Latest / Brian Thomas / Empower U - Colin Madine - Technology and AI: What’s Coming Our Way?

Transcript

Empower U - Colin Madine - Technology and AI: What’s Coming Our Way? · Brian Thomas

• 0:01
  Speaker 1: Eighth five Here at fifty five KRCD talk station, Brian
• 0:04
  Speaker 1: Thomas wishing everyone a very happy Tuesday and always looking
• 0:06
  Speaker 1: forward to this time a week. We have appointment listening
• 0:08
  Speaker 1: here in the morning show. One of them is right now.
• 0:10
  Speaker 1: We call it the breit Bart Inside Scoop, and today
• 0:13
  Speaker 1: happy to welcome back to the bright Bart Inside Scoop
• 0:15
  Speaker 1: in the fifty five krc Morning Show Breitbart's tech editor
• 0:18
  Speaker 1: Colin May nine. And of course, as I always say
• 0:20
  Speaker 1: before we start talking, book mark this site b R
• 0:23
  Speaker 1: E I T b a art dot com. Welcome back, Colin.
• 0:26
  Speaker 1: It's always great having you on the show. Happy to
• 0:28
  Speaker 1: be here, Brian, and sort of welcome to the city
• 0:31
  Speaker 1: of Since then, I know, it's a virtual class. Anybody
• 0:34
  Speaker 1: in America or wherever, whoever you're listening, whoever you're listening
• 0:36
  Speaker 1: to my voice, including Germany, Steve, Good morning again. Empower
• 0:40
  Speaker 1: you America dot org. A virtual class. Just register ahead
• 0:43
  Speaker 1: of time, log in at seven pm because Colinen, Collin's
• 0:46
  Speaker 1: got to be talking about technology and artificial intelligence and
• 0:49
  Speaker 1: what's coming our way with a question mark after that,
• 0:51
  Speaker 1: what is coming our way?
• 0:52
  Speaker 2: Colin?
• 0:55
  Speaker 3: Well, Brian, you know they kind of they kind of
• 0:58
  Speaker 3: put me under the microscope with that headline, because anyone
• 1:01
  Speaker 3: who says with complete confidence they knows what's coming is
• 1:05
  Speaker 3: lying to you, you know, right. I like to say,
• 1:08
  Speaker 3: I don't even really write a speech because if you
• 1:11
  Speaker 3: write a speech about AI, it's going to be out
• 1:14
  Speaker 3: of date by the time you give it. Yeah, no question.
• 1:17
  Speaker 1: That's like technology generally speaking. Man, I'm old enough to
• 1:20
  Speaker 1: remember having to learn MS DOS to deal with, you know,
• 1:24
  Speaker 1: with computers, and I learned basic when I was in
• 1:28
  Speaker 1: K through twelve education. And you know, fast forward to today.
• 1:31
  Speaker 1: We have moved at light speed and every day you
• 1:33
  Speaker 1: wake up something has dramatically changed. So the world is
• 1:36
  Speaker 1: crazy with technology. So I certainly get your point on that.
• 1:40
  Speaker 1: And artificial intelligence is going to be the most moving
• 1:42
  Speaker 1: target out there. I mean, I read articles where we're
• 1:44
  Speaker 1: all going to be unemployed. AI is putting everybody out
• 1:47
  Speaker 1: of a job. To no, no, no, the human element
• 1:49
  Speaker 1: will never be replaced. And of course, the only thing
• 1:53
  Speaker 1: I think it is immune from some impact by AI.
• 1:55
  Speaker 1: And you can feel free to chime in on this generalized,
• 1:57
  Speaker 1: generalized statement trades. AI can't install plumbing or electric or
• 2:03
  Speaker 1: along those lines, So maybe that's the one area of
• 2:06
  Speaker 1: refuge from AI. But your response to that, Colin, well,
• 2:11
  Speaker 1: in large.
• 2:12
  Speaker 3: In part, I agree with that. You know, AI cannot
• 2:17
  Speaker 3: do the trades. They can't fix your plumbing, et cetera.
• 2:20
  Speaker 3: They can't lay you know, they can't do the electrical
• 2:23
  Speaker 3: wiring that they need for their data centers.
• 2:25
  Speaker 2: So that would be the view of Nvidia CEO.
• 2:29
  Speaker 3: Johnson Wong that you know there's going to be an
• 2:32
  Speaker 3: explosion of jobs in the trades.
• 2:33
  Speaker 2: And related to the trades.
• 2:36
  Speaker 3: You know, certainly the powers that be want AI to
• 2:41
  Speaker 3: do those things. They envision the future, and this would
• 2:45
  Speaker 3: be elon mush for example, where robots do everything. So
• 2:49
  Speaker 3: you know it's not AI necessarily that turning the wrench
• 2:54
  Speaker 3: on that pipe, but it's a robot controlled by AI.
• 2:58
  Speaker 3: That's directly you get determinators and you know our party.
• 3:04
  Speaker 1: Colin, I literally just got done writing Fighting Wars for
• 3:08
  Speaker 1: us when you said that. It's the first thing that
• 3:10
  Speaker 1: came to my mind anyhow, Right.
• 3:13
  Speaker 3: So you know that's there's certainly some of these people
• 3:17
  Speaker 3: working towards that. I'm a big believer in the human
• 3:20
  Speaker 3: element because I see a lot of problems with AI,
• 3:24
  Speaker 3: and I don't you know, I don't believe that the
• 3:28
  Speaker 3: proper application is trying to replace humanity and have us all,
• 3:32
  Speaker 3: you know, sitting on virtual reality, eating crickets and smoking wheed.
• 3:38
  Speaker 3: You know, that's frankly what Silicon Valley wants for our
• 3:41
  Speaker 3: children and grandchildren. You and I don't want that. For
• 3:44
  Speaker 3: our chance. Our children and grandchildren.
• 3:46
  Speaker 1: Get are stipend from government for buying food. Otherwise it's
• 3:49
  Speaker 1: naval contemplation and poetry writing for all of you. Interesting,
• 3:54
  Speaker 1: But you mentioned problems, and you know, depending on people's
• 3:58
  Speaker 1: perceptions of what AI represents in terms of problems, I
• 4:01
  Speaker 1: think a lot of it springs largely from at least
• 4:03
  Speaker 1: from you know, my listeners standpoint, a lot of people
• 4:06
  Speaker 1: who are concerned about AI. You know, the whole garbage in,
• 4:09
  Speaker 1: garbage out philosophy that who's programming it has sort of
• 4:14
  Speaker 1: pre programmed an implicit bias towards things that might not
• 4:18
  Speaker 1: sit well politically, ethically, or morally with a lot of
• 4:20
  Speaker 1: people out there that might use AI. How do we
• 4:23
  Speaker 1: address that component.
• 4:26
  Speaker 3: That's a tremendous problem. And there's two forms of bias
• 4:31
  Speaker 3: for one, you know, as if one wasn't bad enough. Brian, So,
• 4:34
  Speaker 3: first of all, you know, we have the bias of
• 4:38
  Speaker 3: the people making the AI. You know, pro tip there's
• 4:43
  Speaker 3: not a lot of red hats in sitting in open
• 4:47
  Speaker 3: AI and meta right.
• 4:49
  Speaker 2: And right shocking. He gets not the HR for doing that.
• 4:56
  Speaker 3: So how that impacts things is AI almost works in
• 5:02
  Speaker 3: two stages. It formulates answers and then it sort of
• 5:05
  Speaker 3: checks those answers against almost.
• 5:08
  Speaker 2: The rules for behavior. Here's how I'm supposed to act.
• 5:11
  Speaker 3: Okay, And so those very biased progressive leftists create those
• 5:16
  Speaker 3: rules for how you're supposed to act. That causes AI
• 5:19
  Speaker 3: to do a lot of wacky things, like when Google.
• 5:22
  Speaker 3: You know, one of the infamous moments in recent AI
• 5:25
  Speaker 3: history was Google launched the Gemini Gemini AI image creator.
• 5:31
  Speaker 3: One of the rules that Google set for that tool.
• 5:34
  Speaker 2: Was make everything diverse.
• 5:36
  Speaker 3: You know, it doesn't matter.
• 5:37
  Speaker 2: What they asked for, make it diverse. To hope.
• 5:39
  Speaker 3: People said, you know, make a picture of founding fathers,
• 5:42
  Speaker 3: and you got Asian founding fathers, Black founding fathers.
• 5:46
  Speaker 1: Black George Washington, if I recall correctly, it generated that one.
• 5:49
  Speaker 3: Yeah, and you know, make a picture of Nazis and
• 5:52
  Speaker 3: you got Indian Nazis. You know, like it was doing
• 5:55
  Speaker 3: what it was programmed to do the other portion. And
• 5:59
  Speaker 3: this is another form of bias is AI operates off
• 6:03
  Speaker 3: training data. You know, you know, imagine your hard drive
• 6:07
  Speaker 3: at home times ten million. Okay, that's the amount of
• 6:12
  Speaker 3: training data these things work with. It's not much more.
• 6:16
  Speaker 3: And that training data is not sort of clean factual data.
• 6:21
  Speaker 3: For example, they love to use Reddit to train oh no,
• 6:25
  Speaker 3: to train these things, and you I did from your
• 6:28
  Speaker 3: action you know who's on Reddit?
• 6:29
  Speaker 1: Oh my god.
• 6:30
  Speaker 3: Well, we don't want the generation, the next generation to
• 6:33
  Speaker 3: become since all day on Reddit, that's being sucked into
• 6:37
  Speaker 3: AI as factual, good info and it's operating on that.
• 6:42
  Speaker 3: So you now see there's sort of two forms of
• 6:45
  Speaker 3: bias in one.
• 6:47
  Speaker 1: Well that kind of goes back to my garbage in
• 6:49
  Speaker 1: garbage out, because one of the other components is erroneous
• 6:52
  Speaker 1: information that not only would you have biased information if
• 6:55
  Speaker 1: you're just looking at Reddit. Not only there's random musings,
• 6:58
  Speaker 1: people with sphincters we all got one, and offering their
• 7:00
  Speaker 1: thoughts and comments on any given topic. But you know,
• 7:03
  Speaker 1: when it comes to generating what we hope are is
• 7:06
  Speaker 1: legitimate content, like for example, the many times stupid lawyers
• 7:10
  Speaker 1: relied on AI generated briefs and then come to find
• 7:14
  Speaker 1: out when they go to court to judge, you know,
• 7:15
  Speaker 1: fund them in contempt because the cases that are cited
• 7:17
  Speaker 1: in the AI generated brief are fake. They are whole
• 7:21
  Speaker 1: cloth made up. How can that even be?
• 7:24
  Speaker 3: There's actually, you know, I love talking about this with you, Brian,
• 7:28
  Speaker 3: because you're a lot background. There's actually now a worst
• 7:31
  Speaker 3: case where a person on appeal not only did they
• 7:35
  Speaker 3: have made up cases and stuff, but they made up
• 7:38
  Speaker 3: from whole cloth. The AI made up from whole cloth
• 7:41
  Speaker 3: quotes and statements that the prosecutor as supposed to know,
• 7:46
  Speaker 3: and they're defamatory. They basically said, look how evil and
• 7:49
  Speaker 3: wrong this prosecutor is when they said quote unquote and
• 7:53
  Speaker 3: it was not in the record.
• 7:54
  Speaker 1: It's just oh my god. So yeah, there's other lawyers
• 8:00
  Speaker 1: out they're reacting an equal shock on that one that
• 8:01
  Speaker 1: I have Geese Louise making up the lower court record.
• 8:06
  Speaker 2: Wow.
• 8:07
  Speaker 3: So, Brian, you know one of the things I want
• 8:09
  Speaker 3: to do tonight with empower you and I think this
• 8:11
  Speaker 3: is really valuable for people is I'm in the front
• 8:15
  Speaker 3: line to this right.
• 8:16
  Speaker 2: I'm dealing with the.
• 8:17
  Speaker 3: News every day and what that's enabled me to do
• 8:20
  Speaker 3: is kind of create some baseline concepts that can help
• 8:25
  Speaker 3: you understand the news about AI. So I'll just share
• 8:29
  Speaker 3: one right now because it goes to what you're talking about.
• 8:32
  Speaker 3: It's very very clear that when people use AI to
• 8:36
  Speaker 3: replace themselves or human activity in a final product like
• 8:41
  Speaker 3: a legal filing, a letter, anything like that, people don't
• 8:46
  Speaker 3: even read that before they turn it in. Oh my lord,
• 8:51
  Speaker 3: it's good, and they turn it in. So that lawyer,
• 8:53
  Speaker 3: it would never you know, no matter how many lawyers
• 8:55
  Speaker 3: get busted for this, it never occurs to the lawyer, Hey,
• 8:59
  Speaker 3: I ought to have my in your associate check, you know,
• 9:02
  Speaker 3: Brian Thomas v. Collin, and make sure that case actually exists. Right,
• 9:06
  Speaker 3: They just they just say the machine said it. The
• 9:08
  Speaker 3: machine is God. I'm turning it in. And so that
• 9:12
  Speaker 3: happens with kids, that happens with lawyers, that happens with
• 9:16
  Speaker 3: everyone in between.
• 9:17
  Speaker 1: But you know, the mind blowing thing about that. I
• 9:19
  Speaker 1: hate to do one of the legal aspects of this,
• 9:20
  Speaker 1: and maybe a little bit weed dwelling, but a case
• 9:23
  Speaker 1: cited in anything generated for me, I'm going to have
• 9:25
  Speaker 1: to look at it anyway, just to find out if
• 9:26
  Speaker 1: it's still valid case law. Not that I would even
• 9:29
  Speaker 1: think first off, that they how could something that's supposed
• 9:32
  Speaker 1: to be intelligent whole cloth create a case that never
• 9:36
  Speaker 1: even existed in law books. Now, that seems to be
• 9:38
  Speaker 1: a glitch that should be easily solvable by our you know,
• 9:40
  Speaker 1: intellectual lords and masters putting this together. But I would
• 9:44
  Speaker 1: first go and grab that case and want to run
• 9:46
  Speaker 1: through shepherds to find out whether it's still case law
• 9:49
  Speaker 1: that's valid or has been overturned by some other decision,
• 9:52
  Speaker 1: because there's no guarantee that they've run it through that
• 9:54
  Speaker 1: process and generating the original brief.
• 9:58
  Speaker 2: Brian, you're so last center.
• 10:00
  Speaker 3: I just can't deal with.
• 10:01
  Speaker 2: People like you.
• 10:03
  Speaker 3: So let me add one guy a little more feel
• 10:07
  Speaker 3: to the fire story is we're not when when we're
• 10:10
  Speaker 3: talking about this happening, Brian, we're almost never talking about like,
• 10:15
  Speaker 3: you know, just out.
• 10:15
  Speaker 2: Of law school of people, and it's like a say, like.
• 10:19
  Speaker 3: A you know, a minor traffic accident kind of case,
• 10:22
  Speaker 3: right right, This is happening on gigantic stages. This is happening.
• 10:27
  Speaker 3: One of the classic cases this has happened is a
• 10:30
  Speaker 3: major mega law firm doing Walmart and you know, the
• 10:35
  Speaker 3: judge called them out on it. They had to remove
• 10:38
  Speaker 3: the one of their partners from the case, and I
• 10:40
  Speaker 3: think they lost the case against Walmart eventually in part
• 10:43
  Speaker 3: because you know, they had this stain on them of
• 10:46
  Speaker 3: doing these fake filings. And so you know, that's not
• 10:48
  Speaker 3: a one man firm, that's not Matt Locke messing up
• 10:51
  Speaker 3: a case or something. Right, this is you have a
• 10:54
  Speaker 3: million people in suits and no one can check that
• 10:57
  Speaker 3: that citation.
• 10:58
  Speaker 1: And billing the out of fifteen hundred dollars or more
• 11:00
  Speaker 1: an hour, good gig.
• 11:03
  Speaker 2: If you can get it well.
• 11:04
  Speaker 1: And then pivoting over to something even more profound, and
• 11:07
  Speaker 1: I think probably more people can connect with the idea
• 11:09
  Speaker 1: that AI is certainly going to be used in the
• 11:12
  Speaker 1: field of medicine, and the idea that a erroneous diagnosis,
• 11:17
  Speaker 1: an erroneous treatment protocol, or something is relied upon, maybe
• 11:21
  Speaker 1: by a physician who's looking for a little help trying
• 11:23
  Speaker 1: to figure out his patient's disease state, might rely on
• 11:26
  Speaker 1: something that could ultimately prove to be I don't know,
• 11:28
  Speaker 1: even fatal to a patient.
• 11:31
  Speaker 3: Right. So, you know, this is a concern I've actually
• 11:34
  Speaker 3: talked to with my son about. So my son is
• 11:38
  Speaker 3: graduating college, he's going to medical school to no, you know,
• 11:43
  Speaker 3: neither of us have any concern that you know, by
• 11:47
  Speaker 3: the time he's out there will be no doctors.
• 11:49
  Speaker 2: She'll just get a doctor. AI.
• 11:51
  Speaker 3: We don't lose a minute of sleep about that, right, right,
• 11:53
  Speaker 3: So what we do, what we have talked about is
• 11:58
  Speaker 3: you know, how will that be used? And there's no
• 12:01
  Speaker 3: question AI has some great applications clearly in medicine, so
• 12:08
  Speaker 3: ranging from drug research because it can sit there all
• 12:11
  Speaker 3: day and all night running numbers and checking things and
• 12:16
  Speaker 3: flagging the humans.
• 12:17
  Speaker 2: Hey, this might work. It can.
• 12:19
  Speaker 3: Also it has a lot of applications and things like
• 12:23
  Speaker 3: reading you know, X rays and.
• 12:27
  Speaker 1: I agree, CT scans, MRIs. They could probably identify tumorous
• 12:31
  Speaker 1: things or little things that may be overlook by a
• 12:33
  Speaker 1: human being. I get that, and of course what positions to.
• 12:36
  Speaker 3: Brian, there's actually it's funny because when you look at
• 12:39
  Speaker 3: test results scans, AI is good at catching certain things,
• 12:44
  Speaker 3: but humans.
• 12:44
  Speaker 2: Are better at catching very experienced.
• 12:47
  Speaker 3: Right, So that's why they should work together. AI should
• 12:49
  Speaker 3: always help us, not replace us. But like to your point,
• 12:53
  Speaker 3: one of the problems with AI, and it goes to
• 12:57
  Speaker 3: the same thing with the lawyers, is people assume the
• 13:00
  Speaker 3: AI is right when the AI is basically a moron
• 13:03
  Speaker 3: that has a huge database and can google that right,
• 13:07
  Speaker 3: So we doctors are going to have to be trained
• 13:11
  Speaker 3: how to use it and how not to say, well
• 13:15
  Speaker 3: the computer said this, so it's this exactly. Well, that
• 13:18
  Speaker 3: will not be true no matter how much the technology
• 13:21
  Speaker 3: advances learning.
• 13:22
  Speaker 1: As a learning prager, we'll have to.
• 13:25
  Speaker 3: Use it as a friend and a support, not a
• 13:28
  Speaker 3: not a replacement.
• 13:29
  Speaker 1: Seven PM log in only empower your America Dot or
• 13:32
  Speaker 1: just register ahead of time and enjoy what Colin has
• 13:33
  Speaker 1: to say and learn a lot. And finally, before we've
• 13:35
  Speaker 1: part company here real quickly, I note that from the
• 13:38
  Speaker 1: notes that you're going to addressing the ideas for the
• 13:40
  Speaker 1: conservative movement to master AI for its own purposes. Briefly,
• 13:44
  Speaker 1: where's that going, Colin?
• 13:47
  Speaker 3: Well, what that's all about is we're behind on the
• 13:51
  Speaker 3: left because the left made this stuff. So we have
• 13:54
  Speaker 3: to quickly put together, you know, some framework of how
• 13:58
  Speaker 3: we're going to use AI and not just you know,
• 14:01
  Speaker 3: the answer is not to just be a latite who
• 14:04
  Speaker 3: doesn't use it, because you're using it today even if
• 14:06
  Speaker 3: you don't know it, if you call your credit card company.
• 14:09
  Speaker 1: I know.
• 14:09
  Speaker 3: But but you know, we can use it effectively for
• 14:14
  Speaker 3: our ends, but it's mostly not falling into the traps,
• 14:19
  Speaker 3: you knows, not falling, not falling into the danger zones
• 14:23
  Speaker 3: where we brought our brains, especially for young people.
• 14:26
  Speaker 1: That's why we have you. Collum Adine, Tech editor, Breitbart
• 14:28
  Speaker 1: Learn tonight seven pm. Empower you America dot org. Don't
• 14:31
  Speaker 1: follow in the traps, listen to what Colin has to say. Colin,
• 14:34
  Speaker 1: you'll be back on the Morning Show at some point
• 14:35
  Speaker 1: in the future. I'll look forward to that. In the meantime,
• 14:37
  Speaker 1: thank you for everything that you and the team at
• 14:39
  Speaker 1: Breitbart does. I appreciate your efforts and I certainly rely
• 14:42
  Speaker 1: upon you in preparation for the Morning Show. Outstanding reporting there, sir.
• 14:47
  Speaker 3: Thank you, Brian, and I hope everyone joins us tonight
• 14:49
  Speaker 3: and power you.
• 14:50
  Speaker 1: Amen to that. Empower you America dot or you'll be
• 14:53
  Speaker 1: linked on my blog page fifty five KC dot com.
• 14:55
  Speaker 2: Have great