte, all these big companies who go around and try to scarf up a whole
bunch of interns from universities around the country.
22:47 - MM
They're getting less Interns who are using AI.
22:55 - MJ
So now to become an intern, you have to have some AI skills
around research and report development and all these kinds of things. The other
thing is I sat in a meeting today and I heard that, you know, at Walmart for,
it doesn't matter what job, they're going to ask you about AI. And they do the
same thing with vendors. You know, they're not interested in working with
anybody who's not going to be bringing bringing their AI game to the table. So
this is very much an AI and kind of capability.
23:28 - MM
So I think that it's going to be important for people
looking for a job to say, you know, my real expertise is whatever.
23:35 - MJ
I'm not talking about software developers so much, you know,
because the majority of jobs are not for software developers. The majority of
the jobs are actually like applied capability in a field. So if you are, you
know, what do you know, you know, that you can apply to a field, and then how
can you stand out and get the job because you have AI, in addition, you know,
knowledge and experience and, and, and like you were saying, EG, a bit of a
brag book, right, some examples of some of the things you've done. And why the
GitHub, you know, they want to actually see the work you've done. Yeah, so I
think that that's the winning combination. And it's going to be harder to get
internships. And a lot of companies are just trying to cut down on some jobs.
And those are the ones they're cutting down on. H?
24:38 - HA
Can you hear me? Yes. OK, I'm having a little bit an issue
with the new mouse here.
24:47 - HA
Two things. I am here in Florida. I'm actually a high school
teacher of cybersecurity and AI. And I'm also a graduate student of MSIS with
UALR. And I'm looking at getting a job in AI as my area of interest. I'm
finishing my course now this semester. And so I've been going around different
places here and talking to people. And what I found is two different approaches
to career placement. And I don't know if you, if UALR is thinking about it or
already has it or has something better. One is, there is a college, Miami Dade
College, is actually very big. They actually pair very early on as the students
are coming into the AI program, they paired them with Google, Microsoft, and
some other even local, potential employers, and they do a mentorship, whereas,
you know, the people from Google, Microsoft, the local companies select, they
go around the room, talk to students, you know, have lunch with them, interview
them. And then they kind of get that bond, which later may develop into
internship or just even if it's just a mentorship, somebody who can kind of
help guide you into the job market, because they're working for an AI-related
company. So that's one college that they're doing that. I find it amazing
because as you get in, they already are talking about placement, which I find
amazing. What's the name of that college again?
26:19 - MM
I'm sorry. Miami Dade College.
26:22 - HA
Miami-Dade College. And then there is another one, Florida
Atlantic University. And what they do is they have the senior career events.
And one of them is computer science computer science, cybersecurity, but they
have one themed for AI. And then people come in and the students just show
their project related to AI, right? And it could be like, you know, how to find
a certain type of problem in a tissue, you know, or there was one about finding
whales so ships wouldn't hit whales here in the waters of Florida. There was
one about computer vision with LiDAR versus this is just camera, et cetera.
There was all kinds of things. Some of it's kind of already in the market. Some
of it were new, but it was interesting to just see all that these students are
doing. And there were really people from, you know, all these companies, you
know, NVIDIA, Google, Microsoft, and local companies going around with a little
piece of paper and ranking these people, you know, and I found that interesting
that I don't know how many people really because I was just participating in
it. But that was an interesting event because they were seeing what these
students are capable of doing.
27:40 - Unidentified Speaker
So I know you have career fairs. I don't know if how those
work, if it's just like you go around, interview, pass resumes, et cetera.
27:49 - HA
But I just found these other two very interesting. Of
course, I'm not with them, so I don't have access to it. So that's one of the
reasons I want to see what we have here. Here at UALR.
28:04 - MJ
So I know Miami Dade well, and I'm a big fan. They started
out as a community college, so they're very focused on workforce development.
That's kind of in their DNA. Yes. They're really, they're really, I mean,
they're now, they're not a community college anymore, but they're, yeah, but
they've always on getting people to work.
28:30 - MJ
Day one, I was impressed.
28:31 - HA
Yeah, I'm impressed too with them.
28:34 - MJ
All I was going to say is that I've also spoke to a guy from
George Washington University a couple of weeks ago, and he is responsible for
teaching a class. And M, I think this is such a great idea for UALR. He teaches
a class to students to prepare them before they to go in to try to get those
kind of jobs that you were talking about, those internships that lead to
placement in these companies. And he teaches them how to stand out from the
rest of the students. He teaches them a lot of those soft skills. He teaches
them to seek out a mentor. He teaches them a lot of things. And even though
George Washington University doesn't have the reputation, maybe some other Ivy
League They actually, their interns getting into full-time placement, they hit
high above their weight. They punch high above their weight for that in terms
of getting students into those jobs.
29:35 - EG
I put a couple of links in the chat. One was the Kaggle. The
other is I got my master's at Georgia Tech. They actually have an innovation
lab where they actually partner with. Google or. Sometimes health care because
they do a lot of health care innovation. But one of the things that they do is
they create a An environment where the students as they are going through their
classes are innovating and solving problems for companies.
30:19 - DB
Okay, any other comments at this point? If not, I asked an
AI to, I gave them my problem. I'm running a discussion group session on the
topic, how to find a job in AI. What are some useful discussion questions to
present?
30:50 - DB
I've got more than we have time for. I've got a lot of them.
Here's one of them. Any comments?
31:04 - HA
Can you put the link to this in the chat so we can have a
copy of it? I didn't.
31:11 - DB
I guess I could probably find the chat. Well, nowadays, you
can't share chat links anymore, I discovered. But anyway, you can get to this
page, which has the essence of it. I mean, I got a ton of questions here, all
generated by AI.
31:26 - DB
So yeah, I mean, and to get to it, you just go to today's
study group, just today's minutes, and then click how to find a job in AI. And
that'll take you straight to this page, which I posted yesterday on this blog.
31:51 - Unidentified Speaker
Okay. Any comments on this question? How about this one?
Take a minute to and see if you have any comments.
32:11 - EG
I have a comment about the previous one. You cannot be an AI
person without being a data engineer. The two are synonymous.
32:24 - EG
Is built on how well you're able to pull together data.
That's why I'm able to oscillate. Right now, I'm a data architect for UPMC to
build the infrastructure to support their AI systems, their whole
modernization. They recognize that you can't do AI without building the
foundation. Data engineering, that's a cornerstone.
33:00 - DB
All right.
33:04 - MJ
The second question is a great question that if you're
applying for a job you should be prepared to take to get, which I guess is what
this is, right? It sounds like to me an of your question, D.
33:22 - DB
I mean, this is the problem with AI. They'll tell you
anything, but how do you know they're not hallucinating? It's a good first
step, but you as the AI user need to be able to verify and own it and verify
it.
33:39 - Unidentified Speaker
Yeah.
33:41 - MJ
It's an important skill to be able to critique your own AI
responses to validate them. Of course, you can always ask for sources. You can
ask for citations. And when you get them, by the way, you should check them.
Oh, absolutely.
33:54 - DB
Because I have done that before.
33:56 - MJ
And where I went was not, it didn't validate. Oh, yeah.
34:01 - DB
When you ask AI for citations, you're guaranteed that 20% of
them, them 25% are going to be bogus. You have to check them. Dr.
34:10 - EG
Brilliant, previous to AI in the machine learning aspect.
When we did a problem, we would test it. We would check the variance inflation
factor. We'd do ANOVA tests. We'd do Pearson tests. We would interrogate it.
We'd assume the answer was wrong until everything proved out it was right. We
don't have the same luxury.
34:41 - DB
All right, is anyone who hasn't had a chance to speak up
have any comment or question at this point?
34:49 - HW
Well, I have a question. For me, I think AI is magnified. So
if the user, as so we should educate our student capacity, so get to have a
creative ideas, also create creative thinking capacity to judge what the AI
said is trustable, accountable or not. So for us, I have been trying to learn
and I took a Udemy course. I find if we find the right way, the AI really can
do many, many things for us. So if the user, ourselves, do not have idea no
matter which is useless for us. So we need thinking about, so for student to
find the job in AI, we maybe need to concentrate on this capacity, this aspect
to, you know, to make students become a powerful. So when the people, I mean,
the company interview those students, they will find those students has the
idea and also has the capacity to wisely and effectively use AI. That way, I
think our students will be very, very strong in the job market, no matter which
field Like last week, we have the, how to say, the company people show us the
software. I think those software are very useful for us, like to write the
chemistry things information. So I think maybe we, that's just my thought.
Okay, thank you.
36:48 - DB
Anyone else who hasn't had a chance to speak up or very
little have a comment at this point?
36:54 - VK
Yeah, I agree with Rick on the last one on the time. Most of
the time it'll take on data engineering, 60 to 70%. And if that needs to be
done right, the cleaning preparation and everything before you can do the
actual AI thing. So it's very important to do that, right? The data modeling
and all that, so that sometimes people are not looking look at AI and just go
directly for that. But there's a lot of work before that. On the second one,
companies are getting into AI, trying to get quicker results and all. But
unlike doing other things that's not AI, like for testing, everything, the
duration is much more in which stakeholders are not probably prepared. And
that's a challenge, however, because to accurate, you have to do all the
validations and testing, and a lot of iterations goes into building their
product. So setting up the stakeholder expectations is very, it'll be a big
challenge, actually, because a lot of companies are doing it, but the success
is not much, so it's just but everybody is doing it.
38:31 - DB
So V, do you think … is going to be looking for AI people or
like in your group or anything like that? Or what do you think?
38:42 - VK
Not right now. So it all depends on how gets. The business
case is very important. So we always Like somebody mentioned, most of the time
they have their own rules and it's been rule-based. They come out with some
kind of understanding on the outcome and then they think about AI. It's still
not ready. I'm working on it. I'm in a similar role of Eric here, trying to set
it up. Currently, we don't need. There's no requirement, but I'm hoping there
will be something in future.
39:28 - DB
OK. Anybody else have a comment at this point?
39:34 - Unidentified Speaker
All right. Well, let's look at the next question.
39:39 - DB
Take a minute to Read it, and then let us know if you have a
comment. Thought.
39:54 - EG
Sorry to dominate Dr. Brilliant, but I just put that into
chat. Generalists in AI, it's hard. You have to have a specific area and AI
secondary.
40:12 - DB
Yeah, I'm sort of getting from what M was saying too, that
it was not necessarily getting a job specifically doing AI, it's getting a job
in whatever your field of expertise is and being able to use AI to do it
better. You know, it's like companies are going to be They're going to be
looking for many of the same roles, but they're going to hire people who can
use AI to help with it. At least that's what I was getting.
40:46 - DB
Any other comments on this one?
40:52 - DB
All right, how about this one?
40:54 - EG
At UPMC, you're using UPMC hardware and they block you from it.
41:18 - EG
They have AI, but all of their AI is all local to them. So
no PHI or no PII leaves UPMC.
41:29 - DB
I mean, I use AI, you know, in my job.
41:40 - DB
Oftentimes I just use my free accounts, you know, and UALR
is not authorized. Me to do it. Fortunately, I don't think they have a problem
with it, but some companies would, perhaps.
41:56 - DB
There's no going back. You can't tell people not to use AI.
That'd be stupid, right? But you can worry about confidentiality of information
and things like that. As an organization, you can worry about things like that.
42:16 - Adam
Well, you might have a separate department for red teaming
strategies to try to kind of counter. Where innovation might go South, you
know?
42:26 - Unidentified Speaker
OK.
42:31 - DB
Um? Any other thoughts on this?
42:34 - HW
Then you have questions for you and the dealer has a
limitation for faculty or to use a certain kind of AI software or something?
42:46 - DB
I don't know. I mean, we're sort of limited by FERPA, you
know, Federal Education Rights and Privacy Act, so we couldn't go to a, I think
we'd be legal to go to a general purpose, general AI on a free account and say,
you know, talk about a particular student by name, because that might violate
FERPA, I would suggest. So you do bioinformatics,
43:17 - HW
So you do data analyze, for example, I do a gene profiling
data. So could we use some certain AI software to analyze our data?
43:33 - DB
Well, all right, here's what I know about, if you log into
Gemini from UALR account, Gemini will tell you right off the bat that any data
you share from a UALR will not be used for their training purposes. I mean, I
can show you. It's worth training. That's OK.
43:54 - HW
As long as I know it, that's good enough.
43:57 - DB
Yeah, well, I mean, I just noticed it.
44:02 - DB
Didn't say anything.
44:05 - EG
And I was at Florida Cancer. We did genetic. Profiling, but
as long as you did not add any identifying information to the individual.
44:18 - HW
OK, complete value to feel a bit cautious about where your
data goes. Those days as I I don't have. Yes, too picky or personal, Some
technical. Here's what I was looking for.
44:38 - DB
I mean, you probably can't even Read it. It's printed. It
says your University of Arkansas at Little Rock chats aren't used to improve
our models. OK. OK. And the others don't do that because I assume this is
because UALR uses Google, has a giant Google account for our email and all this
other stuff. And this is part of it.
45:06 - IU
Somebody in the meeting yesterday that the DC STEM Lunch and
Learn was very cautious about how far that Gemini guarantee went because Gemini
is owned by Google and Google has their own overarching statements about, you
know, what they can do with your data and that kind of stuff. There was some
real concern about just how far that Gemini guarantee went. Yeah, I can see
that.
45:49 - DB
But on the other hand, you know, as a user, you're sort of
covering your, you're doing your due diligence, like even if even if even if
Gemini I did use your chats to improve their models or did break
confidentiality, you could say, well, by reading the statement, I am, as a
professor, doing my due diligence.
46:14 - Unidentified Speaker
I don't know if that addresses your concern.
46:19 - IU
Yeah, I mean, I agree that we, as users on the faculty end
can't, I guess we could, but you know, practically are not going to be able to
go to the top level at Google and demand statements about what they actually
are going to do and aren't going to do with our stuff. And I think at some
point we have to say, look, here's what I was told and assured when I looked at
Gemini. But I also, again, several of the folks in the room who know a whole
lot more about this than I do, were cautioning folks not to take that as an
ironclad guarantee that anything you stuck into Gemini would never be used for
training their AI. Interesting.
47:21 - HW
I see you know I'm a plant biologist so my gene profiling
data is from a crop. So as long as the data could keep private before I publish
the paper, that's OK.
47:43 - EG
But what if your data was innovative and you're using that
to start a company? They train your data and and make it part of their next
model. Now your company information, your private information is now public. I
see.
48:01 - HW
I never think about that part. Thank you. Yeah, I mean, it's
the same as you have the same issue.
48:10 - DB
You know, you want to talk about your ask for medical
questions on an And, you know, next thing you know, maybe you're going to start
getting targeted ads from companies.
48:25 - DB
Hopefully not, but big field is a big problem. Yeah, the
whole privacy thing. All right, let me go back to the...
48:37 - DB
Okay. Thank you.
48:40 - HW
Yeah, no problem. All right, next question.
48:46 - DB
GitHub.
49:13 - DB
If you want to put stuff on GitHub about problems I solved
with AI, you'd use the github.io site, right?
49:21 - EG
No, I just use the regular GitHub because at that point, it
shows you the coding. To do AI, you need to know Python, you need to know SQL,
you need to know these languages, you need to know how to build out the
prompts, you need to know how to load the models. You need to be able to
demonstrate that through code, that you're adept at it. And the only way you
can prove that is for them to look at your code to say, yes, you're doing this.
It looks like you know what you're doing.
49:58 - DB
I interrupted someone a minute ago. What was Anybody?
50:10 - Unidentified Speaker
OK.
50:11 - DB
I have a quick question.
50:12 - MM
I'm sorry. So this is my area, career readiness. And of
course, I don't know about AI. So this one is asking the question of switching
traditional resume focus to tools I know. When they're talking about a
portfolio, what would you recommend creating for students begin doing versus
the typical resume format that I work with students on?
50:40 - EG
Students that I mentor who are going into this market, I
tell them to put all their schoolwork in GitHub so that way they can
demonstrate their progression and they have a history. OK.
50:56 - MM
So you said it, but I'm learning the language. So I get it.
OK, thank you. I have a thought on this.
51:01 - DB
So a lot of students, they'll say, well, I have skills in
SQL. I know Python. I know JavaScript.
51:13 - MM
And they'll also list projects. I did this project. I did
that project.
51:18 - DB
I think the project idea could kind of take root here in the
sense that that's a good opportunity for students to say not only what the
project was, but how they used AI in it. You know, I did such and such a, you
know, I marketed hamburgers to the Little Rock market and used AI to, to
enhance, you know, advertising effectiveness by 50%. I'm just bullshitting, but
you know, that sort of thing.
51:50 - Unidentified Speaker
So, you know, people already list projects they've done on
their resumes sometimes.
51:54 - DB
And that's a good opportunity for them to mention the AI
content of their solutions.
52:03 - DB
How about I learn SQL on the job by using an AI to help
maintain existing SQL code? How about that?
52:24 - DB
I'm talking too much. Any other comments from you all on
this aspect, this question?
52:35 - DB
Okay, next question.
52:50 - Unidentified Speaker
I'm sorry, Dr.
52:51 - EG
Brilliant, I am taking over. But you can't have AI build
your application. It will be ugly. It doesn't know how to, it doesn't have the
years of experience of knowing how to code, how to create functions, how to
create single responsibility, how to basically manage code, create a lot of the
processes. Now, if you're an experienced programmer, you can create the prompts
to know how to set AI to do it. But to have AI replace your programmers, no.
Junior programmers, yes, because they don't know what they're doing.
53:36 - DB
Okay, anybody else?
53:44 - DB
Okay.
53:49 - Unidentified Speaker
Recommended, I don't know what that means.
53:55 - DB
Oh, here, no, no, no. Yeah, does anyone know about the
Indeed AI tracker? Anyone know anything about it?
54:03 - HA
Yes, Indeed will automatically track if you have AI jobs
that you're seeking. It will track and send it to you as well.
54:13 - HA
So it tells about the AI jobs out there, but it will also,
if you're interested in AI, because I get them, it will match your skills to a
potential AI jobs there.
54:26 - DB
Okay, so that's a good tool for people to use then. Yes,
yes.
54:31 - HA
And that's true for, I guess, all of these.
54:34 - DB
I have information about this one.
54:42 - Unidentified Speaker
I had premium.
54:43 - EG
It wasn't worth it.
54:49 - HA
The three here, I think, indeed, has more of the tracking
and AI than LinkedIn. And Glassdoor is what it says here, more about the
culture, more about reading, about how the company works, operates, its
employees, et cetera.
55:04 - DB
Well, that's really important for prospective employees to
know about.
55:08 - HA
Yeah, there is another one called The Vault that is very
popular, V-A-U-L-T. Yeah.
55:15 - DB
I was thinking Glassdoor, to my understanding, is a Japanese
company. And I was afraid that if an American company tried to do a Glassdoor,
56:01 - HA
The vault is more like, how is the culture? How do they
treat you? How is the peership? So I, you know, there is some negative as well,
right? In any social media, there's plenty of negative, but you just have to
balance it out yourself, you know, but it's interesting to see what's trending,
you know, for a company, if they say they overwork, you, et cetera, when you
Read 10, 12, 15 times about that, you're like, well, maybe that's not for me.
Or if they say they'll overwork you, but they also reward you with, you know,
10, 15% raises or something, then that may be something you're interested
perhaps, you know, for a few years. So I think the vault is more like that.
56:44 - DB#+#HA
Doesn't Glassdoor do the same thing?
56:47 - Unidentified Speaker
Yeah.
56:48 - HA
Yeah. I just, I just, you said there was a lot of dirt in
it. I think the vault, um, I haven't been in for a while and the vault I check
from time to time. I just think the vault is not, it does have its dirt, but
it's not unhygienic.
57:04 - DB
Well, I mean, I guess what I mean by dirt, it can be, you
know, if people can go to a location and hear about how terrible a company is
to work for, well, that company may not want that website to say, to stay in
business.
57:23 - HA
And of course, yeah, yeah, there was a even somebody who
shut down a business for outing him. I forget who I think was speaking. I
forget who he was. But at any rate, yeah, so so there's these these sites run
always that danger. That's true.
57:41 - DB
So all right. Anyway, we're kind of out of time. I have a I
mean, I have a million questions here, all generated by AI.
57:50 - DB
We could continue the discussion next time if you like about
these questions or we can do something else next time. Any thoughts,
preferences?
58:02 - HA
So my being a graduate student at UALR is that, and I have
seen other universities what they do in terms of placement. I think AI is
demanding a different approach to.
58:21 - HA
Some of the things we talked about going to hackathons and
that's all great, but also in terms of projects, showcasing them, because we do
projects at UALR, right? How do you showcase that to the market? Then the other
one is about, like I was mentioning and M added in is, Miami Dade College that
pairs, you know, and it's not every student has that spared. It's not possible
to do that, but at least have, you know, some kind of matchings that get some
mentorship that later on people can help in navigating the market, right? So
there's some things that can be done, I think, differently, but that could
actually reap some great benefits for the students. So I don't know if that's
something you all will consider at some point, I would have benefited from it.
And I'm feeling that I will benefit still from some of this, but I'm hopeful
that in the future, this will be considered.
59:26 - EG
At Georgia Tech, your final semester, you do a project with
a previous graduate. The graduate will outline problem domain and they will
mentor them through it and they will spend a semester solving a business
problem. It could be a fake one, it could be a real one but they're solving a
business problem. They take what they are doing and then they put it on their
public website so that way people could look innovation, their creativity in
solving problems. That may be something you may want to consider.
1:00:21 - DB
All right, well. Thanks, everyone, for joining, and we'll
I'll think of something to do next next week for sure.
1:00:34 - DD
Thank
1:00:36 - HW
Good night.
1:00:37 - HA
Thank you. Thank you everyone.
