2961: How AI Proves Humans Used Fire 800,000 Years Ago

[00:00:01] [SPEAKER_00]: How do we uncover the secrets of our ancient past with modern technology?

[00:00:08] [SPEAKER_00]: Well, today I want to dive into this fascinating intersection of archaeology and AI.

[00:00:15] [SPEAKER_00]: Well, I'm more about groundbreaking research that as you utilize AI to detect traces

[00:00:20] [SPEAKER_00]: of ancient fire use in archaeological context, and push to boundaries of what we know about

[00:00:27] [SPEAKER_00]: early human innovation.

[00:00:30] [SPEAKER_00]: And I'll learn more about some remarkable findings how they're into disciplinary approach

[00:00:34] [SPEAKER_00]: is transforming our understanding of human evolution and pondered a future potential of combining

[00:00:41] [SPEAKER_00]: AI with archaeological data.

[00:00:44] [SPEAKER_00]: We're currently producing something like 30 to 35 episodes every single month reaching

[00:00:50] [SPEAKER_00]: around about 130 to 140,000 listeners around the world.

[00:00:54] [SPEAKER_00]: So I wanted to mention the sponsor of Tech Talks Daily.

[00:00:58] [SPEAKER_00]: And also shine a light on how legacy managed file transfer tools are dated and lack of

[00:01:04] [SPEAKER_00]: security that today's remote workforce demands.

[00:01:07] [SPEAKER_00]: And companies that continue relying on that outdated technology put their extensive data

[00:01:11] [SPEAKER_00]: at risk, but the good news is you can ensure top tier security with tight works MFT suite.

[00:01:17] [SPEAKER_00]: Their software security hardening includes an ongoing bounty program, regular penetration testing

[00:01:23] [SPEAKER_00]: to minimise vulnerabilities, the convenience of a one-click appliance update, and avoid

[00:01:29] [SPEAKER_00]: the hassle of managing multiple unhardened servers.

[00:01:33] [SPEAKER_00]: And as an XIT guy I know the benefits of that.

[00:01:36] [SPEAKER_00]: So step into the future of secure managed file transfer with tight works by visiting

[00:01:41] [SPEAKER_00]: tightworks.com to get started.

[00:01:43] [SPEAKER_00]: That's tightworks.com to get started today.

[00:01:46] [SPEAKER_00]: But enough rambling for me.

[00:01:48] [SPEAKER_00]: Let's get today's guest on.

[00:01:49] [SPEAKER_00]: So buckle up and hold on tight as I beam your ears all the way to Israel where we're

[00:01:55] [SPEAKER_00]: going to try and uncover the hidden stories of our ancestors using AI.

[00:02:01] [SPEAKER_00]: So a massive welcome to the show.

[00:02:04] [SPEAKER_00]: Can you tell everyone listening a little about who you are and what you do?

[00:02:08] [SPEAKER_00]: And for the help I let you go first.

[00:02:10] [SPEAKER_02]: Yeah, okay thank you so I was born in Portugal.

[00:02:13] [SPEAKER_02]: I'm a word chemist, meaning that I'm interested in chemistry by a chemistry material

[00:02:20] [SPEAKER_02]: sciences and more recently in our geology.

[00:02:22] [SPEAKER_02]: Well, I always am to answer those questions and using scientific tools to answer some

[00:02:29] [SPEAKER_02]: big questions.

[00:02:30] [SPEAKER_02]: And so I'm a chemist and I'm an ultimately bestigated at the vitamin issue of science in Israel.

[00:02:39] [SPEAKER_00]: Fantastic.

[00:02:39] [SPEAKER_00]: I don't know, obviously we've got not one but two guests.

[00:02:42] [SPEAKER_00]: I don't could you drive on this thing a little about you too?

[00:02:45] [SPEAKER_03]: Yeah and so I'm not always born in Israel.

[00:02:48] [SPEAKER_03]: Yes, I did my PhD.

[00:02:50] [SPEAKER_03]: Hey, that's why I said to the police and the police and the police and the police and the

[00:02:54] [SPEAKER_03]: police and the police and the police.

[00:02:55] [SPEAKER_03]: They're pastels of mine.

[00:02:57] [SPEAKER_03]: PhD I found a huge unit of what's in your eye.

[00:03:01] [SPEAKER_03]: And I blacked into meeting the pendant room.

[00:03:05] [SPEAKER_03]: I got a good offer from the lab sensors.

[00:03:08] [SPEAKER_03]: The co-faciliters to join with the researcher and I'm working in the department of

[00:03:14] [SPEAKER_03]: mathematics and the lab cell of co-faciliters are in the past six years on projects

[00:03:21] [SPEAKER_03]: really in medicine, on bio-imaging and computer vision elected by the public like this project

[00:03:29] [SPEAKER_01]: very, very, very, very, very, very, very good project.

[00:03:33] [SPEAKER_00]: Well, it is such a pleasure to speak with you both.

[00:03:37] [SPEAKER_00]: I wanted to be since I invited you on here.

[00:03:38] [SPEAKER_00]: I was reading a two-year-old article about how you are using AI in archaeology and this

[00:03:43] [SPEAKER_00]: immediately set off a few fireworks in my head because I've always been fascinated by how

[00:03:49] [SPEAKER_00]: technology is transforming our lives here in the present and in the future.

[00:03:53] [SPEAKER_00]: But I'd not actually thought about how it can transform the world of archaeology and

[00:03:57] [SPEAKER_00]: how we view the past.

[00:03:59] [SPEAKER_00]: So I've got to ask though, what was it that inspired you to combine AI and research for

[00:04:05] [SPEAKER_00]: non-visual traces of fire in archaeological context?

[00:04:09] [SPEAKER_00]: What was it that made you turn to AI?

[00:04:11] [SPEAKER_03]: So it was going to be visual, as I know, not enough to have your soil, the first is

[00:04:17] [SPEAKER_03]: to fire this.

[00:04:19] [SPEAKER_03]: Actually, it's pretty bad, it had the root of data and these data are identical to the

[00:04:26] [SPEAKER_03]: best metal for this, it was very fine.

[00:04:30] [SPEAKER_03]: Complex patterns get out that can correlate with the fire on this point you can

[00:04:36] [SPEAKER_03]: build over the heat of temperature on the lens.

[00:04:39] [SPEAKER_03]: So like this, I believe, for finding patterns and I didn't do the best

[00:04:46] [SPEAKER_03]: little to analyze all these new data really.

[00:04:49] [SPEAKER_02]: Well, if I can add something, well, one of the things that one of these

[00:04:53] [SPEAKER_02]: mentioned before was that we don't correlate archaeology because the social science

[00:04:59] [SPEAKER_02]: with the technology which is technology oriented by people.

[00:05:04] [SPEAKER_02]: So we always forget that part of the equation that people always involving both the cases.

[00:05:09] [SPEAKER_02]: Now, well, one of the things that people think to forget because it's the social science

[00:05:12] [SPEAKER_02]: and they don't have money and the vision that we have is the type of Indiana Jones

[00:05:16] [SPEAKER_02]: going in the least, we don't know, we've been in some of the work.

[00:05:20] [SPEAKER_02]: So it's not bad, it meant, but it's also, it's when you look at it and you look at the AI.

[00:05:27] [SPEAKER_02]: What people are looking for when you develop methods in our in AI, in data?

[00:05:33] [SPEAKER_02]: So we are talking about 100, probably millions of years of cultural material that have been

[00:05:38] [SPEAKER_02]: produced and what else we built in the museum, we've had hundreds of thousands of things that

[00:05:44] [SPEAKER_02]: have been generated by humans throughout these times.

[00:05:48] [SPEAKER_02]: So you have a large data basis that you need.

[00:05:51] [SPEAKER_02]: The question is, how can you use that?

[00:05:54] [SPEAKER_02]: So archaeology has the potential because he has the data set, a large data set.

[00:06:00] [SPEAKER_02]: Then come different challenges where we'll address later on, but as a huge data set

[00:06:05] [SPEAKER_02]: and can give a lot of information about material culture and even up to some

[00:06:10] [SPEAKER_02]: up to certain things about behavior of people.

[00:06:12] [SPEAKER_02]: How people, when he started making decisions, why is culture to this sort of pattern

[00:06:17] [SPEAKER_02]: and behavior versus that one?

[00:06:20] [SPEAKER_02]: So there's a huge potential because there's a lot of data which is what we look at.

[00:06:24] [SPEAKER_02]: We're generating data every day to feed, fetching, pre-teach and so on.

[00:06:29] [SPEAKER_02]: So the database is there, the technology is there and people are there,

[00:06:34] [SPEAKER_02]: that equation that we tend to forget.

[00:06:36] [SPEAKER_02]: So some of their very different ones are working on the past version of the one that we have

[00:06:40] [SPEAKER_02]: hasn't and it's been closer into the what we're adding to the future.

[00:06:44] [SPEAKER_02]: We're regarding also that Wi-Fi and then you look at, I think, well, the reason of

[00:06:50] [SPEAKER_02]: FIRE is that was one of their very few breakthroughs in human evolution.

[00:06:56] [SPEAKER_02]: And there was a, there's an hypothesis called the cooking hypothesis that said about

[00:07:00] [SPEAKER_02]: when human transit from bodies to all of our activities based on bone that they've

[00:07:07] [SPEAKER_02]: changed by chemistry.

[00:07:08] [SPEAKER_02]: So the bones were bigger and so on.

[00:07:10] [SPEAKER_02]: So they came up with formulating this idea that about 2 million years ago,

[00:07:17] [SPEAKER_02]: we started making FIRE and FIRE allow us to eat because the FIRE what does it just shocks?

[00:07:24] [SPEAKER_02]: The molecules, if they're small, digestible elements which is what we do and people

[00:07:29] [SPEAKER_02]: actually tied to the experiments with raw fruit, raw footage, that's what they call them.

[00:07:35] [SPEAKER_02]: And so the is this theory that says that the cooking hypothesis, what is what made us humans

[00:07:41] [SPEAKER_02]: because they're at men or humans or genus or mal were able to handle fire and took a

[00:07:49] [SPEAKER_02]: 12 fire after a certain point and then cook their meat and so let allow them to have large

[00:07:54] [SPEAKER_02]: energetic intake. So and then there would be less depend on an imaginary panda that is always eating

[00:08:00] [SPEAKER_02]: leaves. So it needs to constantly be eating because the an additional value of the bamboo leaves are

[00:08:07] [SPEAKER_02]: very low so it just keeps eating what is the large part of.

[00:08:10] [SPEAKER_02]: Eating if you would have eaten bamboo leaves probably would have a space of three hours in between

[00:08:16] [SPEAKER_02]: eating them. So a bit like it's a bit far-fetched comparison, but still it gives you the

[00:08:23] [SPEAKER_02]: fairly the idea of what that means of having cooking food so that allow people to move away

[00:08:30] [SPEAKER_02]: from the center of the food and from the trees and does start exploring and that would be

[00:08:35] [SPEAKER_02]: that another only the bone in the structure but also the brain and the cognitive and spatial

[00:08:40] [SPEAKER_02]: recognition patterns that are so people could walk and explore and stop and the gaitlin

[00:08:46] [SPEAKER_02]: freedom. So to say and there's no technology for it to be proved or this is theory because

[00:08:53] [SPEAKER_02]: it's too many years ago and everything is perishable every organic experience that you don't find

[00:08:58] [SPEAKER_02]: any source of carbon. Yeah by anything like rocks they've been around for millions of years and

[00:09:04] [SPEAKER_02]: those can ask this for us to detect the failure mark on the rocks so that's kind of the idea behind.

[00:09:11] [SPEAKER_00]: And that's one of the reasons I invite you on the podcast today because if we look forward

[00:09:15] [SPEAKER_00]: into the future there's a lot of talk at the moment that AI needs data and we're almost running out

[00:09:22] [SPEAKER_00]: of data as we're racing towards the future but you guys have got a very different problem. You've

[00:09:26] [SPEAKER_00]: hundreds of millions of years of data that AI can bring to life which is just mind blowing to me.

[00:09:33] [SPEAKER_00]: So can you elaborate on some of the specific AI techniques that you employ to detect things like

[00:09:39] [SPEAKER_00]: the ancient use of fire and how do these techniques differ from let's say the traditional

[00:09:44] [SPEAKER_00]: archaeological methods and in the other Jones figure that people might have in there?

[00:09:50] [SPEAKER_03]: So yes, actually we wanted to trick the flint on the rock and just evaluate the temperature

[00:09:57] [SPEAKER_03]: and walk it out of the hollow couple of years. So what really led the model?

[00:10:07] [SPEAKER_03]: So what he has thought he did with the toilet, this is like the flint as and actually

[00:10:13] [SPEAKER_03]: rings spectroscopy yes to take extreme care point from this flint. So we use a

[00:10:21] [SPEAKER_03]: specific Arama spectroscopy and you have a signal. The signal contains the finger print or this

[00:10:29] [SPEAKER_03]: main. Now there are flint that will intercept for many energy of the battle and this thing

[00:10:37] [SPEAKER_03]: I think they didn't put around those days. It's not that you're getting them on the

[00:10:44] [SPEAKER_03]: couple of years and so it must have been very late about let all that net.

[00:10:51] [SPEAKER_03]: 8-bit frame, the signal to the entire hole. Now take a real signal and turn in there

[00:11:07] [SPEAKER_03]: of what's happening. It was a fog well in us, no? Yeah. It was a huge order to split down the net

[00:11:15] [SPEAKER_03]: or anything last 10-15 years. You cannot play it yet. I did because the final deep,

[00:11:23] [SPEAKER_03]: it helped me get in the deep. We feature it exactly, it's not insights from the signal. Yes.

[00:11:32] [SPEAKER_03]: And on the decade, they used people on the record.

[00:11:36] [SPEAKER_03]: They were models like Tisha Rondipoks.

[00:11:39] [SPEAKER_03]: They were using much better,

[00:11:41] [SPEAKER_03]: as well as the Tatya and Rukovishinar

[00:11:43] [SPEAKER_03]: not the other new models.

[00:11:45] [SPEAKER_03]: And all the more modern people,

[00:11:47] [SPEAKER_03]: it's all the time.

[00:11:50] [SPEAKER_03]: And actually using the Dekyan

[00:11:53] [SPEAKER_03]: sound like the Nareen model.

[00:11:56] [SPEAKER_03]: Another aspect is that we're supposed to

[00:12:32] [SPEAKER_02]: the digital ideology and there will be a day to turn off side between social sciences

[00:12:38] [SPEAKER_02]: and mental science, whatever you want to categorize them. There's not the discussion

[00:12:41] [SPEAKER_02]: of the thing you write here in today but for the second simplicity let's keep it

[00:12:46] [SPEAKER_02]: this way. But what the traditional they still rely on are some people are reluctant to use

[00:12:55] [SPEAKER_02]: certain data so it's very difficult sometimes to analyze human behavior based on just a few

[00:13:01] [SPEAKER_02]: measures. And so in that case the traditional ideology is very focused on dating,

[00:13:08] [SPEAKER_02]: like dating is the specific site. This is 14,000 and so on and so on and so forth

[00:13:12] [SPEAKER_02]: to be on the history all of a very visual but at the same very subject. So I don't think

[00:13:19] [SPEAKER_02]: that this AI is going to take some subjectivity out of the interpretation but still

[00:13:26] [SPEAKER_02]: it's still very subjective because we're talking about humans and based on technology

[00:13:30] [SPEAKER_02]: not looking at the technology we're using technology something to understand human behavior.

[00:13:37] [SPEAKER_02]: But the traditional ideology we feel that first of all there's a huge amount of data

[00:13:42] [SPEAKER_02]: but there's not so much people going from computer to AI technology. That's the first step.

[00:13:50] [SPEAKER_02]: We need more people for obvious reasons you can get higher in what ever company sophisticated

[00:13:58] [SPEAKER_02]: technology driven company. While probably no university or even any company will hire you

[00:14:05] [SPEAKER_02]: to do AI in our challenge. So in that case we're very fortunate to be at the vitaministic

[00:14:10] [SPEAKER_02]: because it allows us to do all the freedom of expression, scientific freedom that we could have.

[00:14:17] [SPEAKER_02]: And there's not that people are convinced with AI so there's two people using the

[00:14:22] [SPEAKER_02]: same traditional archeological methods. What we put over is just another method that would

[00:14:30] [SPEAKER_02]: have more data which like we once I have a friend that wants to sit there it's a world of probabilities

[00:14:37] [SPEAKER_02]: right I'm talking about human behavior. So what we did with AI we just reduced like that by 2%.

[00:14:43] [SPEAKER_02]: So it's not that it's going to solve it's a download data this they were making fire no it most

[00:14:47] [SPEAKER_02]: likely we're doing fire and that was always in the fantasy world that we've done so we

[00:14:54] [SPEAKER_02]: soon push the boundaries that the traditional technology, the archeology, I still very subjective

[00:14:59] [SPEAKER_00]: and that's what we try to balance. What would you say with a what were the biggest challenges

[00:15:07] [SPEAKER_00]: that you faced when developing and implementing your AI model for archeology and

[00:15:12] [SPEAKER_00]: and if you can share on how you overcome some of those challenges as well?

[00:15:15] [SPEAKER_02]: First of all we didn't know what we were doing. Very honestly because there's nothing that

[00:15:20] [SPEAKER_02]: we could guide ourselves on anyway. We'll take a look at what we're talking about. What would you

[00:15:25] [SPEAKER_03]: mention in your guess world can they go and go on I wanted it. Is that a world you

[00:15:30] [SPEAKER_03]: were doing your things right? The main thing is the camera can be near you. We just don't know where

[00:15:35] [SPEAKER_03]: we were having it. So yeah I think the total challenge yes look for my hypothesis this one challenge is

[00:15:45] [SPEAKER_03]: dead some of the accurate enough it's a 10 by close but one this is you go to companies

[00:15:53] [SPEAKER_03]: by doing what the impulse theory looks like amount pound the bit this is the

[00:16:03] [SPEAKER_03]: little challenge second challenge we need to do this one at a time and it is looking

[00:16:10] [SPEAKER_03]: on well data but then that challenge also especially not for a bit we need to divide the

[00:16:15] [SPEAKER_03]: strength term we need to test it again is new data not the data sake that you know we'll call

[00:16:22] [SPEAKER_03]: they'll go to our talk different time and if I were to say you can define for you also levels

[00:16:28] [SPEAKER_03]: so this is the second challenge it will be can you realize some you're launching data yes

[00:16:35] [SPEAKER_03]: yeah and I'm actually a little very very in the industry because just to

[00:16:41] [SPEAKER_03]: I to eliminate any balance and that we can enter to the modern pipeline and you

[00:16:48] [SPEAKER_02]: call the condition there's no development of the AI approach yeah are we even adding to

[00:16:55] [SPEAKER_02]: the small things to the challenges one is that once we get the data we cannot ask the people any

[00:17:02] [SPEAKER_02]: more because they're not here so like we do that like not silence right you silence okay that's it

[00:17:09] [SPEAKER_02]: so that's one of the things and then actually brings to the other question about

[00:17:13] [SPEAKER_02]: that's a math and the chemistry we got it all right probably our geology as well but

[00:17:19] [SPEAKER_02]: the big question is once it's about the interpretation of the data meaning that it says to

[00:17:27] [SPEAKER_02]: with human behavior so we're trying to use mathematical models and chemistry both combined

[00:17:33] [SPEAKER_02]: material scientists to try to explore some sort of human behavior or to especially file related

[00:17:40] [SPEAKER_02]: now if the models that we can have debate that if a model will tell us listen this is

[00:17:48] [SPEAKER_02]: how it works and it's within the art with what the archaeologists think it is then it's fine

[00:17:54] [SPEAKER_02]: they will come to us and say why are you spending so much resources to do something that we already

[00:17:58] [SPEAKER_02]: know but then it doesn't match what they stay then how do we figure that out so it's

[00:18:05] [SPEAKER_02]: good sort of paradox that how are we going to validate an experiment of people that are not

[00:18:11] [SPEAKER_02]: there anymore and there's no other more elements that we can use to validate them so we kind of

[00:18:15] [SPEAKER_02]: getting the disgracey loop of how to validate the data mathematically we need a accuracy when you

[00:18:22] [SPEAKER_02]: but from the human behavior becomes a very big challenge and that's what I think the ceiling

[00:18:28] [SPEAKER_02]: of the boundaries of AI is not anymore to make an app that you can take a picture and say is mean

[00:18:34] [SPEAKER_02]: a pottery firm that whatever century it was that is not the question anymore I think the question

[00:18:39] [SPEAKER_02]: there for the future of AI is really what up and you validate because you put that thing on a

[00:18:46] [SPEAKER_02]: car you bump the car in an accident and you we train the model and you do it again so

[00:18:51] [SPEAKER_02]: you have a permanent feedback evaluation over there it's opening the air so that's kind of

[00:18:57] [SPEAKER_02]: until we solve that problem I think we can do many interesting things but then this is a very

[00:19:01] [SPEAKER_02]: fundamental question in AI and I call it into my mind. And I'm curious if we just put the

[00:19:10] [SPEAKER_00]: technology to one side for a moment and AI how did your interdisciplinary team dynamics added that

[00:19:17] [SPEAKER_00]: contribute to the success of your research and other any particular memorable moments from

[00:19:22] [SPEAKER_00]: your collaboration I'm sure you've come back with more than a few stories but anything you can share

[00:19:26] [SPEAKER_02]: around that so I think one of the most memorable moments is like okay let's do this and then we'll

[00:19:34] [SPEAKER_02]: reach in your to you do and so why not let's do this okay let's do this okay let's do this

[00:19:44] [SPEAKER_02]: and then we need that stuff you can come to your office of course let's go and meet

[00:19:49] [SPEAKER_02]: and it shows me like a little bit works and that was the wall moment where every scientist is

[00:19:53] [SPEAKER_02]: looking for it was like oh no okay we have something big coming up and that's for me and the

[00:20:00] [SPEAKER_02]: rest was technical challenge that we had to overcome but that was the moment we said oh we have

[00:20:05] [SPEAKER_03]: something in our hands that is worthwhile exploring. He's a wrong one and you're going to tell you

[00:20:12] [SPEAKER_03]: the same as minute it is the case I'll start with my difficulty. I'm not about that moment

[00:20:22] [SPEAKER_03]: it is when we have the first paper on the kitchen kid it was on the view

[00:21:31] [SPEAKER_03]: And it's go to the local blog, they have a world of France.

[00:21:37] [SPEAKER_03]: They have more the way I was the best in the world.

[00:21:40] [SPEAKER_03]: The project and tools.

[00:21:42] [SPEAKER_02]: And I think my experience of working at the interfaces and I've worked in many different interfaces with our architects and companies and fun and support.

[00:21:52] [SPEAKER_02]: At the end of the day what is really important is that people know how to communicate and not to behave

[00:21:57] [SPEAKER_02]: and take the other side.

[00:21:59] [SPEAKER_02]: It's like a sort of relation that you need to create and show that communication with sounds.

[00:22:04] [SPEAKER_02]: And so Benchland dynamics is like, okay this doesn't work.

[00:22:07] [SPEAKER_02]: What can we do?

[00:22:09] [SPEAKER_02]: Let's think about this altogether.

[00:22:10] [SPEAKER_02]: And because it's uncharted territory, actually even more fun.

[00:22:13] [SPEAKER_02]: It's like, okay, is this not working?

[00:22:15] [SPEAKER_02]: Wow cool.

[00:22:16] [SPEAKER_02]: Okay, instead of it's like being depressed like, oh, we didn't necessarily plan and so we had to feed them because

[00:22:21] [SPEAKER_02]: we're not on the spotlight.

[00:22:22] [SPEAKER_02]: Nobody knew how to do this.

[00:22:23] [SPEAKER_02]: We didn't know how to do this.

[00:22:24] [SPEAKER_02]: So the exploratory thing was very exciting and also promoted the motivation and dynamics of the team, right?

[00:22:31] [SPEAKER_02]: Like always going to do that.

[00:22:32] [SPEAKER_02]: I've no idea.

[00:22:33] [SPEAKER_02]: Okay, let's think about this and come back next week.

[00:22:36] [SPEAKER_02]: This dynamic was not based on any of the grand proposal pressure or anything we have to,

[00:22:43] [SPEAKER_02]: we didn't do anything because we knew that nobody else was doing it.

[00:22:47] [SPEAKER_02]: But at least nothing was published within know that.

[00:22:50] [SPEAKER_02]: And that created a certain very, really style dynamic of constructive and thoughtful and constructive, like,

[00:22:59] [SPEAKER_02]: our science is robust.

[00:23:01] [SPEAKER_02]: Not to slow nothing but constructive and inclusive manner.

[00:23:05] [SPEAKER_00]: And for anybody that missed the article that I was talking about at the beginning of this episode,

[00:23:10] [SPEAKER_00]: can you just explain the significance of the findings at the everyone quarry and have a compared to other known evidence of early fire use?

[00:23:18] [SPEAKER_00]: Just to bring that to life for people listening, hearing about it for the first time and what you discovered there.

[00:23:24] [SPEAKER_02]: The grand quarry is an archaeological site located at currently the most active video.

[00:23:30] [SPEAKER_02]: He's located in the north and it dates about a million years ago.

[00:23:35] [SPEAKER_02]: And so there were no evidence.

[00:23:37] [SPEAKER_02]: One of the archaeological sites, probably the one on our records, the one that was roaming around this thing.

[00:23:44] [SPEAKER_02]: And and therefore discovered, backing the 70s and was excavated.

[00:23:48] [SPEAKER_02]: Now it's a dump and so we cannot expedited that anymore but we have this project and we thought, well,

[00:23:55] [SPEAKER_02]: don't get hurt to a lot.

[00:23:56] [SPEAKER_02]: Our geology set will be you guys want to find anything because there's no clear evidence of fire.

[00:24:02] [SPEAKER_02]: And it's like, well, that's a good challenge.

[00:24:04] [SPEAKER_02]: We've shown before that the model works as the needle mentioned before.

[00:24:08] [SPEAKER_02]: So it's very generalized. We can work. The math is right there. Curacy is right. Everything looks great.

[00:24:14] [SPEAKER_02]: Let's, so another place. So we went again on this exploration.

[00:24:18] [SPEAKER_02]: Again, it sounds like Indiana John's type of thing.

[00:24:22] [SPEAKER_02]: And we just started this and actually when we saw the results, they say not the one we were going to be able to say.

[00:24:29] [SPEAKER_02]: Oh, there was fire.

[00:24:32] [SPEAKER_02]: But even their archaeologists mind below it's like, I don't believe this. This is impossible.

[00:24:35] [SPEAKER_03]: I don't know what they are about these. It is absolutely long another number of moments.

[00:24:42] [SPEAKER_03]: This is a very far, I can't remember.

[00:24:45] [SPEAKER_03]: Final, it has an aquarium that we can lay in there and it's quite a bit.

[00:24:53] [SPEAKER_03]: And then we're going to be done with the questions without decisions.

[00:24:57] [SPEAKER_03]: We have the confidence that we see in our development of the main.

[00:25:01] [SPEAKER_02]: And so what happened is that they are only a handful of archaeological sites with fire.

[00:25:07] [SPEAKER_02]: They're discovering bugs in every decade.

[00:25:11] [SPEAKER_02]: And they are only five archaeological sites that do it run in evidence of fire below half a million years.

[00:25:18] [SPEAKER_02]: That's basically me.

[00:25:20] [SPEAKER_02]: We cannot make a series theory because they are only evidence of like,

[00:25:25] [SPEAKER_02]: something like this.

[00:25:26] [SPEAKER_02]: The narrative really, there's one charcoal about a million years in South Africa.

[00:25:32] [SPEAKER_02]: They have only handful of sites.

[00:25:34] [SPEAKER_02]: And so what they impact that our findings were is that this is probably the second site out of Africa.

[00:25:40] [SPEAKER_02]: But it means that all more of the actors was roaming around this place out of Africa,

[00:25:45] [SPEAKER_02]: which everybody thinks that everything studied in Africa and stayed in Africa.

[00:25:48] [SPEAKER_02]: And then all of the sapiens came later about 200,000 in different ways.

[00:25:51] [SPEAKER_02]: Now, what was the genetics I was saying in the fifth and broad?

[00:25:56] [SPEAKER_02]: No, but what we say is like not only this kind of, they will guys who are here.

[00:25:59] [SPEAKER_02]: They're also doing fire.

[00:26:01] [SPEAKER_02]: So that means that they brought it with them.

[00:26:03] [SPEAKER_02]: A lot of culture.

[00:26:04] [SPEAKER_02]: Or they were doing the local.

[00:26:06] [SPEAKER_02]: So that's we don't know.

[00:26:08] [SPEAKER_02]: That's kind of the resolution we don't have.

[00:26:10] [SPEAKER_02]: And so that is the impact of having the second site.

[00:26:14] [SPEAKER_02]: The other one is often a million in stain.

[00:26:16] [SPEAKER_02]: And this one is probably the oldest archaeological site that shows evidence is a fire out of the Atlantic.

[00:26:21] [SPEAKER_02]: And then, you know, what we did also what he showed for us was very important.

[00:26:28] [SPEAKER_02]: Like the signal preserved in the rocks like this one.

[00:26:33] [SPEAKER_02]: But as he did this, it will stay there for a million years.

[00:26:36] [SPEAKER_02]: It's the signal on this rock.

[00:26:38] [SPEAKER_02]: So like the timestamp of fire's thing.

[00:26:41] [SPEAKER_02]: And so what man was also listen, even if you don't have any archaeological site.

[00:26:48] [SPEAKER_02]: More archaeological sites they don't have any evidence is it doesn't mean that the lack of evidence isn't.

[00:26:54] [SPEAKER_02]: It doesn't have no evidence of fire.

[00:26:56] [SPEAKER_02]: So go out there and explore your archaeological site using this technology.

[00:27:01] [SPEAKER_02]: Only we'll find more we'll start understanding the origins in this spatial temporal distribution of fire throughout.

[00:27:09] [SPEAKER_02]: We both in the general, we're not the last one.

[00:27:11] [SPEAKER_02]: So we don't know what I'm working with.

[00:27:13] [SPEAKER_02]: It's not the only business.

[00:27:14] [SPEAKER_02]: Or you need to go and be teach.

[00:27:17] [SPEAKER_02]: But probably evidence is a fire there.

[00:27:19] [SPEAKER_02]: I never noticed because there was this evening with the show.

[00:27:21] [SPEAKER_02]: So the open up outdoors for us to say, well, it's any guys go out and we might find something that you're not expecting.

[00:27:29] [SPEAKER_02]: And not only in terms of space, but also in terms of time.

[00:27:32] [SPEAKER_02]: So you can go earlier, but you also can go lower, meaning that we can go down to two million or even lower.

[00:27:39] [SPEAKER_02]: To improve our disproved, they're cooking like processes.

[00:27:42] [SPEAKER_02]: So did we use it really co-glomerate as humans?

[00:27:47] [SPEAKER_02]: So that's kind of the brilliant range of what this everyone, the broad impact as whatever any core in mind or this point is to mean.

[00:27:59] [SPEAKER_00]: And of course, that was a few years ago an artificial intelligence now as advanced.

[00:28:04] [SPEAKER_00]: He's going to continue to advance and I'm curious.

[00:28:07] [SPEAKER_00]: How do you see AI based methods transforming the field of archaeology and the future?

[00:28:13] [SPEAKER_00]: Or do any other ancient human behaviors or indeed technologies that you think could benefit from this approach.

[00:28:19] [SPEAKER_00]: So AI might be able to reveal about our history.

[00:28:23] [SPEAKER_03]: Actually, we use this technology.

[00:28:26] [SPEAKER_03]: Well, I'll just find human behavior.

[00:28:30] [SPEAKER_03]: Okay, actually, we use this method and we show the overlay, the overall set of the four other, other than the years ago,

[00:28:44] [SPEAKER_03]: how the cognitive ability to go to do and to get the core.

[00:28:53] [SPEAKER_03]: One example of the failure to meet some stones and get tools to the take a new thing.

[00:29:01] [SPEAKER_03]: So we can use this.

[00:29:03] [SPEAKER_03]: To learn about the answer to us.

[00:29:07] [SPEAKER_02]: It was the first game for us was like, mumbling because we all engaged into these models and AI and what to do and not then mumbling and rops and all sorts of different things.

[00:29:18] [SPEAKER_02]: And so when we realized that we could actually be essentially on this first application for research on core, which like 400,000, more than the anatomical human that we realize that we could they were able to keep the rocks and then depending on the heating temperature.

[00:29:35] [SPEAKER_02]: They were able to be used to make one sort of two years, right, is another sort of tools.

[00:29:41] [SPEAKER_02]: So basically, that's called abstract thinking and for us like when it happened, we using any I on some rocks and suddenly we look at human.

[00:29:48] [SPEAKER_02]: We haven't actually been thinking meaning that you have to think in advance, but if you're taking one rock like this, you call to put it under the moon.

[00:29:57] [SPEAKER_02]: We talked about a lot of things. So you have to put it at a certain depth under the file, it's actually created an environment that reaches a certain temperature so you can put it into either a blade or just a chain.

[00:30:10] [SPEAKER_02]: And that makes you different in terms of the use of the ability of a rock. So we could clear to see a different creation between those types of rocks and the temperature that we heat it, being very wrong with them to be heated.

[00:30:24] [SPEAKER_02]: And then for us, it was like, what would mean it can we do?

[00:30:27] [SPEAKER_02]: It's like, oh, oh, oh, when did we have to go?

[00:30:29] [SPEAKER_03]: That's way too much. We're all about it, so I'd say the environment, this work that we understand human, whatever.

[00:30:39] [SPEAKER_03]: And we're completely new and we're going to change the energy and update information from the application.

[00:30:50] [SPEAKER_02]: Yeah, yeah, yeah.

[00:30:52] [SPEAKER_02]: There are different, I mean, both cases, there are new implications, but we're different, yeah, consequently.

[00:30:58] [SPEAKER_02]: Yeah, yeah.

[00:30:59] [SPEAKER_02]: The way we see the, the way but in the basement that's, well, I think getting to your question, I think that there's still a lot to do based on this data set in terms of the archaeology.

[00:31:11] [SPEAKER_02]: I was probably as far as the app that you can have any different place and there's also very dangerous because archaeology, there's a lot of there's a black market going around so we can use this as an act.

[00:31:27] [SPEAKER_02]: Because if you have something that is what we think, oh, it's worth nothing and they stole it from somewhere. They did the picture and suddenly it's like the home is a third or whatever.

[00:31:37] [SPEAKER_02]: It starts working like millions of dollars in the black market and so it's a day and years as long as not everything will be just putting a model of feeling that this can't from a hamster so it can't and everyone or even while for it.

[00:31:51] [SPEAKER_02]: It has social intrigues which is waiting beyond that. So there are a lot of things that people are doing to establish patterns.

[00:32:00] [SPEAKER_02]: I think this is sort of the large scale models with a lot of data will give you information about how people were handling technology in the health technology involved for a time.

[00:32:09] [SPEAKER_02]: Even at this crazy idea, the model is that can generate future products based on ancient technology. So you could have like pick an iPhone or whatever.

[00:32:20] [SPEAKER_02]: The model, I only think that is practical and it also, and we designed that.

[00:32:27] [SPEAKER_02]: For what we've learned from there's still a lot to learn from using AI in our ecology because that's the advantage of our colleges, what we call the, the home of the unknowns.

[00:32:38] [SPEAKER_02]: Because there's a large data thing.

[00:32:41] [SPEAKER_02]: But we don't know how things relate to each other because archaeology is very local.

[00:32:45] [SPEAKER_02]: Oh, yeah, there's an archeological site here and I'm focusing on this. I might wait something about that.

[00:32:50] [SPEAKER_02]: But it's it there's no correlation and it's a continuum even evolution and human culture is a continuum.

[00:32:57] [SPEAKER_02]: It's not like we stopped in one place and he got isolated. He got influences or to leave they've been there around them because unfortunately it's pain because of the South America.

[00:33:06] [SPEAKER_02]: No, they're in there. They've done what they've done.

[00:33:10] [SPEAKER_02]: Well, that all the next change of culture everywhere and it still does with unfortunately some situation but this work makes you and this exchange of information that we have been doing for hundreds of thousands of years.

[00:33:27] [SPEAKER_02]: And those patterns is something we don't know and we like to export that thing.

[00:33:31] [SPEAKER_02]: And we are able to create a very few role on understanding those patterns that are being known of the unknown.

[00:33:37] [SPEAKER_03]: Oh, I don't hear the past year in Google, well, traditional machine learning to get the unknown going to see the earth above me.

[00:33:48] [SPEAKER_03]: In archaeology, it tells the faculty about that.

[00:33:51] [SPEAKER_03]: I'm going to tell the results and we've got a challenge with the real and let's say both of us take it to my daughter.

[00:34:01] [SPEAKER_03]: They're dealing with much better records than the course.

[00:34:04] [SPEAKER_03]: And what is also depends on the quality of the data but seeing that moving from the machine with the code did counting.

[00:34:15] [SPEAKER_03]: And it was there.

[00:34:16] [SPEAKER_03]: Yeah, and being just going to go over.

[00:34:21] [SPEAKER_03]: Do you remember on?

[00:34:23] [SPEAKER_00]: Of course you mentioned earlier, initially faced skepticism from archaeologists, especially regarding the potential findings at that everyone quarry.

[00:34:33] [SPEAKER_00]: So I've got to ask, what was the most surprising discovery that your team made there anything that you can share around there?

[00:34:41] [SPEAKER_02]: Yeah, I think yeah.

[00:34:42] [SPEAKER_02]: I think that this method is it requires a lot of money investment in international investment.

[00:34:52] [SPEAKER_02]: And people couldn't believe and they still don't.

[00:34:57] [SPEAKER_02]: But as is always the typical position that we get when we approach new technologies, there's a matter of if it's an archaeology and so I think there's sometimes what the lacking is the right.

[00:35:09] [SPEAKER_02]: So you can get that step to it will be very difficult to argue against the mouthy bone.

[00:35:16] [SPEAKER_02]: And behind there's models which you don't know is like perfectly old les employions, which I also follow but is the physicist chemist.

[00:35:27] [SPEAKER_02]: So computational.

[00:35:29] [SPEAKER_02]: But I think it's it still needs some time to get more in a need to get people that do all the informatics that are in the international agenda and sometimes difficult because when what would the people do if they started to work in on this.

[00:35:45] [SPEAKER_02]: Are they going to be highlighted bit that company was the money that is given to our geology.

[00:35:50] [SPEAKER_02]: For example, we do in science in the sense of natural sciences and allow it.

[00:35:56] [SPEAKER_02]: Yet that is of hundreds dollars to run that in archaeology doesn't exist.

[00:36:02] [SPEAKER_02]: It doesn't get 10,000 dollars just like it's neat to approach it.

[00:36:05] [SPEAKER_02]: These are the month consumption on our lap.

[00:36:08] [SPEAKER_02]: So you see the difference.

[00:36:11] [SPEAKER_02]: The distance right to just make these database was like 10,000 dollars.

[00:36:15] [SPEAKER_02]: Just to measure the raw inspection.

[00:36:19] [SPEAKER_02]: And so that makes a huge difference this barrier.

[00:36:24] [SPEAKER_02]: You need to have people that need to simply need to have the resources.

[00:36:29] [SPEAKER_02]: And you need to have the interest.

[00:36:30] [SPEAKER_02]: So people were like, okay I've been working and so facing the set this is like, I've been saying that this doesn't have fire for the last 20 years.

[00:36:37] [SPEAKER_02]: Who are we going to kind of tell me that isn't it's probably as fire based on something they already I don't even have an explain what you're talking about.

[00:36:43] [SPEAKER_02]: And so that is sort of the reaction we get.

[00:36:46] [SPEAKER_02]: This is on one side something that the baller technology on it you from the interpretation point of view is I mentioned before, is that if it matches what people think you're good.

[00:36:57] [SPEAKER_02]: But it's a waste of time and money and resources.

[00:36:59] [SPEAKER_02]: But if it doesn't match, how many value did they that and how do you go without all the set this isn't that people help money.

[00:37:05] [SPEAKER_02]: And that's.

[00:37:07] [SPEAKER_03]: I want to let the multiple.

[00:37:09] [SPEAKER_03]: So that's the score of is finding the population on paper and the.

[00:37:13] [SPEAKER_03]: Yes, okay, we talked on what you want to do most.

[00:37:19] [SPEAKER_03]: But we're question it is a good good confidence in the client.

[00:37:24] [SPEAKER_03]: But this is all that.

[00:37:26] [SPEAKER_03]: And then the question of God or ask is if.

[00:37:34] [SPEAKER_00]: I was able to unlock some of those patterns from our history.

[00:37:38] [SPEAKER_00]: What implication do you think the discovery will have on our understanding of early human innovation and adaptation and how do you think these findings could even reshape the narratives of human evolution is anything that excites you here.

[00:37:54] [SPEAKER_02]: I think that this sort of.

[00:37:57] [SPEAKER_02]: We've seen nothing last year, the Nobel Prize winning was so on top of on the genetics of the Neon of all.

[00:38:04] [SPEAKER_02]: I think it's not yet it's still very early stages and what we contribution to is a small drop on that and showing that it works on its able.

[00:38:19] [SPEAKER_02]: It's still a far.

[00:38:21] [SPEAKER_02]: He requires a lot more momentum more people we get to get more people engaged from different fields to come together and talk and build those interactions and dynamics and start pushing this area.

[00:38:34] [SPEAKER_02]: It's just under that we don't understand definitely is a very something.

[00:38:38] [SPEAKER_02]: And we don't know where to go and what to do not us as our little project, but when you look large tip we don't know what's having the good thing which that will mentor and people doing it.

[00:38:50] [SPEAKER_02]: And then we will look away that's the exciting thing of.

[00:38:54] [SPEAKER_02]: And he does have the potential to to wish we know the same way as DNA in the past and the like the following spell is like for the years ago.

[00:39:03] [SPEAKER_02]: So you know so it does have a potential.

[00:39:06] [SPEAKER_02]: He just needs to be fully explored look at our everybody's days and yeah, yeah, we certainly essentially have the ending of the year with all those and so on and we're talking about evolution I think it does they are this potential.

[00:39:17] [SPEAKER_02]: It just needs to find.

[00:39:21] [SPEAKER_02]: Yeah, that's all.

[00:39:21] [SPEAKER_02]: So, so.

[00:39:23] [SPEAKER_03]: So, you know what I'm about to do what people want.

[00:39:27] [SPEAKER_03]: But the next question is why let the product question.

[00:39:31] [SPEAKER_03]: Oh, wrong.

[00:39:39] [SPEAKER_00]: The other big question of course is how do you ensure reliability and accuracy of those AI driven results because these are incredibly passionate community in archaeology.

[00:39:51] [SPEAKER_00]: Especially around such complex and data sensitive areas there and anything that you can share around ensuring that reliability and accuracy is it all about the mask because that's what people call an archaeo with.

[00:40:04] [SPEAKER_03]: Well, no, it's not a little bit of the first thing you could get the most accurate experiments that they will.

[00:40:13] [SPEAKER_03]: Yeah, it was thought it means something better you get to do.

[00:40:16] [SPEAKER_03]: Yeah, you know, the moment you ever get the moment to ever get the out to very important.

[00:40:23] [SPEAKER_03]: Tell us them to care you can ensure the other but the well-obeility and the health and the nature is relatively better.

[00:40:30] [SPEAKER_03]: The step is to the one.

[00:40:33] [SPEAKER_03]: And it's been driven the optimal component of the boarded pipeline.

[00:40:40] [SPEAKER_03]: Okay, and here.

[00:40:42] [SPEAKER_03]: And that's it everyone to ensure that it is working.

[00:40:47] [SPEAKER_03]: He is proven on vision experiment because if you don't have the right decision, experiment.

[00:40:53] [SPEAKER_03]: Which is not related to the model development.

[00:40:56] [SPEAKER_03]: It is difficult to call.

[00:40:59] [SPEAKER_03]: Constitence and get to commit it with a very well.

[00:41:02] [SPEAKER_03]: Okay, yeah.

[00:41:03] [SPEAKER_03]: And when you also develop the pipeline, it is very important to try to reduce any bias.

[00:41:12] [SPEAKER_03]: Don't there's development using the appropriate and both decision, you know, midpoints and investigation methods.

[00:41:20] [SPEAKER_02]: What a, I also do that is bring more data which humans can't do.

[00:41:26] [SPEAKER_02]: And so then we taking the video get subjectivity which is always associated by telling you.

[00:41:34] [SPEAKER_02]: I mean, people would just argue, he's like, oh, next.

[00:41:36] [SPEAKER_02]: Like, no, it's not my zone is a chick. No, it's all the chick.

[00:41:39] [SPEAKER_02]: Then we start to do it and you can have several coffees sitting in the puppy office.

[00:41:43] [SPEAKER_02]: And also gives a global vision of all the data.

[00:41:48] [SPEAKER_02]: It's not only the data there, we have locally which people will have.

[00:41:52] [SPEAKER_02]: I've served to the metabolism as saying, well, those are my archaeological sites.

[00:41:57] [SPEAKER_02]: And that's what I know about it.

[00:41:59] [SPEAKER_02]: So they might tend to forget the bit picture and they AI,

[00:42:03] [SPEAKER_02]: it's able to bring that up together and find potants.

[00:42:06] [SPEAKER_02]: If people that they've never came up, so it might be a social renewal idea like why it those guys came here.

[00:42:13] [SPEAKER_02]: Why did Romans came to whatever.

[00:42:16] [SPEAKER_02]: Why did they brought the second things?

[00:42:18] [SPEAKER_02]: So it talks to us, laying more that now reliability and accuracy we can only apply to the map.

[00:42:27] [SPEAKER_02]: But for the people, David and a penman wrote this book about thinking slowly getting sounds.

[00:42:35] [SPEAKER_02]: And we think that we can only rely on people because we're buying it.

[00:42:39] [SPEAKER_02]: So the reliability of that the word reliability and accuracy.

[00:42:45] [SPEAKER_02]: It's been from the mathematical and that we haven't read before the people.

[00:42:50] [SPEAKER_02]: Well, question mark.

[00:42:54] [SPEAKER_00]: A question I've got to ask you both is what future projects or areas of research?

[00:43:00] [SPEAKER_00]: Are you most excited about pursuing next?

[00:43:03] [SPEAKER_00]: Particularly with that intersection of AI and archaeology.

[00:43:07] [SPEAKER_00]: Your passion for the subject really comes to life today and I'd be interested in what exercise you both about this and where it's heading.

[00:43:15] [SPEAKER_03]: Okay, it's a lot of stuff.

[00:43:17] [SPEAKER_03]: My name is I'm a working on a digital download and I'm a journalist.

[00:43:23] [SPEAKER_03]: I want to get you here in a learning world.

[00:43:26] [SPEAKER_03]: I've learned from the context of the industry guy, I'm playing in ML page.

[00:43:30] [SPEAKER_03]: And I'm going to do an important quality of the knowledge and the opportunity.

[00:43:37] [SPEAKER_03]: And maybe I cannot get a digital specific project.

[00:43:44] [SPEAKER_03]: But maybe it will be in the front of the museum by one of the young and one of this.

[00:43:49] [SPEAKER_03]: And I'll just ask you how many years you've been following me.

[00:44:26] [SPEAKER_02]: we say it was measured, but I also viewed it with an DNA in our body care.

[00:44:30] [SPEAKER_02]: A lot of we did this yesterday, Chairman Dunn II, right?

[00:44:32] [SPEAKER_02]: Images everywhere to you.

[00:44:34] [SPEAKER_02]: It's an equation with the problem with that whenever you look at it,

[00:44:38] [SPEAKER_02]: on that specific, on all the more, but any sort of material is always

[00:44:41] [SPEAKER_02]: even we live on a 3D space.

[00:44:44] [SPEAKER_02]: And so everything is 3D, so a picture, a 2D picture,

[00:44:49] [SPEAKER_02]: we take a lot of information that you state, and even if you take a picture like this, I'm

[00:44:52] [SPEAKER_02]: going to answer for it.

[00:44:53] [SPEAKER_02]: So, while I have this project on Talbond, and try to look for the early stages of the

[00:44:59] [SPEAKER_02]: investigation.

[00:45:00] [SPEAKER_02]: So we spent thousands of dollars just 3D scanning bones from crazy places like Romania

[00:45:05] [SPEAKER_02]: and British, we went to London Natural Museum at Natural History in London.

[00:45:11] [SPEAKER_02]: We scan dozens of bones to spend almost three years cleaning up those bones on the scanning.

[00:45:18] [SPEAKER_02]: We put them into to see whether the information that you see in 3D would give any sort of

[00:45:25] [SPEAKER_02]: information about the investigation.

[00:45:28] [SPEAKER_02]: And we've very out completely because the models were not picking up the signals.

[00:45:33] [SPEAKER_02]: And we could three years just to clean the beta and then one year of testing.

[00:45:41] [SPEAKER_02]: And we've failed to complete it.

[00:45:42] [SPEAKER_02]: And so 3D is still a challenge, is still a challenging non-MODA model in AI.

[00:45:47] [SPEAKER_02]: 3D is important to know because all of these features, because everything means something that

[00:45:53] [SPEAKER_02]: was to on a specific angle that was all this specific way, because when you tell things like

[00:45:59] [SPEAKER_02]: this, you're not called things the same way.

[00:46:01] [SPEAKER_02]: And even almost a lot of the sensors on that they had the handwork system, the cognitive

[00:46:06] [SPEAKER_02]: abilities were different.

[00:46:08] [SPEAKER_02]: The weight, the whole thing.

[00:46:09] [SPEAKER_02]: So it's very interesting if you give this rock, the different people that will hold the

[00:46:13] [SPEAKER_02]: necessary ability of information that is hidden on each check that was removed from this.

[00:46:19] [SPEAKER_02]: So we need information in 3D.

[00:46:22] [SPEAKER_02]: I and 3D is very challenging, and it's super computers to train those models.

[00:46:27] [SPEAKER_02]: So that's one thing, 3D models and 3D AI.

[00:46:30] [SPEAKER_02]: Ever seen anything like that?

[00:46:32] [SPEAKER_02]: So back to my computer future.

[00:46:33] [SPEAKER_02]: Another one is cultural evolution, cultural style technology and illusionary relation between them.

[00:46:39] [SPEAKER_02]: So in recent testing, we're done of people that have been working on cultural evolution and

[00:46:47] [SPEAKER_02]: science.

[00:46:48] [SPEAKER_02]: So one of the big things would be cultural technology, genomics, putting all together in the same

[00:46:54] [SPEAKER_02]: package and see our things related to each other.

[00:46:57] [SPEAKER_02]: And there's been more of a prize winners and science dedicated to one of those thoughts.

[00:47:02] [SPEAKER_02]: We don't want infinite.

[00:47:03] [SPEAKER_02]: Yeah, I think that those combining those large data set is what AI really do that.

[00:47:12] [SPEAKER_02]: And that can reach out to what we know about the future.

[00:47:17] [SPEAKER_02]: And I just gave a few examples, but it can be like hundreds of thousands more examples because

[00:47:22] [SPEAKER_02]: a very new emerging.

[00:47:25] [SPEAKER_02]: What about?

[00:47:26] [SPEAKER_03]: Well, I'm looking at this.

[00:47:28] [SPEAKER_03]: I've been doing a good job of modernity and combine everything.

[00:47:33] [SPEAKER_03]: Yeah, you have a lot of images and signals.

[00:47:37] [SPEAKER_03]: Let's start with our thoughts on the data.

[00:47:40] [SPEAKER_03]: And I feel that no thoughts in this problem, we can get with AI from this data.

[00:47:45] [SPEAKER_03]: I don't know, we can get new, the mutuality, all the good.

[00:47:51] [SPEAKER_02]: For example, and there's also a technical challenge.

[00:47:53] [SPEAKER_02]: But for example, we got one like the logical side then was an update for hundreds of years.

[00:47:57] [SPEAKER_02]: So how many deal with this data?

[00:47:59] [SPEAKER_02]: Where they send populations, what do they look like?

[00:48:02] [SPEAKER_02]: So you have a geographer, geographical coordinates.

[00:48:06] [SPEAKER_02]: But in temporal resolution, within things like this, you can have 10 different populations

[00:48:11] [SPEAKER_02]: in the same place, especially in the Levant.

[00:48:14] [SPEAKER_02]: So that's kind of a very challenging system to work with.

[00:48:19] [SPEAKER_02]: And there are many other challenges that I think once they're both, some of the technical

[00:48:24] [SPEAKER_02]: others find them mental.

[00:48:26] [SPEAKER_02]: But I think this is just the beginning of something that could be a big, big, big, big, big, big, big, big, big, big, big, big things.

[00:48:30] [SPEAKER_02]: We're very excited to join this from the get-go.

[00:48:33] [SPEAKER_02]: And that is like, and yeah, that's looking to be more on our people.

[00:48:39] [SPEAKER_00]: Yeah, exciting times ahead.

[00:48:42] [SPEAKER_00]: But people listening or reading the article will accompany this podcast episode.

[00:48:47] [SPEAKER_00]: Maybe they want to find out more information about both of you.

[00:48:50] [SPEAKER_00]: Obviously I will link to your link to you.

[00:48:53] [SPEAKER_00]: So people can find you.

[00:48:54] [SPEAKER_00]: Is that a website or anything that you'd like to point people to that want to find out more information?

[00:49:00] [SPEAKER_02]: Yeah, I can send you my website.

[00:49:02] [SPEAKER_00]: Perfect. Well, thank you so much for joining me today.

[00:49:05] [SPEAKER_00]: I hope you're going to ignite a conversation for once, not about looking forward, but looking backwards.

[00:49:11] [SPEAKER_00]: And what we can learn. But say, be great, or hear more about your journey.

[00:49:15] [SPEAKER_00]: As it evolves and the things that you uncover, so I'd love to get you back on.

[00:49:18] [SPEAKER_00]: Maybe next year, see how things are evolving.

[00:49:21] [SPEAKER_00]: But thanks for bringing this topic to life today. I really appreciate your time.

[00:49:25] [SPEAKER_03]: And because what I'm much more forward is to see the virtual point related.

[00:49:27] [SPEAKER_03]: You talk with a pleasure on the topic again. Yeah.

[00:49:31] [SPEAKER_02]: In terms of this conversation, we should be able to do this.

[00:49:33] [SPEAKER_02]: Yeah, and I'm going to read a pleasure. Thank you, Martin.

[00:49:35] [SPEAKER_02]: We're the invitation and I'll also thank you, Doug, or inviting you.

[00:49:38] [SPEAKER_00]: As we've discussed, the past has never been more accessible.

[00:49:41] [SPEAKER_00]: Thanks to the integration of AI in archaeological research.

[00:49:45] [SPEAKER_00]: And listening to my guest today, I think it challenges our assumptions about early human

[00:49:50] [SPEAKER_00]: fire use. But also, of our lights, the transformative potential of

[00:49:54] [SPEAKER_00]: into disciplinary collaboration and emerging technologies.

[00:49:58] [SPEAKER_00]: But also gets me thinking about what other mysteries of our ancient past might be unlocked with

[00:50:04] [SPEAKER_00]: the cutting edge technologies. And this is the part where the microphone gets put in from a view.

[00:50:10] [SPEAKER_00]: I'd love to hear your thoughts on the future of archaeology. And AI, please join the conversation

[00:50:16] [SPEAKER_00]: share with me your insights by emailing me techplogwriteroutrook.com, Twitter,

[00:50:22] [SPEAKER_00]: LinkedIn, Instagram just at Neil Seekyves. But that's it for today. So,

[00:50:26] [SPEAKER_00]: stay curious, gang, keep exploring the intersections of technology and history.

[00:50:31] [SPEAKER_00]: And most importantly of all, just put in me a guide to my own,

[00:50:34] [SPEAKER_00]: I'm no other great guest. I'll see you all then. Bye bye.