1
00:00:16,000 -->
00:00:25,000
Dr. Kent Gustavson: Welcome to Man vs. Slop. There's two men here on the on the podcast. I'm Kent Gustavson.
2
00:00:25,000 -->
00:00:27,000
Aleck McCathie: I'm Aleck McCathie.
3
00:00:27,000 -->
00:00:33,000
Dr. Kent Gustavson: And then I guess the other featured guest this week is Slop.
4
00:00:33,000 -->
00:00:36,000
Aleck McCathie: Yeah, recurring guest.
5
00:00:36,000 -->
00:00:54,800
Dr. Kent Gustavson: I find translation of Slop very interesting between countries. I forget which country or what what language I was I was speaking to someone about, but trash or rubbish is another way to say slop. It seems like a more, you know, nice and proper way to say it.
6
00:00:54,800 -->
00:01:05,000
Dr. Kent Gustavson: But I kinda like slop because it's so it's so messy and so... Yeah. It's not it's not part of everyday language, slop.
7
00:01:05,000 -->
00:01:25,000
Aleck McCathie: Yeah. It's definitely a word that is used more now than ever before just in the context of AI, I guess. I I really like it too because it just reminds me of this sort of amalgam, like, unidentifiable thing that you have to sort of identify in context every time you talk about it.
8
00:01:25,000 -->
00:01:47,500
Dr. Kent Gustavson: I feel like it could also be... You know, we work with a lot of kind of motivational speakers or, you know, thought leaders. I feel like it should be s period s s period l period o period p period. We gotta figure out what... Maybe we should have a Slop of the Week, like what that acronym means.
9
00:01:47,500 -->
00:02:07,000
Dr. Kent Gustavson: But I... The one I just came up quickly with with was like silly little old something that starts with p. Silly little old person. There you go. But Yeah. More, you know, heavy leadership principles like sincerity, you know, likability, you know, all of that.
10
00:02:07,000 -->
00:02:09,000
Aleck McCathie: Originality.
11
00:02:09,000 -->
00:02:12,000
Dr. Kent Gustavson: There you go. Except those are anti SLOPs. That's the... Yeah.
12
00:02:12,000 -->
00:02:12,700
Aleck McCathie: Yeah. True.
13
00:02:12,700 -->
00:02:17,000
Dr. Kent Gustavson: It'd be the ironic SLOP, acronym.
14
00:02:17,000 -->
00:02:36,500
Aleck McCathie: Yeah. Just within the context of, like, what is slop? Reading sort of the headlines this week, it's interesting. I would... Even things that I would maybe would disagree as being slop are being sort of named as this sort of sloppiness.
15
00:02:36,500 -->
00:03:01,100
Aleck McCathie: The one headline that kind of was the most interesting to me is this idea that OpenAI has been going crazy working on these, like, ancient math problems, like the ones that just haven't been solved by humans yet because they're too complicated and they take too long, I guess. And apparently, one of the latest models took, like, just three hours and it, like... I think there's tens of thousands of these things, if not thousands.
16
00:03:01,100 -->
00:03:23,100
Aleck McCathie: And it worked through like 700 of these like old ancient sort of math manuscripts and just like dumped them into the industry of people who are are working on these things and they have to go through now all of these, like, human checks. And it must be just a load of work for the people who do that kind of thing. And they're mad about it. They're like, no. We don't wanna work with this slop.
17
00:03:23,100 -->
00:03:42,000
Aleck McCathie: Like, it hasn't actually been determined whether or not they're, like, the correct answers yet, but it must be, like... It's annoying for them, obviously, because they didn't do the work, and now they have a lot of stuff to sift through. But, like, they're referring to it as Slop where I'm like, those might actually be breakthroughs. You know?
18
00:03:42,000 -->
00:03:59,500
Dr. Kent Gustavson: That's kind of interesting. It's sort of like mountaintop mining, approach. You just take off the top of a mountain, and there must be something good in there. So then they, you know, they just chop it into chunks and sift through it and take out all the, in quotes, good stuff. I'm not a fan of mountaintop mining.
19
00:03:59,500 -->
00:04:14,400
Dr. Kent Gustavson: I just... It occurred to me... Or I guess panning for gold would be the same thing. You know, just take all the muck and, but what's interesting there, that's a that's a good analogy, either the mountain or the panning for gold because you have to be in the right place.
20
00:04:14,400 -->
00:04:35,900
Dr. Kent Gustavson: And so if if AI is actually in the right place to solve important math problems and so on, that's super interesting, far beyond the scope of math. Obviously, it can solve the problems of our time. It can figure out solutions to things that the Nobel Prizes this week are being awarded for.
21
00:04:35,900 -->
00:05:02,000
Dr. Kent Gustavson: It's fascinating to hear people try to describe what what they're getting Nobel Prizes for because I... It... They're they're they're, way outside of my knowledge in some ways. You know, like how the the left and the right hand of a of a molecule, you know, like, there's a mirror image effect that's... This is this is... Permeates through chemistry and all this stuff.
22
00:05:02,000 -->
00:05:17,000
Dr. Kent Gustavson: And it's like, wow. That's amazing. I really... I wish I could go back and study all that stuff again. And if you're in the right ballpark, AI can actually really help with that stuff, I think. Anyway, I'm... Who knows?
23
00:05:17,000 -->
00:05:44,000
Aleck McCathie: Yeah. I mean, certainly that math thing is a good example of, like, something potentially good happening. But I think it's funny that, like, the sloppiness is that... Like, the production is practically free, I guess. Like, it... It's a whole lot of junk and the thing that's not free is the verification, the human side of it.
24
00:05:44,000 -->
00:06:15,600
Aleck McCathie: I think that it was GPT and maybe another company in the last week or so who... They've just been, like, cutting their prices back on some of their lower models to almost, like, to the point where it would be silly to not use AI if you're in some kind of content industry, some kind of industry that that requires you to sort of, I guess, put things out into the world with the idea of quantity over quality.
25
00:06:15,600 -->
00:06:39,000
Aleck McCathie: It's just making me think that, like, we're in a time where it's it's practically free to produce, and it's becoming more expensive and more burdensome to verify, to human check, to even read. There's there's gotta be more AI books and more AI content coming out now than humans would ever be able to parse through, at least in our lifetime.
26
00:06:39,000 -->
00:06:57,500
Dr. Kent Gustavson: Yeah. Well... And I... What I think is interesting is the speed of things. Like, I I I said the same stuff about YouTube a few years after it came out because, like, it was so hard to find... When I was a kid, it was hard to find a certain CD or a record or a tape.
27
00:06:57,500 -->
00:07:22,400
Dr. Kent Gustavson: And I remember in college, I would save up for CDs so I could I could have it and listen to it. And then the iPod came out. And then all of a sudden, got streaming and just... You're you're just driving down this superhighway, and all of a sudden, the entire history of filmed concerts is available within a second, and it's free. It's unreal.
28
00:07:22,400 -->
00:07:42,000
Dr. Kent Gustavson: Any of your favorite artists used to have to get bootlegs under the desk of some weird drugstore, and they'd be like, yeah. I got the bootleg, which you want a Tom Waits bootleg? Yeah. I got four options, but don't tell anybody. And they'll put them on the desk and let you look at them like it's a drug deal, and it was. It was illegal, I'm sure. But amazing.
29
00:07:42,000 -->
00:07:57,200
Dr. Kent Gustavson: And the first time I went to Europe, I I I got bootlegs of of a couple different artists, and it was so cool. And now there's YouTube. And so it feels like it's a different thing. It feels like we're just paradigm switching. We we... It's like you you're...
30
00:07:57,200 -->
00:08:25,700
Dr. Kent Gustavson: So so here's an interesting thing. I have people in my life, I've said this before, who don't like AI, and they go to Google and they use AI. Right? And, oh, wasn't it wasn't it helpful that the search engine gave me exactly the answer I needed at the exact right moment? Gave me the exact steps how to get into my, whatever, my email when I got locked out or something. Right?
31
00:08:25,700 -->
00:08:47,700
Dr. Kent Gustavson: Like these, the things used to have to go to YouTube or whatever. It's incredibly practical the way that AI is being rolled out, and it's absolutely free when you look at Google. Google has made it free. If you can access the Internet, you can use AI for free. And that I find fascinating.
32
00:08:47,700 -->
00:09:15,300
Dr. Kent Gustavson: At the same time as you've got people saying that the AI companies are too big to fail. That sort of dangerous language from The Big Short. Right? Like when you've got these massive trillion dollar companies throwing snowballs at each other and determining what's gonna happen in society and all this other stuff and the... You know, there's a council at the White House and everything else.
33
00:09:15,300 -->
00:09:32,700
Dr. Kent Gustavson: So you've got this matter of scale where you've got every individual having a free search on Google about anything in the world, and it probably has an idea about it, might hallucinate it, but it probably has a pretty good idea, Google... The Google search engine's better than the other AIs.
34
00:09:32,700 -->
00:09:48,000
Dr. Kent Gustavson: And then, at the top, you got these trillion dollar companies playing with our future. Are they worth it? Are they not worth it? What if they fail? What if they don't fail? And they tank the world economy? Anyway, that's what's been occupying my brain this week.
35
00:09:48,000 -->
00:09:59,100
Aleck McCathie: Yeah. The volume thing is huge. I I like the paradigm shift you're talking about because I think, you know, in our generation, I'm a little bit younger, but I think that, like, there's been so many.
36
00:09:59,100 -->
00:10:23,100
Aleck McCathie: Like, when I was a kid, there was certainly a lot of people I knew who had things like Napster, which are, you know, similar to that idea of bootlegging your music. That was filled with slop. I mean, you could certainly go on there and probably find some, like, songs that you weren't supposed to find, but I mean, I knew a lot of kids who downloaded viruses and stuff and screwed their family PCs. You know, that that kind of stuff happened.
37
00:10:23,100 -->
00:10:40,900
Aleck McCathie: But there was no, like, guidelines or safety regulations. Of course, that was a pretty gray area. YouTube was very similar though. I know there's a lot of, like, nefarious stuff on there too. Probably not to the degree of, like, harming your computer, but certainly content that you wouldn't want a child to see.
38
00:10:40,900 -->
00:11:02,000
Aleck McCathie: Unlisted videos and stuff like that can contain nearly anything. Again, a place that has no guidelines. I think there's like parent safety features, but you have to have a knowledgeable parent in order to set that up. And those are like low volume. I mean, you have to actually search for the stuff that's bad or find it accidentally.
39
00:11:02,000 -->
00:11:27,400
Aleck McCathie: But now with AI, it just seems like there... There's not a corner of the Internet you can go or not a software that you can open that doesn't have some kind of AI built into it, and the amount of stuff you can do with it that becomes accidentally nefarious or, you know, gives you information that might be detrimental is crazy.
40
00:11:27,400 -->
00:11:41,000
Aleck McCathie: And it's weird that it's only now after it's been developed for several years and gotten into the homes and onto the smart devices of everyone that we're starting to think about, like, maybe kids shouldn't have access to this.
41
00:11:41,000 -->
00:12:09,400
Dr. Kent Gustavson: So I love that you said smart devices because that's one of those words. Right? It's a word that was borrowed from the standard syntax and went to now explain why this phone isn't for making calls, it's a smart device. What does that mean? Well, it's a computer. So somebody was really smart and they said this is Vitaminwater, this is a smartphone.
42
00:12:09,400 -->
00:12:31,100
Dr. Kent Gustavson: And surprisingly, what it what it looks like is Mark Zuckerberg might have been whispering in the president of the United States' ear, because, you know, Meta has its own, I think it's called, like, super something, Superintelligence Labs, or some... I forget what the name of that is.
43
00:12:31,100 -->
00:12:59,800
Dr. Kent Gustavson: And now all of a sudden, artificial intelligence, AI, or around the world it has different initials depending on what language, now it's super intelligence. And we talked about that last week. AI really isn't always super intelligent based on the word, But it actually looks like this initial might be sticking because the big tech companies are are, talking about it.
44
00:12:59,800 -->
00:13:18,000
Dr. Kent Gustavson: Elon Musk is talking about it even though he was initially against the whole super thing. So whoever bought up the domain, you know, like, opensi.com is probably a rich person right now.
45
00:13:18,000 -->
00:13:34,700
Aleck McCathie: Yeah. I heard that, I think it was this week, that they sort of finished, like, the official federal renaming of of super intelligence. Now there's like a force and like, I guess, a government body sitting behind this superintelligence idea.
46
00:13:34,700 -->
00:13:49,000
Aleck McCathie: For me, I mean, I I haven't read much about it yet, but it feels like it's just like the same thing as, like, the Gulf of Mexico being renamed to the Gulf of America. Like, is everyone gonna start saying that or or no? Like, I'm not sure.
47
00:13:49,000 -->
00:14:19,500
Dr. Kent Gustavson: So so both of those are kind of interesting because both of them have valid arguments. They might have come from somebody a little off their rocker, but it is North and South America. I know everybody says America is the United States, but theoretically it's North America South America.
48
00:14:19,500 -->
00:14:36,800
Dr. Kent Gustavson: Gulf of America sounds patriotic in the US, but theoretically I can see where it's coming from. It sort of makes sense. Similar superintelligence to me makes more sense than artificial. So let's talk about the word artificial for a second.
49
00:14:36,800 -->
00:15:01,000
Dr. Kent Gustavson: Because now Trump said, because artificial is like fake. Well, I think that is one definition. Artificial is fake. Although I am a vegetarian and when I eat artificial meat, am still eating something that's maybe somewhat good for me nutritionally, and it's still food. Right? Even though it's artificial meat.
50
00:15:01,000 -->
00:15:18,100
Aleck McCathie: Yeah. My first sort of spark in the brain is thinking about artificial sweeteners. We don't use sugar very much. We use, like, stevia and stuff like that. It's often sold as, like, artificial sweetener, I guess, along with aspartame and all those other ones.
51
00:15:18,100 -->
00:15:33,700
Aleck McCathie: But there's always something like... Even with artificial meat, there's always something that feels a little bit, like, icky about it. Like, it's... The word artificial sort of has this condescending meaning in my head. I don't know whether that's the same for everybody, but it just...
52
00:15:33,700 -->
00:15:45,000
Aleck McCathie: Even if it's a good product, like, you know, tofu is artificial meat. Nothing wrong with tofu, I guess, but, like, the... Just hearing it being called artificial makes me feel like, ugh, like, something's wrong with it, you know?
53
00:15:45,000 -->
00:15:53,000
Dr. Kent Gustavson: I I have a big a big... Really good idea. We could call tofu super meat.
54
00:15:53,000 -->
00:15:55,000
Aleck McCathie: Yeah. I like that.
55
00:15:55,000 -->
00:16:08,000
Dr. Kent Gustavson: And and sugar sugar could be super sweetener, except then it's not artificial. I guess stevia could be super sweetener. Stevia, the super sweetener.
56
00:16:08,000 -->
00:16:26,300
Aleck McCathie: Yeah. I don't hate that, actually. It's... It sounds better. It does. It makes me feel less icky about it. But I think I like the idea of having the ickiness applied to artificial intelligence. I don't know if I'm ready to think about AI as being nonartificial.
57
00:16:26,300 -->
00:16:42,000
Aleck McCathie: I know that ickiness helps me keep it separated from the idea of a human brain. And I... There's even people doing work around AI and I guess processing power that are using like real, like, brain tissue? Like, they're growing...
58
00:16:42,000 -->
00:16:54,000
Dr. Kent Gustavson: I might not wanna know about this. They have, like, they're... They've lined up 20 humans, and they're actually... They sit there for eight hours, and they they power AI devices with human brains?
59
00:16:54,000 -->
00:17:15,900
Aleck McCathie: I don't know all of the semantics, but there's, like, people right now who are building, like, the the... Something adjacent to, like, you know, the processing power of a computer, out of, like, grown human brain cells, essentially. Like, there's processing power in our brain.
60
00:17:15,900 -->
00:17:35,000
Aleck McCathie: They're now plugging that into machines so that they can do the same stuff that AI does and compute things very quickly in weird ways. It seems like it's something that should be probably, I don't know, not being done. Like, there's a lot of sort of questioning behind that.
61
00:17:35,000 -->
00:18:07,900
Dr. Kent Gustavson: Yes. I was I was about to say earlier, depends on the AI. I... That's what I find, interesting about all the terms. It's as if we've, as if we've said the word tool represents one thing. Like, if we said, Hey, tools are dangerous. Well, it depends which tool. Like, if it's a toothbrush, you're probably fine. You know? If it's an axe, yeah, it might be a little more dangerous. And that's AI.
62
00:18:07,900 -->
00:18:32,200
Dr. Kent Gustavson: Some of them are really benign. Some uses of AI, spell check that we've had for years. Right? Like your your phone's auto correct, which some people think is not so benign, and I... I'm sometimes included. Like, you do bilingual texts, it's very frustrating. Right? But auto auto, correct, that's AI.
63
00:18:32,200 -->
00:18:54,200
Dr. Kent Gustavson: But then you have AI that's like, I don't know, acting like a butler. You have robotics, right? You've got, I don't know, whatever you were talking about, some brain attached devices. You've got, yeah, like, imagine, it's a jar with a brain in it that's attached to a computer. Like, I guess that's also AI, like anything Asimov might have come up with, or any science fiction heroes.
64
00:18:54,200 -->
00:19:12,000
Dr. Kent Gustavson: And then you've got your full on, like, medical related AI that's figuring out how to cure cancer. It's figuring out how to do a million different things. Well, I guess that cuts both ways though, right? Because you can also craft awful things using that.
65
00:19:12,000 -->
00:19:37,300
Aleck McCathie: Yeah. I my immediate knee jerk reaction is that maybe we should be calling, like, the different tiers of AI different things because I feel like I'm always gonna wanna, like, think about Slop as being artificial. Like, that... When I read the content that has been clearly generated by AI that didn't have much thought put into it, it feels artificial. Like, it didn't feel like it came from a human.
66
00:19:37,300 -->
00:20:08,000
Aleck McCathie: Whereas, I guess, some of the really profound work being done, like medical research and these math solutions that are like beyond human capacity, that feels like super intelligence. That's super interesting, profound work being done. And on the other hand, I just looked up that... The brain thing, and apparently, they're referring to that as OI, which is organoid intelligence or biocomputing. So it already has its own name.
67
00:20:08,000 -->
00:20:39,300
Dr. Kent Gustavson: Wow. That's terrifying. And Yeah. It's amazing in some ways, and, I'm kind of in awe or in amazement or in terror of where things are gonna go from here. Because if this is the tip of the iceberg, what are the young geniuses gonna come up with who grow up in this this world, this stuff?
68
00:20:39,300 -->
00:21:00,200
Dr. Kent Gustavson: And then all bets are off in some ways, like, they're they're figuring out how, you know, they're figuring out the the untold story of, you know, what people have been calling the Big Bang or about the spark of life, and they're figuring out how to get to distant planets and all this other stuff, what is AI gonna do.
69
00:21:00,200 -->
00:21:25,800
Dr. Kent Gustavson: Then, you know what? It's... I I don't hate the idea of super intelligence as long as it's plugged into a human being, because I think it helps... Ah, there you go. Yeah. So that... And so, Batman was just a dude, right, with some devices, that that made him super.
70
00:21:25,800 -->
00:21:50,000
Dr. Kent Gustavson: And that's... Batman is different than some of these these others who got bit by spiders or whatever else. And so it's almost like this super intelligence can make normal human beings with great ideas and a limited lifespan. They can take human intelligence and spin it up several more times, maybe.
71
00:21:50,000 -->
00:22:20,000
Aleck McCathie: Yeah. The blur between how humans interact with it though is starting to become a little bit harder to identify. Like, another thing I read this week was that alongside all of these math breakthroughs, if you want to call them that, I guess they're not really confirmed yet. There's been labs quietly sort of checking whether these new sort of models that are capable of this are capable of breaking like encryptions.
72
00:22:20,000 -->
00:22:40,500
Aleck McCathie: And I I... This whole idea of, you know, Bitcoin and all these sort of very mathy encryption-y sort of things that we built as humans. We built these things where they're like, this is so incredibly complicated. It would take your home computer years to, like, potentially mine a Bitcoin or something like that. Right?
73
00:22:40,500 -->
00:23:04,000
Aleck McCathie: So we've built this, like, human construct that we intentionally have no ability to understand. And we're like, machines are more complicated. We don't have to worry about it. But now, I mean, how long has it been? And we have this potential threat of someone being able to use a publicly available model to say like, go break that encryption. Like, I I want I want a Bitcoin.
74
00:23:04,000 -->
00:23:34,000
Dr. Kent Gustavson: And then and then comes quantum within the next few years sometime, and it just gets faster. And then comes fusion for power. That... Like, it's it's crazy. The the speed of innovation of humans, and that's part of the danger too. Right? Is the... How far can we innovate ourselves into a trap or into something more like utopia? And that's, yeah, interesting to think about.
75
00:23:34,000 -->
00:24:03,300
Dr. Kent Gustavson: I would and I would I would say that another homage to our title, the goo, the primordial goo of a world is where things step out and become life or where nothing happens or where you don't know. And that's in in all biblical traditions and so on. It's like the the wind over the water and there's there's nothing. Right? And out of that nothing comes something. Right?
76
00:24:03,300 -->
00:24:16,700
Dr. Kent Gustavson: Or the Big Bang. There was, you know, that, you know, my very scientific friends tell me about the Big Bang, and I I believe this is a thing. But I'm like, what was around the Big Bang? Were there other Big Bangs? Were there other... Right?
77
00:24:16,700 -->
00:24:42,000
Dr. Kent Gustavson: And so it's... There there are questions, but there are also more questions and answers and more answers. And as we discover things, things get more insane in a way. And so I think we're headed back to primordial goo unless we figure some of this stuff out. And then something else will crawl out of it, probably, but you never know.
78
00:24:42,000 -->
00:25:04,000
Aleck McCathie: Yeah. I do think that we have some backtracking or some different ways to think about how our societal expectations are gonna have to shift. You know, the idea of security is huge, especially in America. I mean, who doesn't think about their home security, security, you know, their passwords, their banking.
79
00:25:04,000 -->
00:25:23,800
Aleck McCathie: There's so many aspects of your life that are, like, intentionally private. They're intentionally behind a firewall or a password or another company's ability to help you feel secure. And and it just seems like we're we're nearing this place where that that paradigm is gonna shift.
80
00:25:23,800 -->
00:25:44,500
Aleck McCathie: And, like, will there be such thing as a password in a year from now? If everyone has access to, like, you know, an AI that they can just have on their PC and it can just, you know, break up encryption that's been stumping humans for hundreds of years in, like, a minute? Like, what will there be such thing as the blockchain?
81
00:25:44,500 -->
00:26:15,000
Aleck McCathie: Will there be such thing as, you know, countries using encrypted messages to speak during, like, wartime or, like, you know, the the the passwords that are, like, behind the the launching of, like, nuclear missiles and stuff. Like, there's so many layers of things that are, like... We think of as being these, like, ultimate secure, unbreakable things. That might not just... That might not be a thing pretty soon, I guess. I don't know.
82
00:26:15,000 -->
00:26:20,000
Dr. Kent Gustavson: That's a good place to end. I say let's pull the plug and keep watching the news.
83
00:26:20,000 -->
00:26:34,000
Aleck McCathie: Let's do it.
84
00:26:34,000 -->
00:26:37,900
Dr. Kent Gustavson: podcaststhatmatter.org.
