WEBVTT
00:00:00.160 --> 00:00:02.160
Welcome to another episode of FinTech Talks.
00:00:02.240 --> 00:00:06.000
I'm Sanjeep Khalida and I'm here at FinTech Meetup.
00:00:06.160 --> 00:00:08.720
And I'm so happy to be here with my friend Dave Birch.
00:00:08.880 --> 00:00:10.480
Dave, would you mind please introducing yourself?
00:00:10.800 --> 00:00:11.919
Hi, I'm Dave Birch.
00:00:12.000 --> 00:00:15.199
I'm an author, advisor, and commentator on digital financial services.
00:00:15.439 --> 00:00:18.160
And yeah, I'm very flattered to be invited, Sanjeev.
00:00:18.239 --> 00:00:18.960
Thank you for having me.
00:00:19.199 --> 00:00:24.960
No, it it it's it's I've I've been around the block a little bit and I never miss a chance to get a chance to talk to you.
00:00:25.039 --> 00:00:36.079
And and I always love your takes on on the state of the world or state of the industry that are non-traditional, and I I find them hilarious as well.
00:00:37.039 --> 00:00:43.759
And and um so you know what what what's what's what what's interesting to you at the world in the world right now in in our industry?
00:00:44.079 --> 00:00:47.119
Uh you mean by interesting, you mean what are clients paying me for?
00:00:47.280 --> 00:00:50.240
That's like that's my that's my definition of interesting.
00:00:50.479 --> 00:00:56.320
Um I'd say there's probably kind of three areas at the moment where so obviously AI, yeah.
00:00:56.640 --> 00:01:07.680
So uh you can't get over the fact that AI is the most disruptive technology we're going to be dealing with, and that's true for fintech, payments, everything we're involved in.
00:01:07.760 --> 00:01:09.120
So AI.
00:01:10.480 --> 00:01:20.079
The second thing, and this may not be as obvious in the US, but certainly outside the US, the issue of the resilience of payment systems has become a real issue.
00:01:20.560 --> 00:01:22.480
Uh national and regional resilience.
00:01:22.799 --> 00:01:24.959
And so, and for me it's very interesting.
00:01:25.439 --> 00:01:33.280
So, uh doing a lot of work on resilience, okay, and you won't be surprised to hear me say, of course, the other thing is identity.
00:01:33.439 --> 00:01:35.760
You know, we still haven't solved the identity problem.
00:01:35.920 --> 00:01:36.079
Yeah.
00:01:36.239 --> 00:01:43.040
And actually, because of well, because of AI, apart from anything else, yeah, um, identity is back up the agenda at the moment.
00:01:43.280 --> 00:01:46.879
So yeah, they're they're they're probably the top three things I'm spending time on.
00:01:47.120 --> 00:01:52.000
And uh, and I I I would love to dive into each of those, but I uh let me start with resilience.
00:01:52.079 --> 00:01:57.200
I uh like how how what does it what do you when you say resilience that can be taken many different ways?
00:01:57.439 --> 00:02:03.040
How how are your clients or how are our nations thinking about resilience?
00:02:03.519 --> 00:02:07.040
Well, I mean, first of all, there are more there's more pressure on resilience now.
00:02:07.120 --> 00:02:12.080
You've got things like DORA, uh the uh European Digital Operational Resilience Act.
00:02:12.240 --> 00:02:17.280
Regulators want banks to be more resilient, they want the financial infrastructure to be more resilient.
00:02:17.759 --> 00:02:29.599
You've had some disasters recently, uh, not just kind of earthquakes and floods and things like that, but the power going out in Spain, for example, or and that's kind of got people focused on it.
00:02:29.680 --> 00:02:32.159
Um you've got a lot of regular day-to-day stuff.
00:02:32.240 --> 00:02:38.400
You know, you go to the supermarket and the car machines aren't working, or banks go down for a couple of days, you can't make payments.
00:02:39.199 --> 00:02:48.800
But of course, there's also now the issue of well, what if we have to use regional or national solutions?
00:02:48.960 --> 00:02:54.080
What if we can't use proprietary we can't use they say proprietary, they mean American.
00:02:54.159 --> 00:02:55.680
We can't use American solutions.
00:02:55.759 --> 00:02:56.000
Yeah.
00:02:56.159 --> 00:02:58.560
How are we gonna be more now?
00:02:59.280 --> 00:03:09.919
Historically, you know, when you were talking about resilience, like in a bank or something like that, yeah, you'd be talking about okay, how much does it cost to get us 99% resilience?
00:03:10.400 --> 00:03:11.520
Okay, that's not good enough.
00:03:11.599 --> 00:03:14.560
The r how much is it gonna cost to get us 99.9?
00:03:14.800 --> 00:03:15.919
That's 10 times as much.
00:03:16.000 --> 00:03:16.159
Yeah.
00:03:16.560 --> 00:03:19.199
99.99, 10 times more than that.
00:03:19.280 --> 00:03:23.840
Yeah, you know, pretty soon you hit a wall, you can't afford to go any further.
00:03:24.080 --> 00:03:27.680
And yet you're utterly dependent on the same underlying systems.
00:03:28.000 --> 00:03:32.400
So the focus on resilience now has got a lot to do with diversity of infrastructure.
00:03:32.719 --> 00:03:41.520
Because now we have things like stable coins, you know, distributed ledgers, digital assets, CBDC coming in in some places.
00:03:41.919 --> 00:03:47.520
We've got in many parts of the world domestic networks that can be developed further.
00:03:48.719 --> 00:03:55.280
So I think resilience now is more about trying to look at what the strategic mix downstream will be.
00:03:55.439 --> 00:03:57.759
I mean, to take it like a very simplistic view.
00:03:58.080 --> 00:04:04.240
If the you know, if the settlement network goes down, can we just send things over stable coins instead?
00:04:04.479 --> 00:04:05.439
Or, you know, whatever.
00:04:05.680 --> 00:04:11.120
If the card network is down at the supermarket, can we go to an offline alternative?
00:04:11.280 --> 00:04:11.759
Yeah.
00:04:12.000 --> 00:04:15.199
I mean, obviously without interrupting the consumer experience.
00:04:16.000 --> 00:04:20.480
So, yeah, so it's but it's it's a priority now because just because of recent events.
00:04:20.560 --> 00:04:20.800
Yeah.
00:04:20.959 --> 00:04:23.839
So that's and it's fun and it's intellectually quite interesting.
00:04:23.920 --> 00:04:25.439
So I quite like that work actually.
00:04:25.680 --> 00:04:41.759
I I I I've also uh one thing that I'm I mean, we're seeing in the world is like there's a lot more I'd say like thinking about these payments rails in terms of sovereignty, and and that also fits into the resilience as well.
00:04:41.839 --> 00:04:48.240
Like almost like an I I I like how and I mean how how do you think about that?
00:04:48.480 --> 00:04:56.800
Well, I think in lots of places people have seen, I mean, not just the I mean everybody knows the key examples, UPI, yeah, PICs in Brazil.
00:04:57.040 --> 00:05:01.839
I mean, everybody looks at these and looks at the numbers, yeah, and whoa, you know, yeah.
00:05:02.160 --> 00:05:06.720
I mean, to a first approximation, everyone in Brazil has PICs, basically.
00:05:06.800 --> 00:05:14.160
So um so people are looking at that and saying, well, first of all, why can't we have a national solution like that?
00:05:14.319 --> 00:05:21.600
Yeah, which is uh universal, mandated, you know, everyone's got the same interfaces, etc., etc.
00:05:22.399 --> 00:05:28.560
But there's also the issue of what about alternatives that are being developed by the private sector in different places?
00:05:28.800 --> 00:05:31.279
This is why I mentioned stable coins, for example.
00:05:31.600 --> 00:05:45.199
So could we achieve let's imagine if you're in Southeast Asia, for example, could we achieve regional sovereignty you know, by interconnecting, for example, the real-time payment networks?
00:05:45.279 --> 00:05:49.839
Yeah, could we interconnect domestic card networks or domestic payment schemes?
00:05:50.160 --> 00:05:55.360
Or could we look at developing parallel rails you know, using stable coins and so on?
00:05:55.680 --> 00:06:03.120
And actually, the idea of using parallel rails to achieve higher levels of resilience than building continuous landscape, that's pretty interesting.
00:06:03.839 --> 00:06:06.399
Now, and also you said about regional things.
00:06:06.480 --> 00:06:15.759
I mean, if you look at the example of Scandinavia, you got MPE and VIP, so all of these things are coming together and interconnecting, I think you see different paths to resilience now.
00:06:15.920 --> 00:06:23.839
So uh and working out what strategy is going to work best in which nation/slash region, uh it's interesting work, you know.
00:06:24.000 --> 00:06:58.879
Yeah, and and um shifting now to AI it is um yeah, you know, I I'm I remember as you know for when different technology platforms come out, you know, there there's a bit of the the platform itself, then the tools you can get from the platform, and then how that changes even how like for for example, like we use computer files, but they're named after physical paper files, and and but they're they have nothing to do with one one or the other.
00:06:58.959 --> 00:07:05.439
But it's it's just sort of a way that we as humans are able to manage you know that this new archetype.
00:07:05.680 --> 00:07:06.480
No, I know exactly.
00:07:07.600 --> 00:07:09.920
This is the problem challenge.
00:07:10.000 --> 00:07:24.240
I mean, this is the this is what finance faces because once you go agentic, once it's agents that are doing these transactions, there's no particular reason why you would think that the existing infrastructure that we have is the best for them.
00:07:24.800 --> 00:07:27.120
In fact, you it's got it's almost certainly not, yeah.
00:07:27.920 --> 00:07:30.720
So, and you know, stable coins are an interesting example.
00:07:30.800 --> 00:07:30.959
Yeah.
00:07:31.120 --> 00:07:34.480
So one of the discussions on the stage here this morning.
00:07:34.720 --> 00:07:40.639
So we can make some great stablecoin solutions, and you have people like Tempo and so on that are building this stuff out right now.
00:07:41.199 --> 00:07:46.720
Once you can run stable coins over those rails, actually, you can run any digital asset over those rails.
00:07:46.959 --> 00:07:54.639
There are plenty of people who know more about finance than I do that say, actually, a US dollar stablecoin, maybe that's not the best.
00:07:55.040 --> 00:08:02.160
Maybe actually tokenized money market funds would be a better solution, or even tokenized treasury bills, or I mean there's other kind of things.
00:08:02.399 --> 00:08:05.360
So now you've got tokens that uh yield-bearing tokens.
00:08:05.439 --> 00:08:05.759
Yeah.
00:08:06.000 --> 00:08:15.199
So if you can exchange those seamlessly over those networks, yeah, why would you ever come out of those into fiat stable coins?
00:08:15.279 --> 00:08:16.480
Why would you ever bother with that?
00:08:16.639 --> 00:08:16.800
Yeah.
00:08:17.040 --> 00:08:24.000
So so the AI isn't just uh, you know, clever ways of doing what we do now, but with different processes.
00:08:24.399 --> 00:08:29.439
It's bringing into existence very different ways of doing business.
00:08:29.920 --> 00:08:34.559
And I, you know, I take a mild amount of credit.
00:08:34.799 --> 00:08:40.000
I mean, I wrote a thing a few years ago for Wired magazine actually about AI and banking.
00:08:40.159 --> 00:08:40.559
Yeah.
00:08:40.799 --> 00:08:44.159
I can't remember how this came about, but it wasn't about banking.
00:08:45.200 --> 00:09:00.799
And, you know, I looked at what the banks were saying about all of this, and I I just it didn't seem right to me, and I couldn't figure out like, why is this not like they're gonna use the most powerful technology ever invented, and if the and they're gonna use it to send slightly better targeted credit card mailings?
00:09:00.879 --> 00:09:02.240
I mean, that's doesn't you know what I mean?
00:09:02.320 --> 00:09:15.200
It doesn't so and I realize and I'm not the only person that thought this, but you know, but then I realized you know, there is a revolution coming in financial services, but it's because of customers using AI, it's not because of banks using AI.
00:09:15.679 --> 00:09:22.480
And so this idea that you have the customer's AI becomes your customer, not the customer anymore, if you see what I mean.
00:09:22.559 --> 00:09:23.200
Yeah, yeah.
00:09:23.600 --> 00:09:29.039
And if you take on board, like people are saying, the primary users of stable coins are gonna be AIs, yeah.
00:09:29.200 --> 00:09:40.799
Like when you put all this stuff together, you end up with a different kind of financial infrastructure, which again I think is absolutely fascinating and intellectually very interesting to try and work out where that's going.
00:09:40.960 --> 00:09:46.000
So yeah, the AI stuff I'm really interested in is customers using AI.
00:09:46.159 --> 00:09:54.639
And I mean, don't you think that's kind of I think maybe merchants, banks, I think maybe they don't know what's coming because they don't have a strategy.
00:09:54.879 --> 00:10:02.000
I mean, who has a strategy for customers becoming a thousand times smarter literally overnight?
00:10:02.080 --> 00:10:09.759
Yeah, you know, we have business processes that are I was thinking about car in I had to buy car insurance a couple of weeks.
00:10:10.080 --> 00:10:16.240
You're a like no sane person wants to spend their time comparing car insurance policies, right?
00:10:16.480 --> 00:10:39.200
And and even how do you you know, even when I I get the two best quotes from the guy, I download the 400 pages of like how am I even supposed to compare the like that stuff cannot survive the the advent of agentics, so yeah, so it means a different kind of well, not only does it mean a different kind of financial market infrastructure, it means frankly a different kind of web is coming as well.
00:10:39.279 --> 00:10:40.799
So no, I I'm fascinated.
00:10:40.960 --> 00:10:46.000
Like, am I interested in how the bank uses AI to replace some people in the call center?
00:10:46.159 --> 00:10:47.360
No, I couldn't care less.
00:10:47.440 --> 00:10:49.120
I mean, that's just not interesting.
00:10:49.360 --> 00:10:50.000
Yeah, yeah, no.
00:10:50.080 --> 00:11:17.840
And to me, I'm sure it's good business for some people, but and and and uh sort of maybe jumping to the next that that that third piece, which is also related to AI and and I I I think in terms of identity, and and you know, if you have these agents rather than customers, um yeah, and they're customer driven, but like then what what what is identity at that point?
00:11:18.159 --> 00:11:20.159
I mean you're absolutely right to to point to this.
00:11:20.240 --> 00:11:38.720
This is a this is a fundamental problem with what we have now, which I mean people recognize, but we haven't quite so unless you can work out the identity of an AI, who authorized it, yeah, its reputation, like it's unfortunate like unless you know this stuff, like it doesn't work.
00:11:38.879 --> 00:11:45.360
You draw on the whiteboard, oh well, my bot will open a savings account with your bank's bot, whatever.
00:11:45.440 --> 00:11:47.759
Like that's easy to draw on a whiteboard, yeah.
00:11:47.919 --> 00:11:49.759
But like you drop it down one level.
00:11:49.919 --> 00:11:54.480
Yeah, how does my bot know that you actually are the authorized bot for the bank?
00:11:54.720 --> 00:11:59.200
How does your bot know that I'm an authorized bot who's allowed to access bank accounts?
00:11:59.360 --> 00:12:02.320
How do you know I'm allowed to access on behalf of Dave Birch?
00:12:02.480 --> 00:12:08.159
How do you know what I'm allowed to do on behalf of like when you go down one level, it's really complicated and not fixed.
00:12:08.639 --> 00:12:13.200
So again, intellectually very fascinating as to what to do there.
00:12:13.440 --> 00:12:22.720
And like we kind of the way we semi-fix this at the moment is like you give the bot permission to sort of be you, yeah, but that's not really what we want.
00:12:22.879 --> 00:12:28.480
What we want is an infrastructure where it's a bot, but it's acting on my behalf.
00:12:28.639 --> 00:12:30.159
Like, do I know this is?
00:12:30.879 --> 00:12:38.000
I mean, this is a silly example, but like, does Robin Hood know that this is the real Jim Kramer bot?
00:12:38.320 --> 00:12:47.200
Is the real Jim Kramer bot authorized to access who says, you know, is it authorized on behalf of Dave Birch?
00:12:47.519 --> 00:12:49.440
What can it do on behalf of Dave Birch?
00:12:49.600 --> 00:12:50.159
You know, etc.
00:12:50.399 --> 00:12:50.639
etc.
00:12:51.039 --> 00:12:55.679
So because I like I'm a normal, I'm a normal consumer of financial services.
00:12:55.840 --> 00:12:56.159
Yeah.
00:12:56.320 --> 00:13:01.120
Like the cons most of them are boring, like payments and so on, which I don't want to do them.
00:13:01.200 --> 00:13:02.320
I want bots to do those.
00:13:02.399 --> 00:13:02.559
Yeah.
00:13:02.720 --> 00:13:07.519
And the ones that aren't boring are so baffling and complex that I want an AI to do it.
00:13:07.600 --> 00:13:08.639
I don't want to do it myself.
00:13:08.720 --> 00:13:09.279
Yeah.
00:13:09.519 --> 00:13:18.639
Would you, would you back, I mean, even if you just talk about things like like in the UK right now, it's March, it's coming up to the end of our tax year.
00:13:18.720 --> 00:13:18.799
Yeah.
00:13:19.120 --> 00:13:22.720
Normal middle class people like me have to look at our savings.
00:13:22.960 --> 00:13:26.799
We have like these savings accounts, the tax efficient savings accounts.
00:13:27.120 --> 00:13:30.000
So you you have to open, like, which one do I want?
00:13:30.159 --> 00:13:32.799
Do I want the existing one or do I want one of the new ones?
00:13:33.200 --> 00:13:35.600
I don't want I don't want to spend any time doing that.
00:13:35.679 --> 00:13:37.360
I want a bot to do that for me.
00:13:37.440 --> 00:13:44.559
And I wouldn't back myself to make a better choice of those than even the even the the kickkeeper.
00:13:45.039 --> 00:13:45.200
Okay.
00:13:47.120 --> 00:13:52.639
I wouldn't back myself to make a better choice of savings account than even the most rudimentary bot.
00:13:52.879 --> 00:13:53.600
You know what I mean?
00:13:53.759 --> 00:13:56.559
So so actually, I think that's stuff's coming quite quickly.
00:13:56.639 --> 00:13:56.720
Yeah.
00:13:57.039 --> 00:14:00.080
If the identity problem that you point out is solved, yeah.
00:14:00.240 --> 00:14:08.879
Now, I think one of the bonuses to solving that problem, and this may sound a little odd, but I think it's true, is fraud is completely out of control.
00:14:09.039 --> 00:14:09.360
Yeah.
00:14:09.600 --> 00:14:11.759
A lot of that fraud is caused by AIs.
00:14:11.919 --> 00:14:12.399
Yeah.
00:14:12.720 --> 00:14:16.639
It's not obvious to me that an AI is harder to fool.
00:14:16.720 --> 00:14:17.039
Yeah.
00:14:17.200 --> 00:14:18.879
Like because AIs don't live in our world.
00:14:19.039 --> 00:14:23.200
AIs don't live in the world of, oh, this does this look a bit like Jim Kramer?
00:14:23.279 --> 00:14:23.360
Yeah.
00:14:23.519 --> 00:14:27.519
Did Jim Kramer really say, you know, go and buy this stock or whatever?
00:14:27.679 --> 00:14:33.200
Like AIs live in a world of keys and certificates and encryption, digital signatures.
00:14:33.360 --> 00:14:42.320
And like I can make a fake video of Jim Kramer saying, oh, you should buy Dave Birch's stock, but I can't make a fake digital signature of Jim Kramer.
00:14:42.399 --> 00:14:43.759
Like math doesn't work like that.
00:14:44.240 --> 00:14:51.440
So actually, I think once we get this identity thing sorted out, I actually think that'll be a much safer infrastructure too.
00:14:52.080 --> 00:14:52.960
What's interesting?
00:14:53.200 --> 00:15:08.399
And and uh I I we're we're getting towards the end of our time, and and uh I I wanted to add, I would be remiss if I did not ask you about you know what what's been your experience like here at FinTech Meetup so far.
00:15:08.720 --> 00:15:10.399
Uh for me, it was great.
00:15:10.559 --> 00:15:14.080
I got they they put me on the main stage this morning.
00:15:14.320 --> 00:15:16.159
My panelists were fantastic.
00:15:16.240 --> 00:15:22.320
I mean, you saw that it was just it was just great, and it had the right mix of people, yeah.
00:15:22.480 --> 00:15:26.639
So that the conversation was I said to them, like I only say yes to panel.
00:15:26.720 --> 00:15:30.480
I mean, like I don't want to sound like too much of a prima donna, but I do do this quite a lot, you know.
00:15:30.960 --> 00:15:35.440
I like panels where I'm gonna learn stuff as well, you know, and so wow.
00:15:35.600 --> 00:15:41.759
I mean, in terms of like like knowledge per minute on that stage this morning, it was terrific.
00:15:41.919 --> 00:15:52.960
No one was parroting like PR stuff, and they were all really informed, really so great, yeah, but also it's a bit in less tangible, this kind of thing.
00:15:53.360 --> 00:15:55.600
But don't you think it just like has a nice atmosphere?
00:15:55.759 --> 00:15:59.360
Yeah, like I've been just walking around, bumping into people, chatting to people.
00:15:59.440 --> 00:16:00.639
Yeah, it's been lovely.
00:16:00.720 --> 00:16:02.799
Uh and also by the way, some of the stands here, yeah.
00:16:02.960 --> 00:16:09.279
I have to say the new kind of like base level of swag has gone up significantly.
00:16:09.440 --> 00:16:15.919
Like there's really nice stuff out there on the floor, and so maybe there isn't going to be a global recession after all.
00:16:16.799 --> 00:16:17.759
I love that.
00:16:17.840 --> 00:16:23.759
And and uh I you know I think we're once again we're reached reached the end of our time here.
00:16:23.919 --> 00:16:26.879
But I I would be I just want to thank you personally.
00:16:26.960 --> 00:16:32.480
Like I've known you for so many years, and you have had such an impact on me personally.
00:16:32.720 --> 00:16:38.000
Oh, you're very I I um I I love your intellectual curiosity.
00:16:38.159 --> 00:16:40.159
Um it's second to none.
00:16:40.559 --> 00:16:46.399
And the way that you talk about things and you personalize things, it it has been a North Star for me.
00:16:46.480 --> 00:16:50.159
And that's I once again for my heart wanna want to thank you for that.
00:16:50.480 --> 00:16:51.600
Thank you so much, Sanjeev.
00:16:51.840 --> 00:16:53.279
That means a lot to me coming from you.
00:16:53.840 --> 00:16:54.159
Thank you.
00:00:00.160 --> 00:00:02.160
Welcome to another episode of FinTech Talks.
00:00:02.240 --> 00:00:06.000
I'm Sanjeep Khalida and I'm here at FinTech Meetup.
00:00:06.160 --> 00:00:08.720
And I'm so happy to be here with my friend Dave Birch.
00:00:08.880 --> 00:00:10.480
Dave, would you mind please introducing yourself?
00:00:10.800 --> 00:00:11.919
Hi, I'm Dave Birch.
00:00:12.000 --> 00:00:15.199
I'm an author, advisor, and commentator on digital financial services.
00:00:15.439 --> 00:00:18.160
And yeah, I'm very flattered to be invited, Sanjeev.
00:00:18.239 --> 00:00:18.960
Thank you for having me.
00:00:19.199 --> 00:00:24.960
No, it it it's it's I've I've been around the block a little bit and I never miss a chance to get a chance to talk to you.
00:00:25.039 --> 00:00:36.079
And and I always love your takes on on the state of the world or state of the industry that are non-traditional, and I I find them hilarious as well.
00:00:37.039 --> 00:00:43.759
And and um so you know what what what's what's what what's interesting to you at the world in the world right now in in our industry?
00:00:44.079 --> 00:00:47.119
Uh you mean by interesting, you mean what are clients paying me for?
00:00:47.280 --> 00:00:50.240
That's like that's my that's my definition of interesting.
00:00:50.479 --> 00:00:56.320
Um I'd say there's probably kind of three areas at the moment where so obviously AI, yeah.
00:00:56.640 --> 00:01:07.680
So uh you can't get over the fact that AI is the most disruptive technology we're going to be dealing with, and that's true for fintech, payments, everything we're involved in.
00:01:07.760 --> 00:01:09.120
So AI.
00:01:10.480 --> 00:01:20.079
The second thing, and this may not be as obvious in the US, but certainly outside the US, the issue of the resilience of payment systems has become a real issue.
00:01:20.560 --> 00:01:22.480
Uh national and regional resilience.
00:01:22.799 --> 00:01:24.959
And so, and for me it's very interesting.
00:01:25.439 --> 00:01:33.280
So, uh doing a lot of work on resilience, okay, and you won't be surprised to hear me say, of course, the other thing is identity.
00:01:33.439 --> 00:01:35.760
You know, we still haven't solved the identity problem.
00:01:35.920 --> 00:01:36.079
Yeah.
00:01:36.239 --> 00:01:43.040
And actually, because of well, because of AI, apart from anything else, yeah, um, identity is back up the agenda at the moment.
00:01:43.280 --> 00:01:46.879
So yeah, they're they're they're probably the top three things I'm spending time on.
00:01:47.120 --> 00:01:52.000
And uh, and I I I would love to dive into each of those, but I uh let me start with resilience.
00:01:52.079 --> 00:01:57.200
I uh like how how what does it what do you when you say resilience that can be taken many different ways?
00:01:57.439 --> 00:02:03.040
How how are your clients or how are our nations thinking about resilience?
00:02:03.519 --> 00:02:07.040
Well, I mean, first of all, there are more there's more pressure on resilience now.
00:02:07.120 --> 00:02:12.080
You've got things like DORA, uh the uh European Digital Operational Resilience Act.
00:02:12.240 --> 00:02:17.280
Regulators want banks to be more resilient, they want the financial infrastructure to be more resilient.
00:02:17.759 --> 00:02:29.599
You've had some disasters recently, uh, not just kind of earthquakes and floods and things like that, but the power going out in Spain, for example, or and that's kind of got people focused on it.
00:02:29.680 --> 00:02:32.159
Um you've got a lot of regular day-to-day stuff.
00:02:32.240 --> 00:02:38.400
You know, you go to the supermarket and the car machines aren't working, or banks go down for a couple of days, you can't make payments.
00:02:39.199 --> 00:02:48.800
But of course, there's also now the issue of well, what if we have to use regional or national solutions?
00:02:48.960 --> 00:02:54.080
What if we can't use proprietary we can't use they say proprietary, they mean American.
00:02:54.159 --> 00:02:55.680
We can't use American solutions.
00:02:55.759 --> 00:02:56.000
Yeah.
00:02:56.159 --> 00:02:58.560
How are we gonna be more now?
00:02:59.280 --> 00:03:09.919
Historically, you know, when you were talking about resilience, like in a bank or something like that, yeah, you'd be talking about okay, how much does it cost to get us 99% resilience?
00:03:10.400 --> 00:03:11.520
Okay, that's not good enough.
00:03:11.599 --> 00:03:14.560
The r how much is it gonna cost to get us 99.9?
00:03:14.800 --> 00:03:15.919
That's 10 times as much.
00:03:16.000 --> 00:03:16.159
Yeah.
00:03:16.560 --> 00:03:19.199
99.99, 10 times more than that.
00:03:19.280 --> 00:03:23.840
Yeah, you know, pretty soon you hit a wall, you can't afford to go any further.
00:03:24.080 --> 00:03:27.680
And yet you're utterly dependent on the same underlying systems.
00:03:28.000 --> 00:03:32.400
So the focus on resilience now has got a lot to do with diversity of infrastructure.
00:03:32.719 --> 00:03:41.520
Because now we have things like stable coins, you know, distributed ledgers, digital assets, CBDC coming in in some places.
00:03:41.919 --> 00:03:47.520
We've got in many parts of the world domestic networks that can be developed further.
00:03:48.719 --> 00:03:55.280
So I think resilience now is more about trying to look at what the strategic mix downstream will be.
00:03:55.439 --> 00:03:57.759
I mean, to take it like a very simplistic view.
00:03:58.080 --> 00:04:04.240
If the you know, if the settlement network goes down, can we just send things over stable coins instead?
00:04:04.479 --> 00:04:05.439
Or, you know, whatever.
00:04:05.680 --> 00:04:11.120
If the card network is down at the supermarket, can we go to an offline alternative?
00:04:11.280 --> 00:04:11.759
Yeah.
00:04:12.000 --> 00:04:15.199
I mean, obviously without interrupting the consumer experience.
00:04:16.000 --> 00:04:20.480
So, yeah, so it's but it's it's a priority now because just because of recent events.
00:04:20.560 --> 00:04:20.800
Yeah.
00:04:20.959 --> 00:04:23.839
So that's and it's fun and it's intellectually quite interesting.
00:04:23.920 --> 00:04:25.439
So I quite like that work actually.
00:04:25.680 --> 00:04:41.759
I I I I've also uh one thing that I'm I mean, we're seeing in the world is like there's a lot more I'd say like thinking about these payments rails in terms of sovereignty, and and that also fits into the resilience as well.
00:04:41.839 --> 00:04:48.240
Like almost like an I I I like how and I mean how how do you think about that?
00:04:48.480 --> 00:04:56.800
Well, I think in lots of places people have seen, I mean, not just the I mean everybody knows the key examples, UPI, yeah, PICs in Brazil.
00:04:57.040 --> 00:05:01.839
I mean, everybody looks at these and looks at the numbers, yeah, and whoa, you know, yeah.
00:05:02.160 --> 00:05:06.720
I mean, to a first approximation, everyone in Brazil has PICs, basically.
00:05:06.800 --> 00:05:14.160
So um so people are looking at that and saying, well, first of all, why can't we have a national solution like that?
00:05:14.319 --> 00:05:21.600
Yeah, which is uh universal, mandated, you know, everyone's got the same interfaces, etc., etc.
00:05:22.399 --> 00:05:28.560
But there's also the issue of what about alternatives that are being developed by the private sector in different places?
00:05:28.800 --> 00:05:31.279
This is why I mentioned stable coins, for example.
00:05:31.600 --> 00:05:45.199
So could we achieve let's imagine if you're in Southeast Asia, for example, could we achieve regional sovereignty you know, by interconnecting, for example, the real-time payment networks?
00:05:45.279 --> 00:05:49.839
Yeah, could we interconnect domestic card networks or domestic payment schemes?
00:05:50.160 --> 00:05:55.360
Or could we look at developing parallel rails you know, using stable coins and so on?
00:05:55.680 --> 00:06:03.120
And actually, the idea of using parallel rails to achieve higher levels of resilience than building continuous landscape, that's pretty interesting.
00:06:03.839 --> 00:06:06.399
Now, and also you said about regional things.
00:06:06.480 --> 00:06:15.759
I mean, if you look at the example of Scandinavia, you got MPE and VIP, so all of these things are coming together and interconnecting, I think you see different paths to resilience now.
00:06:15.920 --> 00:06:23.839
So uh and working out what strategy is going to work best in which nation/slash region, uh it's interesting work, you know.
00:06:24.000 --> 00:06:58.879
Yeah, and and um shifting now to AI it is um yeah, you know, I I'm I remember as you know for when different technology platforms come out, you know, there there's a bit of the the platform itself, then the tools you can get from the platform, and then how that changes even how like for for example, like we use computer files, but they're named after physical paper files, and and but they're they have nothing to do with one one or the other.
00:06:58.959 --> 00:07:05.439
But it's it's just sort of a way that we as humans are able to manage you know that this new archetype.
00:07:05.680 --> 00:07:06.480
No, I know exactly.
00:07:07.600 --> 00:07:09.920
This is the problem challenge.
00:07:10.000 --> 00:07:24.240
I mean, this is the this is what finance faces because once you go agentic, once it's agents that are doing these transactions, there's no particular reason why you would think that the existing infrastructure that we have is the best for them.
00:07:24.800 --> 00:07:27.120
In fact, you it's got it's almost certainly not, yeah.
00:07:27.920 --> 00:07:30.720
So, and you know, stable coins are an interesting example.
00:07:30.800 --> 00:07:30.959
Yeah.
00:07:31.120 --> 00:07:34.480
So one of the discussions on the stage here this morning.
00:07:34.720 --> 00:07:40.639
So we can make some great stablecoin solutions, and you have people like Tempo and so on that are building this stuff out right now.
00:07:41.199 --> 00:07:46.720
Once you can run stable coins over those rails, actually, you can run any digital asset over those rails.
00:07:46.959 --> 00:07:54.639
There are plenty of people who know more about finance than I do that say, actually, a US dollar stablecoin, maybe that's not the best.
00:07:55.040 --> 00:08:02.160
Maybe actually tokenized money market funds would be a better solution, or even tokenized treasury bills, or I mean there's other kind of things.
00:08:02.399 --> 00:08:05.360
So now you've got tokens that uh yield-bearing tokens.
00:08:05.439 --> 00:08:05.759
Yeah.
00:08:06.000 --> 00:08:15.199
So if you can exchange those seamlessly over those networks, yeah, why would you ever come out of those into fiat stable coins?
00:08:15.279 --> 00:08:16.480
Why would you ever bother with that?
00:08:16.639 --> 00:08:16.800
Yeah.
00:08:17.040 --> 00:08:24.000
So so the AI isn't just uh, you know, clever ways of doing what we do now, but with different processes.
00:08:24.399 --> 00:08:29.439
It's bringing into existence very different ways of doing business.
00:08:29.920 --> 00:08:34.559
And I, you know, I take a mild amount of credit.
00:08:34.799 --> 00:08:40.000
I mean, I wrote a thing a few years ago for Wired magazine actually about AI and banking.
00:08:40.159 --> 00:08:40.559
Yeah.
00:08:40.799 --> 00:08:44.159
I can't remember how this came about, but it wasn't about banking.
00:08:45.200 --> 00:09:00.799
And, you know, I looked at what the banks were saying about all of this, and I I just it didn't seem right to me, and I couldn't figure out like, why is this not like they're gonna use the most powerful technology ever invented, and if the and they're gonna use it to send slightly better targeted credit card mailings?
00:09:00.879 --> 00:09:02.240
I mean, that's doesn't you know what I mean?
00:09:02.320 --> 00:09:15.200
It doesn't so and I realize and I'm not the only person that thought this, but you know, but then I realized you know, there is a revolution coming in financial services, but it's because of customers using AI, it's not because of banks using AI.
00:09:15.679 --> 00:09:22.480
And so this idea that you have the customer's AI becomes your customer, not the customer anymore, if you see what I mean.
00:09:22.559 --> 00:09:23.200
Yeah, yeah.
00:09:23.600 --> 00:09:29.039
And if you take on board, like people are saying, the primary users of stable coins are gonna be AIs, yeah.
00:09:29.200 --> 00:09:40.799
Like when you put all this stuff together, you end up with a different kind of financial infrastructure, which again I think is absolutely fascinating and intellectually very interesting to try and work out where that's going.
00:09:40.960 --> 00:09:46.000
So yeah, the AI stuff I'm really interested in is customers using AI.
00:09:46.159 --> 00:09:54.639
And I mean, don't you think that's kind of I think maybe merchants, banks, I think maybe they don't know what's coming because they don't have a strategy.
00:09:54.879 --> 00:10:02.000
I mean, who has a strategy for customers becoming a thousand times smarter literally overnight?
00:10:02.080 --> 00:10:09.759
Yeah, you know, we have business processes that are I was thinking about car in I had to buy car insurance a couple of weeks.
00:10:10.080 --> 00:10:16.240
You're a like no sane person wants to spend their time comparing car insurance policies, right?
00:10:16.480 --> 00:10:39.200
And and even how do you you know, even when I I get the two best quotes from the guy, I download the 400 pages of like how am I even supposed to compare the like that stuff cannot survive the the advent of agentics, so yeah, so it means a different kind of well, not only does it mean a different kind of financial market infrastructure, it means frankly a different kind of web is coming as well.
00:10:39.279 --> 00:10:40.799
So no, I I'm fascinated.
00:10:40.960 --> 00:10:46.000
Like, am I interested in how the bank uses AI to replace some people in the call center?
00:10:46.159 --> 00:10:47.360
No, I couldn't care less.
00:10:47.440 --> 00:10:49.120
I mean, that's just not interesting.
00:10:49.360 --> 00:10:50.000
Yeah, yeah, no.
00:10:50.080 --> 00:11:17.840
And to me, I'm sure it's good business for some people, but and and and uh sort of maybe jumping to the next that that that third piece, which is also related to AI and and I I I think in terms of identity, and and you know, if you have these agents rather than customers, um yeah, and they're customer driven, but like then what what what is identity at that point?
00:11:18.159 --> 00:11:20.159
I mean you're absolutely right to to point to this.
00:11:20.240 --> 00:11:38.720
This is a this is a fundamental problem with what we have now, which I mean people recognize, but we haven't quite so unless you can work out the identity of an AI, who authorized it, yeah, its reputation, like it's unfortunate like unless you know this stuff, like it doesn't work.
00:11:38.879 --> 00:11:45.360
You draw on the whiteboard, oh well, my bot will open a savings account with your bank's bot, whatever.
00:11:45.440 --> 00:11:47.759
Like that's easy to draw on a whiteboard, yeah.
00:11:47.919 --> 00:11:49.759
But like you drop it down one level.
00:11:49.919 --> 00:11:54.480
Yeah, how does my bot know that you actually are the authorized bot for the bank?
00:11:54.720 --> 00:11:59.200
How does your bot know that I'm an authorized bot who's allowed to access bank accounts?
00:11:59.360 --> 00:12:02.320
How do you know I'm allowed to access on behalf of Dave Birch?
00:12:02.480 --> 00:12:08.159
How do you know what I'm allowed to do on behalf of like when you go down one level, it's really complicated and not fixed.
00:12:08.639 --> 00:12:13.200
So again, intellectually very fascinating as to what to do there.
00:12:13.440 --> 00:12:22.720
And like we kind of the way we semi-fix this at the moment is like you give the bot permission to sort of be you, yeah, but that's not really what we want.
00:12:22.879 --> 00:12:28.480
What we want is an infrastructure where it's a bot, but it's acting on my behalf.
00:12:28.639 --> 00:12:30.159
Like, do I know this is?
00:12:30.879 --> 00:12:38.000
I mean, this is a silly example, but like, does Robin Hood know that this is the real Jim Kramer bot?
00:12:38.320 --> 00:12:47.200
Is the real Jim Kramer bot authorized to access who says, you know, is it authorized on behalf of Dave Birch?
00:12:47.519 --> 00:12:49.440
What can it do on behalf of Dave Birch?
00:12:49.600 --> 00:12:50.159
You know, etc.
00:12:50.399 --> 00:12:50.639
etc.
00:12:51.039 --> 00:12:55.679
So because I like I'm a normal, I'm a normal consumer of financial services.
00:12:55.840 --> 00:12:56.159
Yeah.
00:12:56.320 --> 00:13:01.120
Like the cons most of them are boring, like payments and so on, which I don't want to do them.
00:13:01.200 --> 00:13:02.320
I want bots to do those.
00:13:02.399 --> 00:13:02.559
Yeah.
00:13:02.720 --> 00:13:07.519
And the ones that aren't boring are so baffling and complex that I want an AI to do it.
00:13:07.600 --> 00:13:08.639
I don't want to do it myself.
00:13:08.720 --> 00:13:09.279
Yeah.
00:13:09.519 --> 00:13:18.639
Would you, would you back, I mean, even if you just talk about things like like in the UK right now, it's March, it's coming up to the end of our tax year.
00:13:18.720 --> 00:13:18.799
Yeah.
00:13:19.120 --> 00:13:22.720
Normal middle class people like me have to look at our savings.
00:13:22.960 --> 00:13:26.799
We have like these savings accounts, the tax efficient savings accounts.
00:13:27.120 --> 00:13:30.000
So you you have to open, like, which one do I want?
00:13:30.159 --> 00:13:32.799
Do I want the existing one or do I want one of the new ones?
00:13:33.200 --> 00:13:35.600
I don't want I don't want to spend any time doing that.
00:13:35.679 --> 00:13:37.360
I want a bot to do that for me.
00:13:37.440 --> 00:13:44.559
And I wouldn't back myself to make a better choice of those than even the even the the kickkeeper.
00:13:45.039 --> 00:13:45.200
Okay.
00:13:47.120 --> 00:13:52.639
I wouldn't back myself to make a better choice of savings account than even the most rudimentary bot.
00:13:52.879 --> 00:13:53.600
You know what I mean?
00:13:53.759 --> 00:13:56.559
So so actually, I think that's stuff's coming quite quickly.
00:13:56.639 --> 00:13:56.720
Yeah.
00:13:57.039 --> 00:14:00.080
If the identity problem that you point out is solved, yeah.
00:14:00.240 --> 00:14:08.879
Now, I think one of the bonuses to solving that problem, and this may sound a little odd, but I think it's true, is fraud is completely out of control.
00:14:09.039 --> 00:14:09.360
Yeah.
00:14:09.600 --> 00:14:11.759
A lot of that fraud is caused by AIs.
00:14:11.919 --> 00:14:12.399
Yeah.
00:14:12.720 --> 00:14:16.639
It's not obvious to me that an AI is harder to fool.
00:14:16.720 --> 00:14:17.039
Yeah.
00:14:17.200 --> 00:14:18.879
Like because AIs don't live in our world.
00:14:19.039 --> 00:14:23.200
AIs don't live in the world of, oh, this does this look a bit like Jim Kramer?
00:14:23.279 --> 00:14:23.360
Yeah.
00:14:23.519 --> 00:14:27.519
Did Jim Kramer really say, you know, go and buy this stock or whatever?
00:14:27.679 --> 00:14:33.200
Like AIs live in a world of keys and certificates and encryption, digital signatures.
00:14:33.360 --> 00:14:42.320
And like I can make a fake video of Jim Kramer saying, oh, you should buy Dave Birch's stock, but I can't make a fake digital signature of Jim Kramer.
00:14:42.399 --> 00:14:43.759
Like math doesn't work like that.
00:14:44.240 --> 00:14:51.440
So actually, I think once we get this identity thing sorted out, I actually think that'll be a much safer infrastructure too.
00:14:52.080 --> 00:14:52.960
What's interesting?
00:14:53.200 --> 00:15:08.399
And and uh I I we're we're getting towards the end of our time, and and uh I I wanted to add, I would be remiss if I did not ask you about you know what what's been your experience like here at FinTech Meetup so far.
00:15:08.720 --> 00:15:10.399
Uh for me, it was great.
00:15:10.559 --> 00:15:14.080
I got they they put me on the main stage this morning.
00:15:14.320 --> 00:15:16.159
My panelists were fantastic.
00:15:16.240 --> 00:15:22.320
I mean, you saw that it was just it was just great, and it had the right mix of people, yeah.
00:15:22.480 --> 00:15:26.639
So that the conversation was I said to them, like I only say yes to panel.
00:15:26.720 --> 00:15:30.480
I mean, like I don't want to sound like too much of a prima donna, but I do do this quite a lot, you know.
00:15:30.960 --> 00:15:35.440
I like panels where I'm gonna learn stuff as well, you know, and so wow.
00:15:35.600 --> 00:15:41.759
I mean, in terms of like like knowledge per minute on that stage this morning, it was terrific.
00:15:41.919 --> 00:15:52.960
No one was parroting like PR stuff, and they were all really informed, really so great, yeah, but also it's a bit in less tangible, this kind of thing.
00:15:53.360 --> 00:15:55.600
But don't you think it just like has a nice atmosphere?
00:15:55.759 --> 00:15:59.360
Yeah, like I've been just walking around, bumping into people, chatting to people.
00:15:59.440 --> 00:16:00.639
Yeah, it's been lovely.
00:16:00.720 --> 00:16:02.799
Uh and also by the way, some of the stands here, yeah.
00:16:02.960 --> 00:16:09.279
I have to say the new kind of like base level of swag has gone up significantly.
00:16:09.440 --> 00:16:15.919
Like there's really nice stuff out there on the floor, and so maybe there isn't going to be a global recession after all.
00:16:16.799 --> 00:16:17.759
I love that.
00:16:17.840 --> 00:16:23.759
And and uh I you know I think we're once again we're reached reached the end of our time here.
00:16:23.919 --> 00:16:26.879
But I I would be I just want to thank you personally.
00:16:26.960 --> 00:16:32.480
Like I've known you for so many years, and you have had such an impact on me personally.
00:16:32.720 --> 00:16:38.000
Oh, you're very I I um I I love your intellectual curiosity.
00:16:38.159 --> 00:16:40.159
Um it's second to none.
00:16:40.559 --> 00:16:46.399
And the way that you talk about things and you personalize things, it it has been a North Star for me.
00:16:46.480 --> 00:16:50.159
And that's I once again for my heart wanna want to thank you for that.
00:16:50.480 --> 00:16:51.600
Thank you so much, Sanjeev.
00:16:51.840 --> 00:16:53.279
That means a lot to me coming from you.
00:16:53.840 --> 00:16:54.159
Thank you.
