00:00.00 archpodnet The timer shit I don't know what's going on on my side now it says it's recording. Okay, let's just take it all right? Yeah, welcome to this era Mark Podcast episode 2 62 segment 3 sorry let me do that again. 00:04.94 Bill White I can see it. 00:06.21 Doug I can sit. Yeah, we're like 13 seconds 14 seconds in. 00:13.34 Bill White Yeah. 00:17.96 archpodnet Welcome back to the Sharemark Podcast episode 2 62 and I don't know I just ended my little rant at the end of the last segment so I don't know who's up now Bill. Ah. 00:30.93 Bill White Ah, well, you know the thing I want to talk about is folks who do have the belief right? Like you know we're we're talking about all these different kinds of scenarios where reports are writing themselves or you know clients can just pay 1 of us to you. Give our certifications to the state and then click in and just get a map of every single site and every single thing that's nearby and then they don't really need. You know they just need us to go out there and check and then Doug's mentioning probably the most pragmatic and probable approach that humans are cheaper than software and we'll just continue using humans either until they break or just stop doing the job I mean that's. That's kind of where crms are at right now where we can't find humans who will work for the low enough wages and so now they're telling me that I need to train more people because they broke them all and didn't pay them but you know I'd I'd like to ask the question of what about all the folks who are trying to make it work right. 01:12.22 archpodnet Um. 01:23.24 Bill White So What about all the folks that do believe that their essay is good that they've conjured out of the internet or what about the companies that do think this is a passable report in actual archaeological science because they just cut and paste it I mean we've we've covered boilerplate before on this show show you know and folks who don't Know. Ah. Companies. Don't write all their reports. They just cut and paste sentences and paragraphs and sometimes entire sections from previous reports and they only update them based on like what few articles that person has had time to read or other things that they've figured out from other sites. But in general, it's really all just kind of cut and paste it and that's going from. Their compliance reports to the proposals to your contract that you got hired if you even signed one I mean all that stuff the company already has so you know if if someone can use the computer and save 7% more I can see how that would be a thing but you know what do we do about folks who believe. That this is the future and start raking the internet and this is just what they're doing. This is how they're doing archeology. 02:26.29 Doug Um, I mean not for archeology. But we're already seeing it in pseudoareology like literally just the other day someone was ah did a whole threat on Twitter about how? um, basically so there's ah, there's a. You guys probably have seen these but they'll be like accounts where people will basically just choose a topic pump out a bunch of stuff that's um, basically tailored to that social media platform get hundreds of thousands of all followers and then basically sell on the account. And they usually if it if it has to do with like archeology. It usually ends up being like a lot of aliens a lot of Atlantis but that's because that's that's what sells and stuff and so it just ends up being pretty shit. But um, you're already seeing that where basically people have started to create like Facebook groups. Um, or you know pages and stuff and they're pumping out an incredible amount of fake crap like they're making up images. They're just complete bonkers and people are just absolutely eating it up because you can make some pretty crazy. Ah, you know photos generated. Ah, photos there that you know looks I think they still look like basically photoshop crap. But I honestly I've seen things in the past where I've it's like incredibly obvious that it's photoshopped and people be like no no, it's real I'm like dude the shadows, the shadows are even like in the same direction like. 03:57.92 Doug You could tell that it's photoshopped just off of that. Also it looks like the wrong pixels the wrong colors. It's just so obvious but like people can't see it. Um and you're already seeing this this deluge of of crap out there and this is what I'm thinking like. A lot of it's probably going to be more negative than positive. So yeah, maybe but we we shave off like point zero five percent of our budget. Um, by being a little quicker on writing our reports but honestly like unless you're having someone check that report. 04:32.19 Doug It's going to be crap. It's Goingnna be the same thing. We're just going to put out a lot more crap and set of quality and I I think it's going to be the same like I mean now it's obviously the examples I've just given are people who are out there to make money and you know do a poor job I don't think a lot of archaeologists are out there to do that. But. Um I don't see this going anywhere well and this goes back to Chris's thing I agree with Chris that like actually honestly we probably don't need a lot of reports like a lot of our reports is we we basically take the data and then we put it into somewhat of a narrative but it's really one of the most. Okay, I'm probably I think I'm not going to insult anyone this but like it's really boring. It's it's like it's like oh and then we and then we dug you know ten centimeters down and then we dug another ten centimeters down and then we dug another ten centimeters down and like trying I've I've gone through reports and like trying to make a report sound. 05:16.40 Bill White Ah. 05:30.51 Doug Halfway not repetitive, not boring is is is a bit brutal but it's because we're just putting data and trying to add like a narrative to what is that I agree like we could. There are better ways to do things I Just don't see this machine learning being able to do like. Honestly getting rid of reports is something we could do without machine learning I think we should um, well reports in the sense of we have these big narrative things where we could just put the data there. Um I think you know there are better ways of doing it now without machine learning. Um I don't see. I Just see machine learning helping us keep putting out the same old things that are not the highest quality. Yeah I think it could probably be a crutch. It'll keep us doing the the wrong things and inside the right things. 06:15.50 Bill White Yeah, yeah, well and you know I agree with a lot of what you said some of those structures on the repetitiveness are because you know the state historic preservation office and the um. You know advisory council we have to have x amount of sections in a report like they have to have these sections and every ship was telling you the report guidelines and we're trying to make it interesting. You're right? but I mean walking next to your buddy thirty feet apart across the forest and digging a hole every you know thirty feet or you know ninety feet that's how creative do you really make that but I'm I'm kind of concerned with some of the stuff like what Chris was saying how many of our clients think that just doing this is good enough like how many of our clients think that just scraping the database and getting a map and making some tables is good enough. And in that case, they really only need like one rpa right? because they can just you know scrape and do the class one and be like oh heck I guess we do actually have to do probes oh nevermind. Let's just change the project area and so now that site's nowhere near there therefore nothing's done and you know we can just do some kind of windshield survey I mean how many how many of our clients could possibly get savvy. And think that just making crap and putting it out there with a map and some tables is good enough right. 07:28.66 archpodnet Ah. 07:31.12 Doug I Mean you kind of say it like that doesn't already happen I mean I mean like. 07:38.30 Bill White Ah, well you worked with me doug it does not ever happen in my office. Ah ah, no serious I was fortunate I was fortunate enough to work with folks who valued the quality of the writing and they valued they were like this is all the archeology we get to do so let's make it good. 07:44.68 Doug Ah. 07:53.50 Bill White And so they did want us to learn how to write and they didn't want us to scrape and do crap and even the boilerplate they were like well don't just assume straight up that something that was written fifteen years ago about this town is still. You still have to go to the library. You still have to find resources and try to figure things out like we can't just only accept this as what's going on and so. You know I came through several companies that that was they did want it to be better and they didn't want us to do something like chat Gpt. 08:21.51 Doug And ah but I think that's the thing is like it's amazing. How much good archeology that happens that ah okay, my wording here is going to be a bit iffy and I realize someone's got rip me a new one in the comments but like. 08:38.51 archpodnet Um, yeah. 08:38.74 Doug We We do a lot of good archeology that technically we don't have to do and that's because like most people that go into archaeology want to do good archeology. So like if you look at like all the regulations across across almost like every country and like if you've. People follow just to the the letter of the law and some people do and you get some real crap out there. But like if we did it. It would all be real crap. Um, it's amazing like legally like yeah okay, some shippers will come back or in the U K you know, um, you know. Council archaeologists and they'll be like actually we need Better. You know higher quality of this this and this honestly like if push came to shove and you you sued them. They would have to back down. Um like a lot of archeology is basically done because we want to do good archeology. 09:37.44 archpodnet Yeah I'm here but I'm back I'm back. 09:40.29 Bill White So that we lost Chios are you there. We found that last east egg. Okay, yeah, no I I fully agree with that and and. 09:42.32 Doug Did I kill the conversation. 09:43.41 archpodnet I Think you lost me for a minute. Okay, you go ahead, go ahead. No. 09:50.31 Andrew Ah, yeah. 09:58.33 Bill White Fully agree with what you were just saying Doug that this thing really does ah you know continue and and keep itself alive because there's a lot of folks who really do care about archaeology and that's why things aren't just absolute trash at every single place because the folks that are doing it. They really want to do archeology. I saw a thing online. You know we've been talking this podcast for a long time and um, you know like my advice to folks who want to do archeology is don't consider this your only career you're going to do this until you decide to do something different and we've even interviewed people who switched careers. But I saw ah you know, ah a thing online someone talking to a couple of folks who did leave archeology and how they were saying. Yeah, they have something that you know pays the bills and gives them benefit but they they really miss being in the field and doing archeology. And so I really think the folks who are in here they care about it and you know we're really putting in more than just ah, you know money and effort into this where this is our dream like this is we're putting a lot of ah you know energy that we really care about that. You don't always find in all careers and that's the only thing that keeps it from being you know Skynet. So ah, there's no way to really calculate There's no way to quantify that and you can't capture that in a software and there's no ai that's going to cover someone who's wanted to be an archeologist from the time they were seven years old and finally gets to do it. 11:24.18 archpodnet Yeah, you, you guys are right? Um, yeah I mean everything is here's what I'm going to to say to to kind of finish out this discussion right? guys I think we need to just look to the future and say what do we want it to look like right and we need to use these new tools. We need to use. But we need to use these new tools. Let me emphasize that part because if we don't it's just going to be like archaeology is always you know 1520 years behind technologically right? We need to acknowledge the fact that new things are coming down the line in a way to do things and again I'll go back to what I said either in the beginning segment 2 lead into segment 1 what is the output. What are we trying to learn here right? We're we're trying to learn more about how past societies you know thought and behaved interacted with their environment. You know what they were what they were thinking what they were feeling. We're trying to learn all those sorts of things and that's really the only thing that we're trying to figure out. Okay, archeology is just one tool that helps us do that. But there's all kinds of other things that we use from you know, geological resources to you know other technologies when are in our excavation and things like that where we can analyze different facets of the environment from pollen to you know, different chemicals in the ground and things like that to help us understand how. Our ancestors and everybody else's ancestors and and maybe somebody who you know is just back there and doesn't have any descendants. You know how they interacted with the world and that's the output who cares how we get there. You know what I mean who cares if it's you know we're we're entering all this information into a ah machine learning algorithm. 12:55.71 archpodnet Because it can do it quicker and see all the connections that we can't see and spit something out that we can interpret and see how it see how it looks I think we'd be foolish not to at least try that sort of technology as it gets better and better and better and see what kind of output it gives us go ahead. Doc. 13:19.56 Doug Right? Yeah I mean I going go back I going say like we should be a bit more amish. Um, and that's to say like so everyone sort of thinks of like the amish with the buggies and that they like give up all technology and that's not true. Um the amish and it depends on different groups and different families and stuff like that. 13:25.94 archpodnet Yeah. 13:37.70 Doug Um, they go through a process where they literally evaluate New Technologies. So All new technologies that comes out the amish have actually tried Um, they try cell phones they try everything um and they evaluate it on the on the process of does it bring them closer to God or further away from God. Um, and I I think that's a good approach that people should take with all technology does it bring you closer to what you want to get and if that's understanding the past and that's great or if it takes you further away then obviously that's a negative I Just don't see the way It's designed so like. 14:04.37 archpodnet Sure. 14:15.81 Doug All those like the the chat and the bard they are not actually again this goes back to their not artificial intelligence. They're machine learning and all they're good at is guessing the next word like there's there's no it can't they're they're specifically design and it's how they're engineered. They basically guess the next word and they guess it really? Well so it sounds like human speech. But there's nothing there. Um, so I don't see that being a tool that's going to very much help us or if it does it's going to be such a marginal bit that like honestly like you know. Ah, bigger deal would be like if you're you're going to lose more money if ah, one of your vehicles breaks down and you can't get out to site for like two days or something like that's going to cost you more to a project than this marginal gains. We're going to get out of these sort of machine learning writing things I honestly think. The damage. It's going to do is going to be so much worse like the internet right now is kind of a dumpster fire. Um in terms of trying to get ahold of the stuff but man once all the spammers just basically turn sites into nothing but gibberish. Like we're gonna like right now. It's it's nearly impossible to fight pseudo archeology because there's just so much of that crap out there. But that's nothing that's like ah ah, ah, ah, a drop a pin in the ocean. Yeah, it's ah it's going to be It's gotta be. 15:35.58 Andrew Well not not impossible, not not completely impossible. You now. 15:45.26 Doug Andrew it's going to be impossible after this because basically people are going to create so many sites that like your anything on Google the first four thousand are going to be just gibberish. 15:52.80 Andrew And that's that's true and you know what that does gives me employment for the next century. Yeah yeah. 15:56.20 archpodnet Nice, nice. 16:04.12 Doug Ah see Angie's taking the jobs of ah robots. 16:05.87 Andrew I Totally am. 16:06.42 Bill White Ah, well, the last thing last thing I'd like to say to people who you know students who are looking to a robot to write your essay You're not learning how to write and the number 1 thing that employers are looking for is people who know how to write. And if you learn if you learn how to write with chat gbt you better hope that the next company wants you to do the same thing there because as we're finding out. It's pretty unlikely the next company is going to have you use chat gbt and you're instead going to have to use your brain and experience. So if you use the software to get through. Yeah you might make it through college because I mean you pay upfront for college like I get paid either way whether you learn or not. It doesn't matter. But if you're trying to rely on that tool or crutch or whatever you want to call it to make it through a career where you have to write every day good luck. My friends good luck with that Thanks. 16:46.40 archpodnet Okay, well we're wrapping this up and I just want to comment 1 more thing on both what you bill and Doug just said I don't think any of us are really talking about this being a tool right now right? It's still I mean in the last year this thing is really exploded, especially chat gp. It's really exploded on to the you know the the consciousness of the everyday person right? I mean they've been working on these kinds of things for a little while now. But it's really just come on so that we are talking about this on the cerra mark podcast but I'm talking about you know, 5 years from now 10 years from now. It's going to get better That's how those things go. It's going to get better and whether or not we should even use it then is ah is an ethical question that we need to ask ourselves right? I don't know if the answer to that is yes or not but we won't know until we get there. You know. 17:37.50 Doug Ah I know Chris like we're supposed to have self-driving cars in 2014 and there has been no improvement in a decade like like we're we're assuming that it's going to get better, but it could be alchemy like literally there is nothing. 17:37.54 archpodnet Right? But right? but that's not but doug. 17:52.80 archpodnet Ah, cars. 17:54.88 Doug Alchemy is never going to be anything because it can't be anything I Ah you're saying that this might be something but it may never like what we're talking about here the the machine learning on these things. It can't do anything but make up references. 18:06.67 archpodnet Right? right? now. Okay, well, that's that's highly unlikely right? because that's not how computers are designed I mean computers get better exponentially and that's just how it is right? look at your. 18:11.32 Doug Because that's how it's designed like I'm not I don't think you're going to see much improvement. 18:23.10 archpodnet Look at the phone that's in your hand today or the computer you're sitting in front of versus the one you had ten years ago right they get better and cars are not a great example either because cars are not able to evolve because of infrastructure and regulation not because of a lack of will in technology I mean they want. There's. Plenty of car manufacturers out there right now that can put self-driving cars on on the road but they're not and there are some actually in California but there are not doing that because of the will of the people and regulatory guidelines that are preventing it and then. You know, different cities not wanting it to happen and the infrastructure not being there for it to actually work has nothing to do with technology right? Absolutely nothing. So I don't know with that. 19:06.94 Doug Well no, but that's just that technology hasn't and like so also like you've just said computers get better. No the transistors the we double on transistors every eighteen months well 1 is one's a physical of what they can do. 19:13.93 archpodnet What's the difference. What's the difference. 19:25.65 Doug The other is actually the algorithm that you put put into it so you can like you you can't you can crunch bigger and bigger numbers. But if you're not crunching those numbers Correctly, it doesn't matter like like you like you could crunch like and that's that's the thing with self-driving cars is it's not. 19:42.88 archpodnet Sure. 19:43.10 Doug You say it's regulatory, but it's actually the problem is it's the ah it's the edge cases. It's always been the edge cases of when something goes wrong, Who's who's responsible for the money. Um, who who's going to pay the insurance who's going to pay for the the person who dies. Um, yes, but it's always been there. 19:49.83 archpodnet Right? That's regulatory. 19:58.49 archpodnet No. 20:02.68 Doug And and and it's never going to disappear and the self-driving cars have never, they've never got to the point where they're they're going to be able to overcome that hurdle. Even though you get you get more and more um you know, better and better computer chips in the cars. That's not going to affect it because. There's physical problems with edge cases that it just can't You'll never be able to understand um and that's the thing with this ah with these is it's just it's not going to be able to do it. 20:27.24 Andrew You know what? you know what? you know? what? No no, this is important. This is very important. Well you guys Yammer on I've built a bunker and I've had my daughter do many many sit ups. So I am ready for the end times. 20:38.74 Doug I. 20:41.91 archpodnet Ah, right? All right guys we are way over time on this segment and this podcast so I'm going to end it right? there. But if you're listening to this I want to hear your thoughts on this. You know if you're over on Facebook or or you know our our website actually you can actually comment on the blog post. Ah well they're actually kind of blog posts as far as format goes. But you can comment on the podcast show notes page right? at apodnet.com/serumareology okay and feel free to email us at any 1 of our contact information or Twitter or whatever we have listed on the website. Okay, so we want to hear your thoughts and opinions on this and where serm is going in the future with all this ai and machine learning or maybe it's going absolutely nowhere. And Doug is right? but I doubt it all right that we'll see you guys later all right keeping a rolling for the outtro. Thanks to everyone for joining me this week thanks also to the listeners for tuning in. 21:37.14 Doug Hold on hold on home Chris Chris hold on I need to go to I need to go chat gt I need to put in and have them do my outro for me real quick if you could just hold on one second um it's going to tell me how much time I have and how I should do the most perfect outro. 21:40.81 archpodnet Oh my god. 21:49.93 Bill White Ah, ah. 21:54.52 Doug That we'll both enjoy Chris and keep a tradition going. Oh okay, in that case goodbye. 21:55.22 Bill White I Think this is it. You did it. You don't need yeah Gbt This is it. 21:55.28 archpodnet Oh my god Jesus all right bye. Ah. 22:06.51 Doug Jesus won't save you Chris Jesus won't save you from me. 22:06.61 Andrew See you guys next time. 22:07.18 Bill White By ah oh gosh.