Latest / Go News To Go: More Good News / Elizabeth Kennedy TALKS ABOUT AI, EARLY DETECTION OF DISEASE AND COMPUTER VISION

Transcript

Elizabeth Kennedy TALKS ABOUT AI, EARLY DETECTION OF DISEASE AND COMPUTER VISION · Go News To Go: More Good News

• 0:04
  Speaker 1: Kyle Paul Slack is good News and work for peace,
• 0:07
  Speaker 1: Peace piece piece. Elizabeth Kennedy, how are you?
• 0:11
  Speaker 2: I'm doing well? How are you doing?
• 0:14
  Speaker 1: I met Elizabeth and her husband and our good friend
• 0:18
  Speaker 1: Jean Fish's reading of Satchel Page in Lincoln Center. Not
• 0:25
  Speaker 1: in Lincoln Center and Key Carnegie Hall. Okay, practice practice practice,
• 0:31
  Speaker 1: and so in any case, she is fascinating. We just
• 0:35
  Speaker 1: interviewed her husband and both of them they're unbelievable of
• 0:39
  Speaker 1: the good things that they've been involved with and continue too.
• 0:43
  Speaker 1: And so give us a little bit of your background
• 0:46
  Speaker 1: and some of the nice positive efforts that you've been
• 0:50
  Speaker 1: doing with in your life that have made you happy.
• 0:54
  Speaker 2: Sure.
• 0:55
  Speaker 3: So I started my career as an engineer at General Motors,
• 1:00
  Speaker 3: and I used to manage the design and installation of
• 1:03
  Speaker 3: assembly plants with a focus on robotic welding.
• 1:07
  Speaker 2: And that's how I started my career.
• 1:10
  Speaker 3: And then I realized that I wanted to be on
• 1:13
  Speaker 3: the business side of decisions and not just on the
• 1:16
  Speaker 3: engineering side.
• 1:17
  Speaker 2: And so I went back to grad school and got
• 1:18
  Speaker 2: my MBA at Duke and then Duke Duke.
• 1:25
  Speaker 1: Where's my Duke hat? My hand goes to Duke Duke really, duke,
• 1:31
  Speaker 1: I have to key, I wear that haty because my
• 1:32
  Speaker 1: daughter's spending a trillion billion and I got it promoted
• 1:36
  Speaker 1: at every moment.
• 1:40
  Speaker 2: Do you want to go get your hat quick and
• 1:41
  Speaker 2: come back?
• 1:43
  Speaker 1: Maybe? All right, go ahead.
• 1:47
  Speaker 3: Well, after business school, I worked at Craft Foods doing
• 1:51
  Speaker 3: brand management, and then I moved back into auto at Nissan,
• 1:56
  Speaker 3: and then eventually I made my way to the tech
• 1:58
  Speaker 3: industry and I've and doing product management in the technology
• 2:02
  Speaker 3: industry ever since.
• 2:03
  Speaker 1: All Right, what's product management? Somebody makes a product, then
• 2:06
  Speaker 1: you make sure you walk it around and talk to it.
• 2:09
  Speaker 2: Yeah, so not exactly that, but they're heading there.
• 2:14
  Speaker 1: No, it's weird guy. I'm a weird guy. Go on.
• 2:18
  Speaker 2: I like weird people, so it's a good match.
• 2:21
  Speaker 3: I So a product manager can be defined in so
• 2:24
  Speaker 3: many different ways. But how I talk about it is
• 2:27
  Speaker 3: that product managers we solve problems. We solve our customers problems,
• 2:33
  Speaker 3: and so whether it is helping to make a purchase
• 2:38
  Speaker 3: go faster and quicker, or to make it a better
• 2:43
  Speaker 3: decision on what you're going to be buying. But there's
• 2:46
  Speaker 3: product managers all over the world, and what we do
• 2:49
  Speaker 3: is we see a problem, we come.
• 2:51
  Speaker 2: Up with a solution for it, and.
• 2:54
  Speaker 3: Then we work with engineers to build it and bring
• 2:57
  Speaker 3: it to market.
• 2:59
  Speaker 1: Okay, all right, I don't want anyone ask this question?
• 3:05
  Speaker 1: Maybe it Should I not ask it? You told me?
• 3:07
  Speaker 1: Should I not ask this question?
• 3:09
  Speaker 2: I'm up for any question. I may not answer it
• 3:11
  Speaker 2: depending on what the question.
• 3:12
  Speaker 1: Tell me if I should answer it, if you should,
• 3:14
  Speaker 1: if I should ask it, say well, let's get to it. Okay,
• 3:19
  Speaker 1: is this a I'm always even say AI? So the
• 3:25
  Speaker 1: do you still have a job?
• 3:28
  Speaker 2: Do I still have a job?
• 3:29
  Speaker 1: Yeah? This AI because AI is uh is walking? Is
• 3:34
  Speaker 1: AI walking the product around?
• 3:36
  Speaker 2: So as of right now, AI is helping to make
• 3:40
  Speaker 2: my like So let me take a step back. AI
• 3:45
  Speaker 2: is a summary. It can be a whole lot of
• 3:48
  Speaker 2: different things.
• 3:49
  Speaker 3: And the AI that most people are using and thinking
• 3:52
  Speaker 3: about right now, which is what's called large language models,
• 3:56
  Speaker 3: but AI can mean so many different things. And I
• 4:02
  Speaker 3: have worked in a form of AI that's called computer
• 4:06
  Speaker 3: vision for years. And so what the what you do
• 4:11
  Speaker 3: to create an AI tools? You have to train a
• 4:14
  Speaker 3: model using data to train a model to ultimately be
• 4:18
  Speaker 3: used to solve some other problems. And so what's happened
• 4:22
  Speaker 3: with AI recently is multiple large language models have come
• 4:27
  Speaker 3: out of into the public use, and they are helping
• 4:31
  Speaker 3: me as a product manager because I am able to
• 4:36
  Speaker 3: do the same things I used to do, but faster,
• 4:40
  Speaker 3: because I can go into an AI tool and instead
• 4:47
  Speaker 3: of me having to write a document from scratch, I
• 4:52
  Speaker 3: can go in put in some information to tell the
• 4:55
  Speaker 3: link the large language model to write a document for me,
• 4:59
  Speaker 3: and then it creates it and I just have to
• 5:01
  Speaker 3: go in and fine tune it.
• 5:03
  Speaker 1: So it's not smart enough to know that it needs
• 5:05
  Speaker 1: to do that itself.
• 5:08
  Speaker 2: Well, it's you have to tell it what to do.
• 5:11
  Speaker 1: Oh right, Yay. That means some jobs are still going
• 5:15
  Speaker 1: to be here.
• 5:16
  Speaker 3: Yay.
• 5:16
  Speaker 1: Somebody's going to tell somebody something.
• 5:18
  Speaker 2: Yeah.
• 5:19
  Speaker 3: And I look at AI as just like with every
• 5:23
  Speaker 3: new technology that comes out in the world.
• 5:26
  Speaker 2: There's ways that.
• 5:27
  Speaker 3: You can use it in a positive way, and there's
• 5:29
  Speaker 3: usually ways that you can use it in a negative way.
• 5:32
  Speaker 3: And it's making sure that we as a society are
• 5:36
  Speaker 3: using it in a responsible manner.
• 5:39
  Speaker 1: Yay, all right, I'm in on that. So and so
• 5:45
  Speaker 1: to today, what kind of work did you do today?
• 5:48
  Speaker 2: Well, I'm not actually working right now.
• 5:51
  Speaker 1: That's a great job. I think my job.
• 5:55
  Speaker 2: Right now is looking for another job.
• 5:57
  Speaker 1: Oh, oh, okay, you're in between I.
• 6:00
  Speaker 3: Am in between jobs. Yes, because I had a health issue.
• 6:05
  Speaker 3: I've recovered from that health issue, and so now I'm
• 6:07
  Speaker 3: ready to get back to work.
• 6:08
  Speaker 1: Oh okay, all right, good and we pray your continue
• 6:12
  Speaker 1: to have this. It's history resolved.
• 6:18
  Speaker 2: I appreciate it.
• 6:19
  Speaker 1: That's the key of it, all right, no doubt about it.
• 6:22
  Speaker 1: I hope it has nothing to do with microplastics.
• 6:25
  Speaker 3: It has nothing to do with microplastics. It had if
• 6:30
  Speaker 3: it was microplastics, then Phil would have already cured me.
• 6:34
  Speaker 1: He would have solved it. Okay, Yeah, Because I was
• 6:36
  Speaker 1: just going to say, I know a good guy and
• 6:38
  Speaker 1: knows a lot about this stuff, and so it happens.
• 6:41
  Speaker 1: I guess it's our husband. So and then well you
• 6:44
  Speaker 1: just about a half an hour learning about these little
• 6:47
  Speaker 1: buggers and so.
• 6:50
  Speaker 3: So I think what I could share with you are
• 6:54
  Speaker 3: just examples of how computer vision, which is the type
• 6:58
  Speaker 3: of AI that I have worked in the great thank you,
• 7:01
  Speaker 3: how that has been used for good. So one of
• 7:06
  Speaker 3: the things that when we first met, that I talked
• 7:08
  Speaker 3: a little bit about is that computer vision is being
• 7:11
  Speaker 3: used to detect illnesses faster than we used to be
• 7:16
  Speaker 3: able to detect illnesses. And so, for example, uh, when
• 7:21
  Speaker 3: an X ray is taken or an MRI is taken
• 7:25
  Speaker 3: of let's say it's somebody's knee. I what has happened
• 7:31
  Speaker 3: in the past is a human will look at that,
• 7:35
  Speaker 3: uh X ray or MRI and they will interpret the
• 7:39
  Speaker 3: results and the radiology.
• 7:43
  Speaker 1: Brother in law, but oh really, let's continue on. He
• 7:46
  Speaker 1: might be you might be firing him. All right, already
• 7:50
  Speaker 1: quit the paperwork. Did him in the amount of paperwork
• 7:54
  Speaker 1: he had to do? Forget it. I'm not doing any
• 7:57
  Speaker 1: more paperwork. I don't even have a chance to look
• 7:59
  Speaker 1: at a X ray. Sorry.
• 8:04
  Speaker 3: Well, what what computer vision has done is it can
• 8:08
  Speaker 3: see better. A computer can see things at an image
• 8:13
  Speaker 3: resolution that is that.
• 8:16
  Speaker 2: The human eye can't see.
• 8:18
  Speaker 3: And so what when the past, what would happen is
• 8:20
  Speaker 3: you'd have to wait until the disease had progressed.
• 8:23
  Speaker 2: So that the human eye could see it and then
• 8:26
  Speaker 2: diagnose it.
• 8:27
  Speaker 1: This is big. This is very very big. This is
• 8:30
  Speaker 1: bigger than big.
• 8:31
  Speaker 3: This is yeah, And so now what can happen is
• 8:35
  Speaker 3: the a computer can look at the results of an
• 8:38
  Speaker 3: MRI or an X ray and detect the cancer before
• 8:43
  Speaker 3: the human could. And then that means you can treat
• 8:46
  Speaker 3: it earlier, which means a higher chance of survival and
• 8:50
  Speaker 3: full recovery along with it.
• 8:53
  Speaker 1: This is the biggest This is really big. Is this
• 8:56
  Speaker 1: is the word? Because I just was an some other
• 9:01
  Speaker 1: investment meeting and this was for Alzheimer's and they were
• 9:06
  Speaker 1: talking about markers, markers that you could mark, and I
• 9:09
  Speaker 1: sat next to somebody at the Arthritis Foundation then a
• 9:12
  Speaker 1: surgeon on the plane on the California So I was into,
• 9:15
  Speaker 1: you know, marker is that you could possibly potentially know
• 9:18
  Speaker 1: whether you're going to get Alzheimer's at a later stage
• 9:22
  Speaker 1: in your life. Okay, so now this is I guess
• 9:25
  Speaker 1: do you have to know that you have a little
• 9:28
  Speaker 1: cancer for it to then say, you know, well.
• 9:35
  Speaker 2: The patient doesn't have to know that they have cancer.
• 9:38
  Speaker 2: It's more of.
• 9:39
  Speaker 3: Their symptoms that would cause a doctor to have their
• 9:44
  Speaker 3: patient get an MRI or an x ray or something
• 9:48
  Speaker 3: along those lines.
• 9:49
  Speaker 1: And now they get it. And now you have this
• 9:51
  Speaker 1: technology that can they take the x ray the MRI
• 9:56
  Speaker 1: and they learn from seeing the MRI, or they give
• 10:00
  Speaker 1: the MRI instead of putting it on a like a film,
• 10:06
  Speaker 1: uh you know thing up on the wall and going
• 10:10
  Speaker 1: there's your knee. So, uh so you when do you
• 10:15
  Speaker 1: put this into a machine? This is what you're talking about.
• 10:18
  Speaker 1: How does it how does it actually accomplish the task?
• 10:22
  Speaker 3: Well, what happens is that so similar like how I
• 10:25
  Speaker 3: was saying that large language models work. When you build
• 10:29
  Speaker 3: a it's you're building a computer vision model, and so
• 10:33
  Speaker 3: what you do is train it based on humans looking
• 10:39
  Speaker 3: at X rays or MRI results, and you have to
• 10:42
  Speaker 3: train it with lots and lots and lots of data
• 10:46
  Speaker 3: before it can ever be used to actually detect. And
• 10:51
  Speaker 3: usually what happens is in the especially in the early
• 10:54
  Speaker 3: days and even depending on how let's say a hospital
• 10:57
  Speaker 3: or doctor wants to use it, you use what's called
• 10:59
  Speaker 3: a human in the loop. And what that means is
• 11:03
  Speaker 3: that you aren't just taking what the computer says as
• 11:09
  Speaker 3: the finished result. A human looks at what the computer
• 11:14
  Speaker 3: is saying the result is and can say yes or
• 11:17
  Speaker 3: no that it's correct. And that's the way that you
• 11:19
  Speaker 3: continuously train it until the model is so good that
• 11:24
  Speaker 3: it's performing better than a human could.
• 11:27
  Speaker 1: So so, but you know, once you do this, it
• 11:31
  Speaker 1: seems to me then the deal is done. I mean, yeah,
• 11:36
  Speaker 1: but you.
• 11:36
  Speaker 3: Also want to keep training it. You also want to
• 11:38
  Speaker 3: keep training it so that it can get better and
• 11:40
  Speaker 3: better and better. At some point, you get to a
• 11:42
  Speaker 3: point of diminishing returns, where the work to train it
• 11:47
  Speaker 3: is no longer actually increasing.
• 11:49
  Speaker 1: Yeah yeah, yeah, yeah, yeah yeah. Check this out. This
• 11:53
  Speaker 1: is unbelievable. This is super wow. Science, science and moving
• 12:02
  Speaker 1: our lives forwards.
• 12:04
  Speaker 2: Yeah.
• 12:04
  Speaker 3: Yeah, And I know I see a lot of people
• 12:08
  Speaker 3: talk about their AI and how it's scary and there's
• 12:12
  Speaker 3: things that you have to worry about. It also comes
• 12:16
  Speaker 3: down that you have to with any technology, you have
• 12:19
  Speaker 3: to use it responsibly.
• 12:21
  Speaker 1: Yeah. Yeah, you're a good one. You're a good one. Yeah,
• 12:26
  Speaker 1: you're bringing up a good, good, good point. So we
• 12:29
  Speaker 1: have a few more minutes. Tell me if you if
• 12:32
  Speaker 1: if I could give you the job that you want,
• 12:35
  Speaker 1: what job would that be.
• 12:38
  Speaker 3: So I am a product manager and I love a
• 12:41
  Speaker 3: being a product manager, and I earlier in my career
• 12:45
  Speaker 3: I had the opportunity to work in fashion technology and
• 12:50
  Speaker 3: I absolutely loved it well, and so being able to
• 12:56
  Speaker 3: work back in that space where I'm helping people find
• 13:01
  Speaker 3: either the right size of an item that they want
• 13:05
  Speaker 3: or give people access to clothing what they wouldn't have
• 13:08
  Speaker 3: had previously, or making it so that you don't have
• 13:12
  Speaker 3: to return items. There's so many things that we can
• 13:16
  Speaker 3: do within the fashion in the apparel space, and this
• 13:21
  Speaker 3: is for me. It's an area that I find fascinating.
• 13:26
  Speaker 3: And yes, I am a bit of a I'm a
• 13:29
  Speaker 3: bit of a clothes.
• 13:31
  Speaker 1: Hog, a fashionista one hundred. I saw you in the
• 13:37
  Speaker 1: streets on seven thousand. You're right in front of Carnegie
• 13:40
  Speaker 1: Hall Cohen the Jeans party, and you had a code on.
• 13:44
  Speaker 1: I'm a fashionista. My son groogup shoe business. I love
• 13:47
  Speaker 1: the we call them schmantas. We love clothes. Best dressed
• 13:50
  Speaker 1: in high school. I love that. And because she had
• 13:54
  Speaker 1: ms Kennedy had a red what describe your coat?
• 14:00
  Speaker 3: The one I was wearing that that evening was a
• 14:03
  Speaker 3: like full trench coat that was bright pink and brown.
• 14:09
  Speaker 1: It was gorgeous. This couple was They were like, I said,
• 14:14
  Speaker 1: this is New York's best, is what I said. And
• 14:17
  Speaker 1: then they said, but we're from Detroit. And then I
• 14:22
  Speaker 1: know about Detroit. I've been in Detroit and I'm a
• 14:25
  Speaker 1: Motown dude, so and.
• 14:28
  Speaker 2: Well, Motown. I listened to Motown still to this day.
• 14:31
  Speaker 1: I had a fashion. I was with a whole bunch
• 14:33
  Speaker 1: of brothers last night from from Motown and just talking
• 14:38
  Speaker 1: talking this, you know, the the scene, and they were
• 14:41
  Speaker 1: all dressed and peccable. You know, I had a regularly
• 14:44
  Speaker 1: ugly gray suit on. You know, you know I was
• 14:47
  Speaker 1: like the nerd amongst the cool dudes.
• 14:51
  Speaker 2: Well, I have a special place for fashion in my
• 14:56
  Speaker 2: heart that I love it.
• 14:57
  Speaker 3: But one of the reasons why I would love to
• 15:00
  Speaker 3: get even more involved in it and continue to work
• 15:02
  Speaker 3: in it is that for so long, some of these
• 15:05
  Speaker 3: this fashion was only attainable for people who looked a
• 15:09
  Speaker 3: certain way or had a certain body shape that was
• 15:12
  Speaker 3: skinnier than buyout. And I love to be able to
• 15:16
  Speaker 3: expand it so that it doesn't matter what body shape
• 15:19
  Speaker 3: or size you have, that you.
• 15:21
  Speaker 2: Can have access to fashion.
• 15:23
  Speaker 3: That's good because when I dress nice and random people
• 15:27
  Speaker 3: on the Seventh Avenue compliment me, it makes me feel good.
• 15:31
  Speaker 1: It's not that about it. Hey, you are what you
• 15:33
  Speaker 1: look you look Billy Chrystly used to say, you look marvelous.
• 15:39
  Speaker 1: All right, I got to run to another something. You're fantastic.
• 15:43
  Speaker 1: You should be coming to New York this week. It's
• 15:44
  Speaker 1: fashion Week.
• 15:46
  Speaker 2: And I've never been to fashion Week. I would love
• 15:48
  Speaker 2: that someday.
• 15:49
  Speaker 1: You should come back this week look into where things are,
• 15:54
  Speaker 1: and because there's a fashionistas and you know the people
• 15:57
  Speaker 1: that run these companies that probably need people someone like
• 16:01
  Speaker 1: you who can help them in a totally different way
• 16:05
  Speaker 1: than most of the other people. They're mostly designers and
• 16:08
  Speaker 1: you know models and things. Well, it's a sort or
• 16:13
  Speaker 1: just go online and see what's going on and find
• 16:15
  Speaker 1: out who's running the shows.
• 16:19
  Speaker 3: But yeah, I unfortunately can't go to New York this week,
• 16:22
  Speaker 3: but I will keep that in mind, and I really
• 16:25
  Speaker 3: appreciate the suggestion because I can always go next year. Yeah,
• 16:29
  Speaker 3: Or you can go online and see what's happening right now.
• 16:31
  Speaker 1: We go online, see what's happening in fashion week. You
• 16:34
  Speaker 1: know the key I always I help kids get jobs
• 16:39
  Speaker 1: in my whatever is the first thing I tell them is,
• 16:42
  Speaker 1: you know, look in the trades and say who's doing what,
• 16:45
  Speaker 1: and then write a letter to the one who's doing
• 16:48
  Speaker 1: something and say, you know, I really think that that's
• 16:50
  Speaker 1: a great idea to put gold buttons on the jacket,
• 16:55
  Speaker 1: and that person who's going to see that you've made
• 16:58
  Speaker 1: an effort and smart. Usually the kids I tell them
• 17:01
  Speaker 1: to try to get an internship with the president.
• 17:04
  Speaker 3: You know, every single job I've gotten has been through networking,
• 17:08
  Speaker 3: and in many cases there wasn't even a job posting
• 17:10
  Speaker 3: for it yet.
• 17:11
  Speaker 1: There you go, all right, rock and roll. Do you
• 17:14
  Speaker 1: have a website too?
• 17:16
  Speaker 2: I don't have a website myself. I have a LinkedIn.
• 17:19
  Speaker 1: That's the that is what's your LinkedIn?
• 17:23
  Speaker 2: Uh? It is? It's long because to me, and.
• 17:29
  Speaker 1: We'll put it in the.
• 17:33
  Speaker 2: Text the best way or email.
• 17:35
  Speaker 1: Either one is cool, We'll be fine.
• 17:39
  Speaker 2: I will email it to you.
• 17:40
  Speaker 1: Then that that does two I can do. I don't
• 17:46
  Speaker 1: send that to me on AI, all right or anything else.
• 17:49
  Speaker 2: I won't do that. Well, have a wonderful rest of
• 17:52
  Speaker 2: your weekend.
• 17:53
  Speaker 1: There you go, all right, it's President's weekend.
• 17:55
  Speaker 2: Yes, I have to work on Monday. Still though, Oh okay,
• 17:59
  Speaker 2: all right, got job to find.
• 18:01
  Speaker 1: Oh okay, there there you've got energy and determination. That's
• 18:07
  Speaker 1: what it takes. Ah, thank you very much. Bye bye
• 18:11
  Speaker 1: bye