Governing Agentic AI: Subodh Shetty & Greg Adams

WEBVTT

00:00.231 --> 00:02.052
[SPEAKER_00]: Listeners of the randomness of nothing podcast.

00:02.072 --> 00:09.295
[SPEAKER_00]: We talk about a variety of topics on this show, but I always think that technology is at the forefront of everything that we utilize in life.

00:09.675 --> 00:14.377
[SPEAKER_00]: It's changed the way we live, the change of music we listen to, our music, the shows we watch.

00:14.477 --> 00:16.778
[SPEAKER_00]: It touches every single aspect of our lives now.

00:16.898 --> 00:22.225
[SPEAKER_00]: And I have here a very high level expert on here, one of the co-founders of Sintara AI, Savoni Shetty.

00:22.265 --> 00:22.926
[SPEAKER_00]: Thank you very much.

00:23.126 --> 00:23.747
[SPEAKER_02]: Thank you, Rishal.

00:23.767 --> 00:31.977
[SPEAKER_02]: This is a pleasure having being on your podcast and absolutely excited to talk to you about AI and the different things that we're doing at Sintara.

00:32.177 --> 00:33.218
[SPEAKER_00]: I'd very much appreciate it.

00:33.238 --> 00:43.722
[SPEAKER_00]: I think sometimes when people hear it, they, you know, there's there's the connotation of of any time something is big at a moment of an inflection point, it brings out emotions and people.

00:43.782 --> 00:52.286
[SPEAKER_00]: So how do you when it comes to this particular segment, let's talk about your background first and then get into the, how you ended up getting it ultimately into this company that you cova.

00:52.406 --> 00:53.047
[SPEAKER_02]: Sure.

00:53.347 --> 00:55.568
[SPEAKER_02]: My background is pretty much building enterprise software.

00:55.908 --> 01:05.759
[SPEAKER_02]: For the past 30 years, I've been in this area, the DC metro area, throughout my career, started off as a software engineer, and then kind of build up on that stack.

01:06.079 --> 01:10.924
[SPEAKER_02]: And then 10 years ago, I kind of moved to like the FinTech and applied AI space.

01:11.615 --> 01:14.378
[SPEAKER_02]: And that's what has been my experience with the past 10 years.

01:14.999 --> 01:21.886
[SPEAKER_02]: And before I joined Centauri, it was Amazon that I was deleting some of the ML engineering initiatives there.

01:22.227 --> 01:25.610
[SPEAKER_02]: And that's been a last met the co-founder of

01:29.093 --> 01:40.538
[SPEAKER_02]: And it was at Amazon that we talked about different things that AI was doing because AI was really hyped up and think of ways we can kind of put some godrails.

01:40.558 --> 01:45.920
[SPEAKER_02]: That's what is interesting because when we talk about AI people are kind of, you know.

01:46.260 --> 01:53.025
[SPEAKER_02]: a little bit of apprehensive because they don't want AI's doing and they don't want AI to go rogue, you know, do things that you don't want to.

01:53.446 --> 02:02.472
[SPEAKER_02]: That's what the thought was to kind of put those guardrails in there so that people are there or what the AI is doing, not like before it does even after.

02:02.552 --> 02:05.655
[SPEAKER_02]: So you kind of have an audit trail as well as preventance,

02:07.556 --> 02:09.217
[SPEAKER_02]: So that's not the very high level what we do.

02:09.477 --> 02:12.939
[SPEAKER_02]: You already put the god rules so that we kind of kind of controls.

02:13.540 --> 02:14.160
[SPEAKER_02]: Right.

02:14.620 --> 02:17.042
[SPEAKER_00]: So real quick, what's the name come from?

02:17.082 --> 02:19.743
[SPEAKER_00]: Because I couldn't find anything that was related to Centaur.

02:19.783 --> 02:22.625
[SPEAKER_00]: I just watched, I just saw the Odyssey yesterday.

02:22.665 --> 02:25.547
[SPEAKER_00]: So I was like, maybe there's something I could like prepare myself for that.

02:25.567 --> 02:27.348
[SPEAKER_00]: I hear maybe that name if it's A.G. Greed.

02:27.728 --> 02:28.989
[SPEAKER_00]: And I couldn't hear anything.

02:29.029 --> 02:30.249
[SPEAKER_00]: So how did the name actually come from?

02:30.369 --> 02:31.330
[SPEAKER_02]: Yes, the name itself.

02:31.490 --> 02:35.372
[SPEAKER_02]: It's like interesting because Tara is like a...

02:36.753 --> 02:54.718
[SPEAKER_02]: a star yeah and we want to do something where it shows like a north star which kind of points beautiful you know where the direction is and that's what every centaur is is one of the domains but yeah okay okay I didn't know if it was some ancient latin verbs that I just couldn't figure out right so that was always fascinating

02:55.198 --> 03:07.271
[SPEAKER_00]: What, you know, and so I think that when people think of AI, they automatically start think of privacy concerns, job losses, you know, all of the negative things that are associated with it, you know, that bring out emotions of people.

03:07.431 --> 03:14.338
[SPEAKER_00]: And so, but it's also very useful, too, because the speed of which things can be done, the complication of things that were

03:14.598 --> 03:19.602
[SPEAKER_00]: you know, previously unattainable to people who didn't have a particular skill set, now have it at their disposal.

03:19.922 --> 03:37.534
[SPEAKER_00]: When it comes to self-governance, I'm curious how that actually works, is it just because that's the next level to get to it or is to stay above, you know, hey, we're putting guardrails upon ourselves early on, just to let us know that this is, we are protecting, you know, proprietary and in certain items related to businesses and people.

03:37.674 --> 04:01.872
[SPEAKER_02]: Yes, so the thing is like with AI, like you said, it's just a tool and what people kind of forget is like when internet came on or when Google came on even then people were afraid that it's going to take away the jobs and maybe in it but also gave a lot of people tools to kind of progress and right now what's happening is very similar so the big previously like maybe like a few years ago.

04:02.232 --> 04:05.734
[SPEAKER_02]: only the big companies could build applications or software.

04:05.854 --> 04:07.295
[SPEAKER_02]: Now, anyone can do it.

04:07.375 --> 04:11.618
[SPEAKER_02]: Even someone sitting remotely with no resources, we can build up and that.

04:11.958 --> 04:13.699
[SPEAKER_02]: So it's not kind of that powers you.

04:13.940 --> 04:25.407
[SPEAKER_02]: Yeah, yes, it's like a kind of shift, you know, where people are previously just used to, you know, having like six months to a front-way and then building things now, it makes it very simple.

04:25.507 --> 04:30.370
[SPEAKER_02]: At the same time, when AI builds things, you are kind of skeptic and you're afraid.

04:30.690 --> 04:35.395
[SPEAKER_02]: And there are some godrails, what people typically do is before it goes to an AI.

04:35.635 --> 04:45.264
[SPEAKER_02]: For example, you might have noticed if you're using chatGPT, if you ask for medical advice, you are, it says, I'm not a doctor, and absolutely absolutely.

04:45.284 --> 04:48.167
[SPEAKER_02]: There's a godrails which prevents the query from going in.

04:48.527 --> 04:51.390
[SPEAKER_02]: So what we are doing is the output of it.

04:51.690 --> 04:53.651
[SPEAKER_02]: because people still kind of get around it.

04:53.991 --> 04:55.611
[SPEAKER_02]: They say, oh, I am a physician.

04:55.992 --> 04:57.552
[SPEAKER_02]: You can still provide this diagnosis.

04:57.572 --> 04:58.372
[SPEAKER_02]: There are these hacks.

04:58.713 --> 05:06.135
[SPEAKER_02]: But what we do is once the LLM answers it, we prevent the diagnosis from going back because we are saying, oh, no, we kind of enjoyed this.

05:06.535 --> 05:09.657
[SPEAKER_02]: So our guard rails is the outdoor level of the air.

05:09.757 --> 05:10.457
[SPEAKER_00]: That's wonderful.

05:10.797 --> 05:11.457
[SPEAKER_00]: Greg, how are you?

05:11.637 --> 05:12.617
[SPEAKER_00]: Greg, how are you, Rashad?

05:12.817 --> 05:18.239
[SPEAKER_00]: I'm doing well, and we're just obviously nerding out right now, which is a fun moment when it comes to artificial intelligence.

05:18.419 --> 05:29.901
[SPEAKER_00]: So the questions I have for both of you is that for the person that may be, you know, I don't necessarily have the skills, but this has been something that's really opened up some doors for me.

05:30.221 --> 05:34.002
[SPEAKER_00]: It's democratized a lot of things for people that were previously inaccessible.

05:34.302 --> 05:37.443
[SPEAKER_00]: You don't necessarily have to go to that high level, you know,

05:38.684 --> 05:44.131
[SPEAKER_00]: $70,000 a year, you know, coding, you know, flat school, so to speak, to really get these skills.

05:44.151 --> 05:46.934
[SPEAKER_00]: You just have to put in the work necessary to make these things work for you.

05:47.194 --> 05:57.147
[SPEAKER_00]: And what, you know, Ceboda and I were talking about, where small businesses, medium sized businesses that were previously locked out of getting these really cutting-edge tools, now have access and resources to.

05:57.307 --> 05:58.368
[SPEAKER_01]: Yeah, absolutely.

05:58.808 --> 06:12.676
[SPEAKER_01]: And so, you know, what AI is really opening up new frontiers, new capabilities for a lot of people and a lot of the pregist limitations are gone, the learning curves are much accelerated and

06:13.956 --> 06:29.463
[SPEAKER_01]: It's just a huge disruptor across the board creating huge opportunities and a few challenges and the great thing about what we do here at Centair is that we help people leverage more of the opportunities But doing it in a safer, more responsible way addressing some of the major challenges

06:29.923 --> 06:30.844
[SPEAKER_01]: that come with AI.

06:30.964 --> 06:31.265
[SPEAKER_00]: Right.

06:31.325 --> 06:41.236
[SPEAKER_00]: So, obviously, you know, there are challenges when it comes to, because typically, in most governments, government moves slow private entities move very fast.

06:41.636 --> 06:44.700
[SPEAKER_00]: How do you all, in a broad sense, I'm not going to ask for details, obviously?

06:45.140 --> 07:02.226
[SPEAKER_00]: navigate those tricky waters by being an inventive innovative company without sitting back and having those those type of situations where somebody's ready legislation, somebody's trying to slow progress of very fast moving technologies because to your point, you know, it can get dicey in those conversations.

07:02.466 --> 07:14.618
[SPEAKER_01]: Yeah, well, we can never wait for legislation because that's always the last one to show up at the party and by the time the legislators get caught up was what's happening in the technologies that already moved on to the next thing.

07:14.718 --> 07:23.146
[SPEAKER_01]: So the thing is, you've got to understand how to fully seize the opportunities that AI brings about, but what are the things that people are most concerned about?

07:23.466 --> 07:24.127
[SPEAKER_01]: And they are,

07:24.447 --> 07:46.297
[SPEAKER_01]: basically preventing the AI from hallucinating or going rogue, every beak, there's a new publicized event of AI going sideways and how embarrassing or scary or risky that is, but for every AI that's publicized, many more are not, either because people don't talk about it because it's embarrassing, but actually the majority is just because they've not even been detected yet.

07:46.637 --> 08:02.768
[SPEAKER_01]: And so, how do you let people use more AI for more mission critical operations and functions with an organization, but without basically unleashing it to be able to do harm or go sideways?

08:03.089 --> 08:03.689
[SPEAKER_01]: So how do you secure it?

08:03.709 --> 08:04.970
[SPEAKER_01]: I'm sure you're in the crowd.

08:05.290 --> 08:13.112
[SPEAKER_01]: And now some of the AI companies have been developing some, you know, what I would call embryonic tooling to help control some of their tools.

08:13.152 --> 08:18.173
[SPEAKER_01]: And you ask them to also, of course, our agent would never go rogue, but unfortunately that's not true.

08:18.753 --> 08:18.913
[SPEAKER_01]: Right.

08:19.113 --> 08:30.236
[SPEAKER_01]: And so we saw this kind of did a, a horizon scan of this industry and said, well, really what, what the industry's needing the most, what companies are yearning for.

08:30.976 --> 08:47.437
[SPEAKER_01]: is a technology neutral horizontal governance layer that can cut across all of your different text acts regardless of whether you've got a few agents training on Amazon, someone and traffic, you know, someone Google and some maybe, you know, using some other technology.

08:47.757 --> 08:55.180
[SPEAKER_01]: you don't have different governance tooling across all of these different functionality that are only specific to those two agents.

08:55.560 --> 09:04.783
[SPEAKER_01]: What you need is a horizontal layer that cuts across all your text stack allow you to govern these agents and make sure that they never misbehave.

09:05.224 --> 09:07.724
[SPEAKER_01]: That's the idea and the concept behind Centera.

09:08.085 --> 09:12.966
[SPEAKER_01]: And we've developed it, started testing it across many use cases in many different industries.

09:13.407 --> 09:15.487
[SPEAKER_01]: We've actually filed three patents around this,

09:17.508 --> 09:23.689
[SPEAKER_01]: And yeah, it's really exciting to be solving such a real problem that nobody else has solved elegantly yet.

09:23.830 --> 09:23.990
[SPEAKER_00]: Yeah.

09:24.130 --> 09:31.091
[SPEAKER_00]: And so the question I have for you, Sebode, is that they're particular industries that are in particular of need.

09:31.291 --> 09:38.093
[SPEAKER_00]: Obviously, it's across the board, hotels, airlines, production and manufacturing, utilize them very well.

09:38.393 --> 09:46.055
[SPEAKER_00]: But do you sit back and say, this industry has, you know, for lack of better term, kind of been in the stone age, it's a little bit, and it really is like, thank goodness for this

09:47.555 --> 09:49.777
[SPEAKER_02]: Yeah, so we are kind of platform agnostic.

09:49.937 --> 09:55.902
[SPEAKER_02]: It's almost like in front as a service and on like an infrastructure But at the same time, I would think of like right now.

09:55.922 --> 10:09.412
[SPEAKER_02]: We have a few pilots running on in a fintech area and Health industry so those two verticals are the ones which require a lot of compliance and that's where I think the guardrails actually makes that But at the same time you could see if

10:09.808 --> 10:18.290
[SPEAKER_02]: even for a very small company, like we talked earlier, where AI actually enables even small and a mom and pop shop to build their own applications.

10:18.650 --> 10:22.871
[SPEAKER_02]: Say you have a small travel agent who wants to build an application which will just book tickets.

10:23.251 --> 10:24.911
[SPEAKER_02]: So you're looking at that, right?

10:24.931 --> 10:31.433
[SPEAKER_02]: That's a very common use case, instead of hiring someone else, you do it in house, you blur it, and guess for it, that doesn't have any garbage.

10:31.853 --> 10:38.117
[SPEAKER_02]: So you say, oh, book me a ticket from DC to LAX, you'll just book it even if it's like $2,000.

10:38.898 --> 10:42.681
[SPEAKER_02]: So what the current AI guardrails will do is prevent even the booking.

10:42.721 --> 10:45.122
[SPEAKER_02]: It comes before you send it to the AI.

10:45.363 --> 10:49.425
[SPEAKER_02]: You say, if the ticket is going to say, you don't have the guardrails on the outputs.

10:49.485 --> 10:51.627
[SPEAKER_02]: What we do is, say, I say, book a ticket.

10:51.947 --> 10:57.753
[SPEAKER_02]: from DC to LAX and then I've a policy behind it saying do not book it if it's more than a thousand dollars.

10:58.294 --> 10:59.655
[SPEAKER_02]: So that's where the normal is.

10:59.676 --> 11:01.197
[SPEAKER_02]: It's in a very, very simple use case.

11:01.517 --> 11:10.367
[SPEAKER_02]: So this will be a use case for a very moment pop shop and this we can kind of, you know, extrapolate to like the fintech and the healthcare industry.

11:10.843 --> 11:12.624
[SPEAKER_02]: and it doesn't use benefits there.

11:12.964 --> 11:17.505
[SPEAKER_00]: Yeah, and you know, while you were saying that, you know, it makes perfect sense that the pharmacies, right?

11:17.545 --> 11:28.368
[SPEAKER_00]: Because there's so many, they deal with so much medication of so many different people and some medicines are very similar and there's certain doses of a certain medicine that if you go, I mean, we all, you know, have health things of something.

11:28.708 --> 11:33.091
[SPEAKER_00]: You have to be very careful in their process thousands of prescriptions a day.

11:33.411 --> 11:50.761
[SPEAKER_00]: So if there's an agent that has those guard rails that says don't prescribe Mr. Johnson more than this milligrams and then you check that guard rail to make sure that prescription didn't go among that certain amount, now you have a good situation where you can speed up the transaction process for people while at the same time protecting patients.

11:51.021 --> 11:51.622
[SPEAKER_02]: Yeah, exactly.

11:51.742 --> 12:03.856
[SPEAKER_02]: In that use case, say Mr. Jones here is like, is there's a guardrail right in front of him saying then do not prescribe like more than 20 mg, but then what if he does like three different prescriptions of 5 mg each.

12:04.296 --> 12:07.240
[SPEAKER_02]: Yes, so that's where the guardrails come in after the outcome.

12:07.460 --> 12:17.086
[SPEAKER_00]: beautiful, beautiful and so I don't you know and I think sometimes when people get scared about certain things they don't realize how much that could save a life, how much that could treat a diagnosis or medicine.

12:17.786 --> 12:29.013
[SPEAKER_00]: I want to pit it real quick because I'm curious about the data centers like what to the average person that's hearing these things you know you know if they watch the sci-fi movies and they immediately think of the worst thing in the world right you know like

12:29.393 --> 12:33.859
[SPEAKER_00]: droids and bots that are going to be walking around like a little smith eye robot kind of situation right?

12:34.219 --> 12:42.509
[SPEAKER_00]: What exactly would people say and then maybe Greg will submit to you about data centers and things like that to kind of you know put that in a more realistic sense?

12:42.729 --> 12:47.475
[SPEAKER_00]: Well, um, I think the issue is that Yes, right.

12:47.715 --> 12:48.376
[SPEAKER_00]: Yeah, of course.

12:48.516 --> 12:48.857
[SPEAKER_00]: Of course.

12:52.311 --> 13:07.824
[SPEAKER_00]: You know, it's crazy sometimes that we love to use things that are easily out of disposal to use them for our benefit, but then we hesitate on the real world use, I mean, you love the fact that sometimes you have the thermostat in your house that can adjust to every single bedroom.

13:08.104 --> 13:09.806
[SPEAKER_00]: You love your curated musical list.

13:10.066 --> 13:18.513
[SPEAKER_00]: You love the recommendations from your, from your various streaming platforms and services or your car gets a diagnosis and you can start your phone, your car from your phone.

13:18.613 --> 13:22.855
[SPEAKER_00]: All those things are AI agents that people love and use every single day.

13:23.195 --> 13:32.680
[SPEAKER_00]: But then when they're kind of told kind of like the wizard behind the curtain so to speak, and it's real to them, all of a sudden they're like, whoa, whoa, whoa, whoa, I'm a little bit like, well, you use it every single day.

13:33.021 --> 13:34.682
[SPEAKER_00]: You use it every single day.

13:35.022 --> 13:36.002
[SPEAKER_00]: Even you just don't know it.

13:36.122 --> 13:38.864
[SPEAKER_02]: Yeah, I'm even coming back to your data center point, right?

13:39.244 --> 13:43.666
[SPEAKER_02]: So data centers weren't even there, like before Amazon started doing AWS.

13:47.322 --> 13:49.245
[SPEAKER_02]: what the importance of the data centers.

13:49.605 --> 13:54.832
[SPEAKER_02]: But then when a lot of people started using data centers, you kind of see how it actually is better for the environment.

13:54.852 --> 14:01.542
[SPEAKER_02]: Because before that, people used to have it on-prem and there was a super expensive lot of waste where it's now with Amazon Cloud.

14:01.762 --> 14:04.005
[SPEAKER_02]: You're kind of sharing it, so it's more efficient.

14:04.185 --> 14:08.186
[SPEAKER_02]: Even with this, the benefits far outweigh whatever the drawbacks are.

14:08.607 --> 14:14.148
[SPEAKER_02]: For example, you talked about using AI and different industries like thermostat and stuff, what about it?

14:14.549 --> 14:15.989
[SPEAKER_02]: It's using like diagnosis.

14:16.289 --> 14:18.510
[SPEAKER_02]: Right now, medical is so expensive.

14:18.910 --> 14:23.172
[SPEAKER_02]: What if AI can make it so easy that you can just go and get your diagnosis for free?

14:23.532 --> 14:26.873
[SPEAKER_02]: It's so many unsap benefits and different industries.

14:27.297 --> 14:42.059
[SPEAKER_00]: right and to also because you know 80% of the population does not live in an urban district area downtown so if you need telemedicine if you have self-diagnostic medical tools then you could you know get mailed and get access to the doctor you know it I don't I

14:42.319 --> 14:49.624
[SPEAKER_00]: Sometimes we think that where we live at is respective to how everybody lives, whether it's high speed, internet, access to certain amenities.

14:49.924 --> 14:52.306
[SPEAKER_00]: That's not the case for a lot of people, every state.

14:52.686 --> 15:00.371
[SPEAKER_00]: In your own state, you can go 10, 20 miles outside of your own crowded downtown district, and the next thing you know you find out is it's completely different world.

15:00.771 --> 15:05.054
[SPEAKER_00]: And so those tools that are democratized to allow it across the board are very valuable.

15:05.074 --> 15:09.517
[SPEAKER_00]: A farmer, you know, with crops and things of that nature and weather patterns,

15:11.218 --> 15:11.679
[SPEAKER_02]: Absolutely.

15:11.719 --> 15:14.782
[SPEAKER_02]: And especially when you say agriculture, you know, A.I.

15:14.802 --> 15:19.527
[SPEAKER_02]: can help like, you know, improve their the efficiency of their land.

15:19.627 --> 15:22.750
[SPEAKER_02]: So you can have these crops at a low price range.

15:22.810 --> 15:23.551
[SPEAKER_02]: And then guess what?

15:23.571 --> 15:25.893
[SPEAKER_02]: Everyone benefits because food prices actually will drop.

15:26.234 --> 15:27.955
[SPEAKER_02]: So there's so much untat potential.

15:28.155 --> 15:28.376
[SPEAKER_00]: Right.

15:28.576 --> 15:29.397
[SPEAKER_02]: It's yes.

15:29.677 --> 15:30.117
[SPEAKER_02]: Absolutely.

15:30.278 --> 15:30.498
[SPEAKER_00]: Yes.

15:30.698 --> 15:30.898
[SPEAKER_00]: Right.

15:33.040 --> 15:38.401
[SPEAKER_00]: What do you say now to the, and I'm going to pivot a bit and maybe this question you guys don't know, maybe just curious.

15:38.742 --> 15:40.682
[SPEAKER_00]: What's going to do for higher education now?

15:40.722 --> 15:46.223
[SPEAKER_00]: And I ask because now you have, you know, this was not really a big field obviously when I graduated from high school.

15:52.125 --> 15:54.826
[SPEAKER_00]: we're all going to date ourselves a little bit on this call, right?

15:55.127 --> 16:02.831
[SPEAKER_00]: But having said all that, this is different because before it was accessible, you know, it gradually made itself to certain segments of the population.

16:03.211 --> 16:08.634
[SPEAKER_00]: Now you have a graduation, you know, of all these kids who all they know is this type of stuff.

16:08.954 --> 16:14.398
[SPEAKER_00]: All they know is the internet, cell phones, technology streaming, they don't have yellow pages, they don't have blockbusters.

16:14.799 --> 16:18.702
[SPEAKER_00]: What's this changing for the educational over kid who's like a sophomore in high school?

16:19.082 --> 16:22.725
[SPEAKER_00]: What they should specialize in when it comes to their field and what's the future for?

16:22.845 --> 16:30.611
[SPEAKER_02]: Yeah, so that's very interesting because right now we have actually spoken a lot of these kids, you know, like in some of these, the conference that we go to.

16:31.332 --> 16:36.596
[SPEAKER_02]: And these kids in high school itself, they are so used to AI that they are building stuff right there.

16:37.036 --> 16:51.026
[SPEAKER_02]: So at the same time, they need to be aware of doing it in the ethical way, so that's what the education would be on how to use it in the right way and make sure it's like things like age is going grow, you know, you need those guardrails, you know, they can actually do it today.

16:51.066 --> 16:54.769
[SPEAKER_02]: I mean, I'm sure even all of us are using charity, but instead of Google now, now,

16:55.650 --> 17:05.058
[SPEAKER_02]: And this kid's throwing up they're only used to all that and where as we were not, we were like, we kind of put it in the air.

17:05.118 --> 17:06.779
[SPEAKER_00]: I was there when you were out, man.

17:06.799 --> 17:08.161
[SPEAKER_00]: I was there when you were out.

17:08.441 --> 17:09.322
[SPEAKER_00]: Remember that website?

17:10.002 --> 17:15.303
[SPEAKER_02]: or NetScape, NetScape, navigate, it was the one with the browser, I don't know of those days.

17:15.783 --> 17:36.548
[SPEAKER_02]: So think of like the kids now, they grew up at like, Chad GPT, so it's endless possibilities and they can learn so much on their own and whereas we had to kind of go to like do your graduate school, undergrad, graduate school, to learn all these things whereas right now they have it in your fingerprints, figure tips and you can actually educate yourself by someone

17:37.528 --> 17:44.015
[SPEAKER_02]: you can ask no question is a stupid question, you know, no one is touching you, you can ask whatever you want and help with learning as well.

17:44.055 --> 17:48.900
[SPEAKER_02]: So I think when these guys come out of even high school, they are like so much more at once than most of us.

17:48.960 --> 17:49.521
[SPEAKER_02]: It's amazing.

17:49.601 --> 17:50.182
[SPEAKER_02]: It's amazing.

17:50.542 --> 17:51.143
[SPEAKER_00]: It's amazing.

17:51.383 --> 17:53.245
[SPEAKER_00]: So yes, yes, absolutely.

17:53.445 --> 17:55.728
[SPEAKER_00]: Yes, I've said that for years, man.

17:56.888 --> 17:57.048
[SPEAKER_00]: Yep.

17:57.248 --> 17:57.489
[SPEAKER_00]: Yes.

17:57.809 --> 17:58.109
[SPEAKER_00]: Yes.

17:59.150 --> 17:59.771
[SPEAKER_00]: Right.

17:59.791 --> 18:02.794
[SPEAKER_00]: You're getting that customized list that we talked about, right?

18:02.874 --> 18:03.735
[SPEAKER_00]: Like you're okay.

18:03.775 --> 18:13.664
[SPEAKER_00]: So if I really do want to go in a particular field, I've got this particular field of skills that are better suited for my personality and more alert, I'm more involved and more engaged.

18:14.004 --> 18:28.111
[SPEAKER_00]: You know, putting somebody who wants to be, you know, a car mechanic in, you know, a history class may not be beneficial for them long term to be able to get the best output out of them.

18:28.231 --> 18:28.352
[SPEAKER_00]: Right.

18:28.372 --> 18:29.232
[SPEAKER_00]: That's just an example.

18:29.272 --> 18:30.913
[SPEAKER_00]: That's kind of, you know, just me throwing that out there.

18:31.273 --> 18:50.402
[SPEAKER_00]: But if you curated this is how you're going to be able to work on these particular set of skills from a computer and it's going to be less physically taxing for you to do certain things, there's a benefit to them to be in these these piloted programs to to get these software and these skills that they need, it ultimately could lead them to have a higher success rate in graduations.

18:50.642 --> 18:51.383
[SPEAKER_00]: Yeah.

18:51.423 --> 18:56.445
[SPEAKER_02]: And even this, even if you're a car mechanic and suddenly now you develop an interest in history,

18:56.858 --> 19:00.902
[SPEAKER_02]: it's easy to pay with it because you have this patient teacher who's going to teach everything that you want to know.

19:01.103 --> 19:01.583
[SPEAKER_00]: Exactly.

19:01.683 --> 19:02.144
[SPEAKER_00]: Exactly.

19:03.045 --> 19:07.810
[SPEAKER_00]: When you're talking about self-confidence and, you know, obviously, you know, there's the guardrails that are in place.

19:08.130 --> 19:12.215
[SPEAKER_00]: Do you think that it's better that when they have these data centers that there are, you know,

19:12.535 --> 19:35.119
[SPEAKER_00]: that there's not one company that has the monopoly over the AI systems and that could probably be beneficial for people to feel more comfortable that there's not one central company that owns all the artificial intelligence and these data centers or these technology today's use because a lot of companies are pivoting towards it that there are those of the guardrails that are in place because there are so many companies that are pivoting towards this particular set of technology.

19:35.399 --> 19:36.980
[SPEAKER_00]: Although yours is the first one that's already

19:38.627 --> 19:40.267
[SPEAKER_02]: Yeah, so there are a lot of options.

19:40.307 --> 19:51.230
[SPEAKER_02]: Even with data centers where you have decentralized CPUs, so it's actually like little machines which are running, which are actually not in machine like processor on different, but it's almost like a blockchain, right?

19:51.270 --> 19:52.631
[SPEAKER_02]: It's everything is decentralized.

19:52.651 --> 19:54.531
[SPEAKER_02]: So not one person is controlling everything.

19:54.911 --> 19:56.852
[SPEAKER_02]: So with the data centers, there are also,

19:57.212 --> 20:00.835
[SPEAKER_02]: I think there's company called Akash which does that which is decentralized compute.

20:01.175 --> 20:03.877
[SPEAKER_02]: It's like a data center, but it's all running in different machines.

20:04.037 --> 20:04.878
[SPEAKER_02]: So it's like spread out.

20:05.298 --> 20:13.184
[SPEAKER_02]: And that's what is going to be there because a lot of these companies would try to monopolize, but then there will be a lot of other decentralized options as well that we have.

20:13.604 --> 20:14.345
[SPEAKER_02]: And at the same time,

20:14.685 --> 20:18.166
[SPEAKER_02]: what we are doing even for our governance when we kind of audit it.

20:18.486 --> 20:23.727
[SPEAKER_02]: One of the options is we do have the other history available as a blockchain transaction.

20:23.767 --> 20:27.547
[SPEAKER_02]: So that's the decentralized part where any transaction cannot be changed.

20:27.968 --> 20:32.908
[SPEAKER_02]: Once I say that, yes, agent has done it, has done a certain action.

20:33.249 --> 20:36.109
[SPEAKER_02]: That's persisted in the blockchain, which is actually decentralized.

20:36.129 --> 20:40.690
[SPEAKER_02]: So anyone from anywhere can actually see and verify that that transaction was done by this agent.

20:41.090 --> 20:42.891
[SPEAKER_02]: So those guardrails are also there, right?

20:42.911 --> 20:45.692
[SPEAKER_02]: It's like not controlled by one specific entity.

20:45.872 --> 20:46.833
[SPEAKER_00]: Craig, I want to ask you this.

20:47.233 --> 20:52.496
[SPEAKER_00]: How is the US has its certain level of laws that are, you know, that this is a being implemented by?

20:52.816 --> 20:58.899
[SPEAKER_00]: Are you guys right now just servicing US customers or have you, what's the navigation taking place when you start dealing with?

20:59.279 --> 21:02.760
[SPEAKER_00]: the EU laws, or even a Canada, or somewhere else.

21:02.800 --> 21:12.444
[SPEAKER_00]: Well, we're on paper, people would think, well, this should be something pretty simple to expand, but you have to navigate some of these different respective laws of them catching up to technologies like this.

21:12.664 --> 21:13.024
[SPEAKER_00]: Beautiful.

21:14.545 --> 21:15.506
[SPEAKER_00]: Right, right, right.

21:15.646 --> 21:16.427
[SPEAKER_00]: Oh yeah, beautiful.

21:16.667 --> 21:20.211
[SPEAKER_00]: Yeah, because that can get very tricky and dicey without having something like that in place.

21:20.231 --> 21:24.956
[SPEAKER_00]: You have to deploy a legal team, you know, and then God forbid somebody has to go there and implement it.

21:25.016 --> 21:31.343
[SPEAKER_00]: And then if you're dealing with mountains of data, if you don't have something automated, eventually something's gonna slip through the cracks, right?

21:31.403 --> 21:37.009
[SPEAKER_00]: So having somebody like yourself and your team being able to sit back and say, no, we already have the right governance in place.

21:37.469 --> 21:38.670
[SPEAKER_00]: we've already planned for this.

21:38.890 --> 21:43.792
[SPEAKER_00]: It's only a benefit to the clients, the customers, and that's one thing that's taken off their plane.

21:44.112 --> 21:45.832
[SPEAKER_00]: Yeah, definitely.

21:45.932 --> 21:56.797
[SPEAKER_00]: And more importantly, it's always about your guys or the step ahead of asking those questions for people because you already have already done your due diligence and your proper research.

21:57.157 --> 22:04.000
[SPEAKER_00]: So that's not even going to be up for a discussion to be had because you're already going to bring that to the table knowing that this is part of what

22:04.720 --> 22:07.385
[SPEAKER_00]: What does the future live for the expansion of the company?

22:07.586 --> 22:07.926
[SPEAKER_00]: Good for.

22:10.245 --> 22:23.633
[SPEAKER_02]: I mean, because right now, maybe we are tied to a theme cloud, in the future, I would see it being like cloud, totally cloud, agnostic, as well as data center agnostic, we will actually a pretty question.

22:23.673 --> 22:36.701
[SPEAKER_02]: So, the decentralized data warehouse is actually one of the ideas that we were exploring as well, so that we can kind of be completely agnostic, and it's like totally decentralized and not monopolized by certain companies, so it's not constant.

22:37.033 --> 22:42.156
[SPEAKER_00]: Yeah, and you don't have the beauty about it is just that at the end of the day this opens up a plethora of opportunities for people.

22:42.216 --> 22:50.720
[SPEAKER_00]: If there was ever a time to really, you know, dive your feet into that new venture, learn that new set of skills, be able to always have something at your fingertips.

22:50.740 --> 22:56.603
[SPEAKER_00]: I mean, the process has been democratized more at the most basic level than it's ever been in the history of this country, right?

22:56.983 --> 22:57.303
[SPEAKER_00]: You know,

22:57.603 --> 23:04.324
[SPEAKER_00]: people who typically had location or it doesn't really exist with this type of situation.

23:04.484 --> 23:06.505
[SPEAKER_00]: Obviously, there are other situations people could be dealing with.

23:06.845 --> 23:11.565
[SPEAKER_00]: But the availability has never been as available as it is right.

23:11.685 --> 23:15.506
[SPEAKER_00]: And that is a huge benefit to people right now, particularly for skills and businesses.

23:15.566 --> 23:16.746
[SPEAKER_00]: You know, it's a huge benefit.

23:16.806 --> 23:20.367
[SPEAKER_00]: It really is, is there anything that I haven't covered?

23:20.707 --> 23:24.608
[SPEAKER_00]: And then the second question I had is, where can people find Centaur?

23:24.928 --> 23:26.028
[SPEAKER_00]: They don't need to show to do this.

23:28.727 --> 23:29.168
[SPEAKER_00]: Absolutely.

23:29.348 --> 23:29.708
[SPEAKER_00]: Beautiful.

23:29.888 --> 23:30.329
[SPEAKER_02]: Yeah.

23:30.429 --> 23:43.958
[SPEAKER_02]: And to that, I just want to say, if you will be hearing a lot of these news articles where agents actually go rogue and do things that they are not supposed to at that point, which one them to remember, Sintara.io is the one which could have prevented it.

23:44.179 --> 23:44.779
[SPEAKER_00]: Absolutely.

23:44.899 --> 23:50.043
[SPEAKER_00]: And you know, I guess I was really pumped for this call because there was, you know, this is right up my alley.

23:50.103 --> 23:56.688
[SPEAKER_00]: And obviously, I think that when you look at the future, it's always great to talk to people who are the innovators because, you know, when you guys are on,

23:57.088 --> 24:02.914
[SPEAKER_00]: You know, on that news channel, I'm glad I got a chance to get an interview before the line got really crowded and sought to go.

24:03.334 --> 24:07.978
[SPEAKER_00]: So I want to thank you very much for your time and it's been an honor to have you on the round of something podcast.

24:08.139 --> 24:08.799
[SPEAKER_00]: Thank you very much.

24:08.959 --> 24:09.200
[SPEAKER_00]: Thank you.