Latest / KAJ Business & Money LIVE / Why Your AI Strategy Is Burning Cash (And How to Fix It) | Andrew Quillen
Transcript
- 0:00Speaker 1: With you today.
- 0:02Speaker 2: You are welcome. Welcome to the show. Andrew, welcome to
- 0:05Speaker 2: India in this online form. And I'm sure not just
- 0:08Speaker 2: in India, but a lot of people, a lot of
- 0:10Speaker 2: businesses across the globe will benefit from what we are
- 0:13Speaker 2: going to talk right now.
- 0:14Speaker 3: We'll talk about AI.
- 0:15Speaker 2: We'll talk about how they can stop wasting money on
- 0:20Speaker 2: AI and only on the right kind of AI. So
- 0:22Speaker 2: we'll talk about all this. But before that, let's talk
- 0:25Speaker 2: a bit. Let's learn a bit about your journey, your background.
- 0:29Speaker 2: so that the audience knows who they are listening to?
- 0:33Speaker 1: Absolutely.
- 0:34Speaker 4: So as of right now, well, really over the last
- 0:37Speaker 4: two years, I've started my own firm. It's And Maverick.
- 0:43Speaker 4: This comes after several years working in AI. My first
- 0:47Speaker 4: actual career, my first job in my career was at Stripe.
- 0:50Speaker 4: So to see where they've gone versus where they were
- 0:52Speaker 4: before has been amazing.
- 0:54Speaker 1: That took me to open AI. Later, that took me
- 0:57Speaker 1: to full AI.
- 0:59Speaker 4: Head of AI roles at Spotify while doing slight consulting
- 1:04Speaker 4: work at Coinbase during not only the FTX fallout, but
- 1:09Speaker 4: also what came after. And it's so interesting now to
- 1:12Speaker 4: see our marketplace continue to evolve, but see how crypto
- 1:16Speaker 4: is starting to become the next pillar that needs to
- 1:18Speaker 4: walk down the AI aisle.
- 1:19Speaker 1: And it's a fascinating time.
- 1:24Speaker 2: Absolutely, absolutely fascinating times indeed. So I want to understand
- 1:29Speaker 2: what is it that you are trying to tell with
- 1:32Speaker 2: what you are saying, with what you are doing, also
- 1:35Speaker 2: what you are doing with and Maverick? Absolutely.
- 1:39Speaker 4: I would say the biggest part of that, and I'm
- 1:41Speaker 4: so glad that you hit on India a few moments ago.
- 1:44Speaker 4: I'm actually currently based in Indonesia, and I have the
- 1:48Speaker 4: fortunate privilege of working with clients around the world. Having
- 1:52Speaker 4: grown up in the West, I saw a lot of division,
- 1:55Speaker 4: not only between borders, but between ideologies. And what Ann
- 1:59Speaker 4: Maverick gets a chance to do now is set the
- 2:01Speaker 4: record straight for where we can take this technology because
- 2:04Speaker 4: it shouldn't be in the simplistic hands of VCs or
- 2:08Speaker 4: it shouldn't be in Silicon Valley.
- 2:10Speaker 1: It should be around the world. How do we take this?
- 2:12Speaker 1: How do we better ourselves?
- 2:13Speaker 4: How do we better the world that's coming after us
- 2:16Speaker 4: and leave it in a better place? That's what Ann
- 2:19Speaker 4: Maverick does. across my clients, across our work with governments,
- 2:22Speaker 4: across the opportunities that I have to give back to
- 2:25Speaker 4: a younger audience who are asking me questions, asking my
- 2:28Speaker 4: company not only thoughts and privileges, but where we can
- 2:32Speaker 4: take this and what we can do together.
- 2:33Speaker 1: And that's what I hope to achieve.
- 2:37Speaker 3: Absolutely. Absolutely. So let's talk about, you know, in terms
- 2:41Speaker 3: of AI, you.
- 2:42Speaker 2: Want it to, you know, to be available to a
- 2:45Speaker 2: lot of people. But you see companies, they are also...
- 2:49Speaker 2: looking at AI from their perspective to make many, many
- 2:54Speaker 2: trillion dollars companies, you know, for whatever it takes.
- 2:59Speaker 3: In that backdrop, do you think you.
- 3:01Speaker 2: Are talking about something utopia, that AI is actually a
- 3:05Speaker 2: power in some hands and only, you know, some bit
- 3:09Speaker 2: of crumbs in the hands of general people in terms
- 3:13Speaker 2: of generative AI? And actually now AI is also becoming
- 3:18Speaker 2: political power with the way it is. And we are
- 3:21Speaker 2: getting to see it much more in all terms. And
- 3:25Speaker 2: actually the cost is being paid by general population, either
- 3:29Speaker 2: in terms of water, the price of electricity, and much more.
- 3:34Speaker 1: You are completely right.
- 3:35Speaker 4: And what I would say is that most companies do
- 3:38Speaker 4: not have an AI problem. They have a haunted house problem.
- 3:41Speaker 4: They have old processes. They have dead decisions. They've forgotten owners.
- 3:45Speaker 4: and nobody willing to admit that the floorboards are actually
- 3:47Speaker 4: rotten underneath them, until you start to pick apart each
- 3:53Speaker 4: part of the map. Let's say that we talk about
- 3:54Speaker 4: Australia for a second, and we talk about what they're
- 3:56Speaker 4: wanting to do with data centers, so that if you're
- 3:59Speaker 4: putting them on the continent, how do you make those
- 4:02Speaker 4: centers actually work for the people? How does it then
- 4:05Speaker 4: reproduce natural resources? How does it then give you clean
- 4:08Speaker 4: water back to it? from all of the individuals that
- 4:11Speaker 4: I've spent time with, whether it be in India, whether
- 4:13Speaker 4: it be in China, whether it be in the United
- 4:14Speaker 4: States or anywhere else in between, you find the same commonality.
- 4:18Speaker 1: There are brilliant people everywhere. And what does that look like?
- 4:22Speaker 4: We need to be able to take a pilot that
- 4:24Speaker 4: proves the technology can work, but also understand how the
- 4:28Speaker 4: transformation proves the organization can survive changing how it works indeed.
- 4:33Speaker 1: And that's a wildly different achievement.
- 4:35Speaker 4: And so to sit down and have a base understanding
- 4:37Speaker 4: of what AI is, is not necessarily a prerequisite to
- 4:41Speaker 4: understanding how it works for you. We need to understand
- 4:45Speaker 4: how we go back to our comprehension. How do we
- 4:47Speaker 4: understand what works for us on a day-to-day basis and
- 4:50Speaker 4: then use the AI technology as an assistant, as the
- 4:54Speaker 4: assistant to our Batman, per se, and utilize that to
- 4:58Speaker 4: then catapult our businesses, our processes, our systematic approach to
- 5:04Speaker 4: the way we do business in a meaningful way.
- 5:06Speaker 1: And that's what AI gives us on a day-to-day basis.
- 5:11Speaker 3: Absolutely, Andrew. We'll come to the business part of it.
- 5:15Speaker 2: First, let's understand from the useful part of it.
- 5:18Speaker 3: You talk about useful AI.
- 5:20Speaker 2: How should a general person understand what useful AI is?
- 5:25Speaker 2: You have worked with OpenAI, you have worked with other companies,
- 5:29Speaker 2: all those companies who are known for whatever the way
- 5:32Speaker 2: people know them. Either looking at non-profit or working as
- 5:36Speaker 2: a profit company.
- 5:38Speaker 3: All those people know about it.
- 5:39Speaker 2: I want to understand how should they understand useful AI
- 5:43Speaker 2: in today's time. Many people are scared of AI in
- 5:46Speaker 2: that sense.
- 5:48Speaker 4: No, you're absolutely right. There is so much fear that
- 5:51Speaker 4: we see on a day-to-day basis. And what I learned
- 5:54Speaker 4: at OpenAI is, you know, it allowed for us to
- 5:58Speaker 4: change the question from how intelligent can this become to
- 6:02Speaker 4: what happens when intelligence becomes abundant. when everyone has access
- 6:05Speaker 4: to intelligence and we live in a smarter world. Well,
- 6:08Speaker 4: that's where I love and where I want to take
- 6:10Speaker 4: AI and where so many of us alike, like myself,
- 6:13Speaker 4: want to see this field continue to evolve. And so
- 6:16Speaker 4: from the user perspective, seeing it less as a deterrent
- 6:20Speaker 4: and more as a third extension, a third arm per se,
- 6:26Speaker 4: is a meaningful shift of approach because that allows you
- 6:29Speaker 4: to understand what it does for you on a day-to-day
- 6:32Speaker 4: basis instead of saying, well, this may offset me, this
- 6:35Speaker 4: may cause my job to be lost. We can take
- 6:39Speaker 4: this same technology and build in this space, in this
- 6:42Speaker 4: creator field, something that outlasts even you and I. And
- 6:46Speaker 4: that is, I don't know that we've had a more
- 6:48Speaker 4: rewarding opportunity in our lifetimes or in lifetimes past to
- 6:52Speaker 4: do that.
- 6:53Speaker 1: And I would say that that is a huge.
- 6:54Speaker 4: Compliment to the technology where it sits today and where
- 6:57Speaker 4: we can continue to take it as long as intelligence
- 7:00Speaker 4: remains the key driver in this conversation.
- 7:05Speaker 3: Absolutely, Andrew.
- 7:06Speaker 2: Let's understand it from the leader's point of view, from
- 7:09Speaker 2: a business perspective.
- 7:12Speaker 3: You talk about durable.
- 7:14Speaker 2: How should I understand a durable AI strategy.
- 7:21Speaker 3: When nothing is concrete? Every day, something new is coming.
- 7:25Speaker 2: Every day, some new model is coming from any other
- 7:28Speaker 2: part of the world.
- 7:30Speaker 3: Then there is in terms of pricing, which is now
- 7:33Speaker 3: becoming a major factor. In that aspect, how does.
- 7:36Speaker 2: A company leader decide what AI's durable strategy would be?
- 7:42Speaker 2: And what are the heights that you talk about in
- 7:46Speaker 2: today's time, especially from a business.
- 7:48Speaker 3: Point of view?
- 7:51Speaker 4: Well, the first way that I look at this is
- 7:52Speaker 4: that companies keep buying intelligence while protecting, I want to
- 7:56Speaker 4: be soft when I say this, but while protecting the
- 7:58Speaker 4: stupidity that's already in the system itself.
- 8:01Speaker 1: And that's not to say that the people within the
- 8:03Speaker 1: system are stupid.
- 8:04Speaker 4: It's that the processes have almost been relegated as antiquated.
- 8:08Speaker 4: And what that does is with this technology, we can
- 8:10Speaker 4: install new systems that preserve, but still preserve the meetings,
- 8:14Speaker 4: still take on the approvals, the incentives and hierarchies that
- 8:17Speaker 4: made the old systems relatively useless.
- 8:20Speaker 1: But as far as a pricing model goes, what I
- 8:23Speaker 1: recommend to.
- 8:23Speaker 4: My clients and what I would recommend to yours as
- 8:25Speaker 4: well is that you need to find part of that
- 8:28Speaker 4: same model.
- 8:31Speaker 1: Let's talk about a couple of models here.
- 8:32Speaker 4: Let's talk about an open AI model versus a moonshot model.
- 8:37Speaker 4: Those models exist in the same landscape, but understanding the
- 8:42Speaker 4: price and what you actually need so that you don't
- 8:44Speaker 4: get into a negotiation where you're being priced at, let's
- 8:48Speaker 4: say USD for $ 200, where you want to actually be
- 8:50Speaker 4: able to pay half of that.
- 8:52Speaker 1: Well, how do you do that?
- 8:53Speaker 4: Well, you find a smaller model that still has the
- 8:56Speaker 4: same intelligence. You're not, and I would push on this,
- 8:59Speaker 4: most people are not utilizing the full capability of a
- 9:02Speaker 4: model given to them on a day-to-day basis or a
- 9:05Speaker 4: month-to-month basis. And they're paying out a premium for that.
- 9:08Speaker 4: And so to step back, understand the landscape, understand the
- 9:12Speaker 4: models that you're using, and don't be so tied down
- 9:14Speaker 4: to flags. That's what I recommend. and continue to lead
- 9:18Speaker 4: my own business the same way.
- 9:22Speaker 3: Absolutely.
- 9:23Speaker 2: In terms of hype, Andrew, how should a business leader
- 9:27Speaker 2: evaluate that? How does he know whether this hype is
- 9:31Speaker 2: actually truth or is it actually, you know, inflated sort
- 9:36Speaker 2: of explanation? Because he or she doesn't know about these things.
- 9:41Speaker 2: They may be listening to the same things, you know,
- 9:44Speaker 2: the first time, people might think, you know, this is
- 9:46Speaker 2: a great idea. It might be the biggest idea of
- 9:49Speaker 2: the future. I shouldn't be left behind. And then you
- 9:51Speaker 2: find it was all uh hype and nothing nothing in reality.
- 9:57Speaker 4: I think it's i think it's a great question because
- 9:59Speaker 4: it it pinpoints the exact issue that that we're all
- 10:02Speaker 4: seeing right now and that's you know we have this
- 10:04Speaker 4: abundance of information but how do you really weed out
- 10:07Speaker 4: what is right and what is wrong it's a there
- 10:09Speaker 4: are there are apps like if you want to write
- 10:11Speaker 4: and you want to continue to do well with your
- 10:13Speaker 4: writing i believe it's called grammarly And Grammarly is something
- 10:16Speaker 4: that I've used for a number of years because it
- 10:17Speaker 4: helps me understand how to be correct in the way
- 10:20Speaker 4: that I am grammatically approaching a conversation with someone. Well,
- 10:23Speaker 4: that same understanding is now free to use in platforms
- 10:27Speaker 4: like Google and platforms like a parent company like an alphabet.
- 10:31Speaker 4: To understand that and to balance that out towards the
- 10:34Speaker 4: greater question here is to look across the board and
- 10:37Speaker 4: understand exactly what tools are new and which ones have
- 10:40Speaker 4: been in place for a long period of time. I
- 10:42Speaker 4: am very, very, very open to the idea that old
- 10:47Speaker 4: ways of doing things are not necessarily so inherently wrong,
- 10:50Speaker 4: but it's finding little pieces, little parts of your ecosystem
- 10:55Speaker 4: that can be improved upon. And spending, even if it's
- 10:59Speaker 4: drops in the ocean and maybe it's five or even
- 11:01Speaker 4: $ 10 per month on small things that can make your
- 11:04Speaker 4: overall system feel like a new way of doing business,
- 11:08Speaker 4: that's without having to get into your whole monthly balance
- 11:10Speaker 4: or your whole monthly budget. And really having your finance
- 11:13Speaker 4: team look at you cross-eyed and thinking that you're crazy.
- 11:16Speaker 1: That's the biggest part of this.
- 11:18Speaker 4: And number two is understanding who to listen to and
- 11:21Speaker 4: who not to listen to. I wouldn't profess myself a sage,
- 11:25Speaker 4: but I would recommend to have more of an approach
- 11:29Speaker 4: towards a worldly view on what's going on so that
- 11:32Speaker 4: that way you're never taking one from one source. And
- 11:35Speaker 4: that tends to have served me well in my career,
- 11:37Speaker 4: and it's what I would recommend.
- 11:38Speaker 1: Absolutely.
- 11:40Speaker 3: Absolutely.
- 11:41Speaker 2: Andrew, so let's talk from a leader's point of view
- 11:44Speaker 2: when selecting AI. When you are looking at it from
- 11:48Speaker 2: an organization's point of view, you may have advisors, you
- 11:52Speaker 2: may have your CTO, you may have your tech team,
- 11:55Speaker 2: whatever it is with their whatever.
- 11:57Speaker 3: Level of knowledge.
- 11:59Speaker 2: How do you evaluate from your organization's needs in terms
- 12:04Speaker 2: of just the needs Or you look at those models
- 12:08Speaker 2: in terms of pricing, in terms of trust, in terms
- 12:12Speaker 2: of reliability.
- 12:13Speaker 3: Because these words are also very wide.
- 12:18Speaker 2: Trust in terms of whether that country may not have
- 12:21Speaker 2: a problem with your country, depending on where you are staying.
- 12:26Speaker 3: Or, you know, somebody who is another part.
- 12:30Speaker 2: Of the world, that country does not have, you know,
- 12:34Speaker 2: political problems with that country. Or there is, you know,
- 12:38Speaker 2: your country is supporting another country which can have a
- 12:41Speaker 2: problem with that country where this particular model is situated.
- 12:46Speaker 2: That is why a lot of countries are talking about
- 12:48Speaker 2: native AIs and all. How do I as a business
- 12:52Speaker 2: leader who do not have much AI presence in my
- 12:56Speaker 2: own country but have a big business, maybe, you know,
- 12:59Speaker 2: in several countries, how do I assess reliability and trust
- 13:03Speaker 2: from that point of view, Andrew?
- 13:06Speaker 1: There's two ways to do this. Number one, I have.
- 13:12Speaker 4: Two people in my corner that I trust with most
- 13:14Speaker 4: decisions that I need to have. And that's an engineer,
- 13:17Speaker 4: a really sound engineer that spent his career doing that.
- 13:20Speaker 4: And then also a PM, a product designer. I want
- 13:24Speaker 4: someone in product management so that that way they can
- 13:27Speaker 4: both look at the system, tell me what they're seeing,
- 13:30Speaker 4: and then really get me to the middle of the pie.
- 13:33Speaker 4: Because both are going to be extracting different parts of that.
- 13:36Speaker 4: And then what I find back in the middle is
- 13:38Speaker 4: exactly where I need to be. I couldn't recommend them enough.
- 13:41Speaker 4: They are roles that I hope continue to evolve. I've
- 13:44Speaker 4: met some of the best engineers of my career in India,
- 13:47Speaker 4: which is not ironic. You guys have been brilliant to
- 13:51Speaker 4: me in my career. But even more than that, in
- 13:54Speaker 4: second part of this, is understanding how to find the
- 13:58Speaker 4: absurdity in all of this. What makes sense and what
- 14:01Speaker 4: just feels like it's blowing smoke at you? And that's
- 14:05Speaker 4: That's being able to understand your own comprehension and understand
- 14:08Speaker 4: where you can redesign the whole damn thing. And that
- 14:11Speaker 4: to me has been a major shift. Because once you
- 14:15Speaker 4: can sit down and let's say that we were looking
- 14:17Speaker 4: at two or three different things and we needed to
- 14:19Speaker 4: have a conversation with two or three different vendors. How
- 14:21Speaker 4: do we get to the ballpark where we are actively
- 14:24Speaker 4: seeking out not only the price that works, but the
- 14:27Speaker 4: tooling that we need? And that's staying relevant with your business.
- 14:31Speaker 4: staying up to date on whatever it may be. Maybe
- 14:33Speaker 4: you're in marketing. Maybe you need to be making sure
- 14:36Speaker 4: that your ad copy or that your email campaigns are
- 14:38Speaker 4: actively and successfully deploying against your marketplace. Or maybe you
- 14:43Speaker 4: work in such a way that you're working with railroads.
- 14:47Speaker 4: Those are still parts of that system, but finding out
- 14:50Speaker 4: the tools that work for you on a day-to-day basis,
- 14:52Speaker 4: instead of just giving your money towards one model or
- 14:55Speaker 4: a different model, you actually are building out your ecosystem
- 14:58Speaker 4: to allow for the system to work for you instead
- 15:01Speaker 4: of against you.
- 15:04Speaker 3: Absolutely, Andrew.
- 15:05Speaker 2: Let's understand it from government point of view, the Wall
- 15:09Speaker 2: Street point of view, powerful institutions point of view.
- 15:12Speaker 3: How do they.
- 15:14Speaker 2: Evaluate AI for their use from whatever they have? Maybe
- 15:19Speaker 2: CIA has its own way. Perhaps they are usually creating
- 15:22Speaker 2: their own AI. But then Pentagon had a problem with anthropic.
- 15:28Speaker 2: And then several organizations have that. But maybe not all
- 15:31Speaker 2: organizations are so powerful. But in businesses, how do these
- 15:38Speaker 2: powerful businesses actually evaluate AI for their needs?
- 15:45Speaker 4: Governments evaluate it like it is a software procurement. So
- 15:49Speaker 4: to me, that's the wrong frame, because what they need
- 15:52Speaker 4: to be evaluating it as is infrastructure.
- 15:55Speaker 1: They need to be making.
- 15:56Speaker 4: Sure that if it touches healthcare, defense, education, finance, or
- 16:00Speaker 4: even public benefits, that you don't risk failure when it
- 16:04Speaker 4: becomes civic. And that's, so we're not only asking how
- 16:07Speaker 4: accurate is the model we need to understand is we
- 16:11Speaker 4: need to understand if and who gets harmed when it
- 16:15Speaker 4: is wrong, who can challenge that decision and whether the
- 16:18Speaker 4: system itself can actually explain after the damage is done.
- 16:22Speaker 4: We've seen increasingly, we've seen increasing reports on this and we're,
- 16:26Speaker 4: what that has shifted that whole perspective, that whole landscape
- 16:29Speaker 4: to look at is not that a government should never
- 16:33Speaker 4: deploy an AI system, but it should never deploy one
- 16:36Speaker 4: that it cannot audit, interrupt, or publicly defend. If nobody
- 16:40Speaker 4: can reconstruct the decision, then nobody truly owns it. And
- 16:44Speaker 4: that's a major issue because with the benchmarks that we
- 16:47Speaker 4: see across the board is we want our governments to
- 16:50Speaker 4: be faster utilizing the technology, but it should only be
- 16:54Speaker 4: that way When a citizen becomes safer, they're healthier, they're
- 16:59Speaker 4: better served, and less trapped inside bureaucratic machinery. We saw
- 17:02Speaker 4: that in crypto. We've seen that across different industries where
- 17:06Speaker 4: we're reducing the real-life time that someone is needing to
- 17:10Speaker 4: be on that platform. If we can continue to steer
- 17:13Speaker 4: that the right direction, then I think it behooves everyone
- 17:16Speaker 4: within not only within those domains, but people at large.
- 17:24Speaker 3: Absolutely, Andrew.
- 17:25Speaker 2: Andrew, you have worked with very, very top organizations in
- 17:29Speaker 2: this field. You are an insider from that perspective. And
- 17:35Speaker 2: now you are also advising a lot of companies. When
- 17:39Speaker 2: you look at all those companies, all those years that
- 17:42Speaker 2: you spent with them, and now you look at their
- 17:46Speaker 2: image management for the outside world, do you think they
- 17:51Speaker 2: are more moving towards... creating, you know, powerful structures, trillion
- 17:57Speaker 2: dollar companies more towards that, creating more leverage in terms of,
- 18:02Speaker 2: you know, controlling things from technology perspective, being part of
- 18:06Speaker 2: the political systems which will benefit them? Or is it
- 18:11Speaker 2: they're actually, you know, actually looking at taking humanity forward?
- 18:17Speaker 3: Because that's a lot of stake for humanity.
- 18:21Speaker 1: It certainly is.
- 18:22Speaker 4: And I love that question because it actually takes two
- 18:27Speaker 4: different lines of thought here. We can either move us
- 18:28Speaker 4: forward or we can do right by our everyday people.
- 18:32Speaker 1: And that's a great question.
- 18:33Speaker 4: I would say that some of the companies that I've
- 18:35Speaker 4: been a part of are now in what I would
- 18:38Speaker 4: consider the reconstruction of their own image. OpenAI jumps out
- 18:43Speaker 4: to me when I say that out loud because due
- 18:45Speaker 4: to their own relations with the US government and how
- 18:49Speaker 4: that's played out over the last eight months back home
- 18:51Speaker 4: in the States has been wildly interesting and wildly different
- 18:55Speaker 4: than another Western competitor in Anthropic. But what that has
- 18:59Speaker 4: done for Spotify and what that's done for Coinbase and
- 19:01Speaker 4: what that's even done for Stripe is Stripe here recently,
- 19:04Speaker 4: and you may know this as well as I do,
- 19:07Speaker 4: that they've recently purchased or are looking to purchase and
- 19:09Speaker 4: close the deal with OpenRouters. And OpenRouters does most of
- 19:13Speaker 4: the business, I think 60 or 65% of the business with.
- 19:15Speaker 1: Fortune 500 companies.
- 19:18Speaker 4: And so we're taking all of these models, all of
- 19:20Speaker 4: these great companies, we're buying them within the infrastructure of
- 19:23Speaker 4: what Stripe is.
- 19:24Speaker 1: And Stripe has continued to take us forward.
- 19:27Speaker 4: Spotify and music has led the way in how we
- 19:29Speaker 4: not only discover music, but also audiobooks and podcasts. And
- 19:35Speaker 4: taking that in account to how we build out a
- 19:37Speaker 4: system that applies and benefits the everyday person.
- 19:41Speaker 1: I am so appreciative of my time at the company.
- 19:43Speaker 1: And then Coinbase.
- 19:44Speaker 4: Coinbase is at a place right now where we need
- 19:47Speaker 4: and we are continuing to see AI. AI agents, AI
- 19:52Speaker 4: and its orchestration layer level into crypto to actually create
- 19:57Speaker 4: a regulated space so that that way we're seeing more
- 19:59Speaker 4: and more currency that is actually being benefited by the technology.
- 20:04Speaker 4: The I think they're not the last two to go
- 20:08Speaker 4: that direction, but it's certainly an overwhelming majority in that
- 20:12Speaker 4: part of the world that sees it as a great thing.
- 20:15Speaker 4: And I think once we come to the reality that
- 20:17Speaker 4: that is a benefiting factor to so many other people,
- 20:20Speaker 4: it also opens the door for that as well.
- 20:25Speaker 3: Absolutely.
- 20:26Speaker 2: So do you think for businesses with the present level
- 20:30Speaker 2: of AI... And most of the AI, my personal opinion,
- 20:35Speaker 2: and I talk to several experts also, you can say
- 20:38Speaker 2: it's much more like, you know, they talk of agentic AI,
- 20:41Speaker 2: I talk about psychophantic AI. You tell them and they
- 20:44Speaker 2: will do whatever you say to them, whether it is
- 20:46Speaker 2: right or wrong. They want to please you rather than
- 20:49Speaker 2: make you, help you reach your mission. In that aspect,
- 20:53Speaker 2: with the amount of data that we have from AI's performance,
- 20:58Speaker 2: Do you think that is enough for business leaders to
- 21:02Speaker 2: take concrete measures for their business or chart a path
- 21:06Speaker 2: for their business for the coming future? And especially from
- 21:10Speaker 2: a small business point of view, mid-sized business point of view.
- 21:13Speaker 3: The bigger ones, they have deep pockets. They can certainly
- 21:17Speaker 3: take a hit if they go a wrong way.
- 21:23Speaker 4: I personally am not as impressed with AI agents, with
- 21:28Speaker 4: agentic AI. Those are two different thoughts, but they're really
- 21:31Speaker 4: overlapping in the way that I'm going to approach this question.
- 21:35Speaker 4: That has progressively decreased, and not because we're looking at
- 21:40Speaker 4: it the wrong way. It's that people have forgotten that
- 21:43Speaker 4: what agents are and what agentic technology was designed to
- 21:46Speaker 4: be was already part of our systems several years ago,
- 21:51Speaker 4: what we would describe as entire workflows, work graphs. If
- 21:55Speaker 4: you were thinking out loud in that conversation, then that
- 21:58Speaker 4: was a part of what you were doing on a
- 21:59Speaker 4: day-to-day basis. And now where we are is almost a
- 22:03Speaker 4: retroactive fit to, okay, well, I need to understand this
- 22:06Speaker 4: market or I need to understand that. And that's all
- 22:08Speaker 4: great and well, but how do we then progress that
- 22:11Speaker 4: to if we're in Q3 of 26, how do I
- 22:14Speaker 4: make sure that the system that I'm speaking with is
- 22:17Speaker 4: pulling the right data so that that way I'm not
- 22:19Speaker 4: two quarters behind, but I'm two or three quarters ahead
- 22:23Speaker 4: And that's restructuring, that's designing your system in a way
- 22:26Speaker 4: that the agents are actually pulling that out for you.
- 22:29Speaker 4: But as a default system to the everyday non-prompt engineer,
- 22:34Speaker 4: that may be a struggle. And that has been something
- 22:37Speaker 4: that has increasingly made me worried for the marketplace, just
- 22:43Speaker 4: because there's not a lot of coverage gap for that.
- 22:46Speaker 4: Some do it better than others, as you hit on,
- 22:48Speaker 4: but the overwhelming majority has been a disappointment in my mind.
- 22:54Speaker 3: Absolutely.
- 22:55Speaker 2: So in your case, how should we understand what Andrew
- 23:00Speaker 2: is trying to do? Is he, in terms of the
- 23:04Speaker 2: naming of your company and Maverick, in terms of the
- 23:07Speaker 2: way everything is not clear, what is Andrew trying to do?
- 23:12Speaker 2: Is he propagating the views of those companies or is
- 23:16Speaker 2: he telling people about to be careful, especially business people,
- 23:20Speaker 2: to be careful about the future, the path that they take?
- 23:25Speaker 4: Well, you know, with so much of my work, a
- 23:28Speaker 4: lot of what I get to do now is I
- 23:29Speaker 4: get to lend my time and I get to lend
- 23:32Speaker 4: my experiences to the opportunity of presenting and forwarding and
- 23:38Speaker 4: progressing an ecosystem, not only of our everyday lives, but
- 23:43Speaker 4: of the lives of those who are coming after us,
- 23:45Speaker 4: whether it be coral reefs, whether it be for future generations,
- 23:49Speaker 4: whether it be in the everyday parts of our conversations,
- 23:52Speaker 4: that the most important AI systems are We're never about
- 23:55Speaker 4: writing emails or generating an ad. They were taking time
- 23:58Speaker 4: to listen to forests and understand how we can help
- 24:01Speaker 4: our ecosystems and how we can take that and track
- 24:05Speaker 4: diseases before symptoms appear. What do model collapsing ecosystems actually
- 24:09Speaker 4: sound like? How do we prevent those in live time
- 24:12Speaker 4: so that that way we don't have another cloud flare
- 24:13Speaker 4: example like we had six months ago? Well, if we're
- 24:16Speaker 4: doing that, can we also at the same time give
- 24:19Speaker 4: humanity a few more years to make a better decision?
- 24:21Speaker 1: Can we increase their lifespans?
- 24:23Speaker 4: Are we accessing and allowing there to be a shift
- 24:28Speaker 4: in modern medicine that takes us into a brighter future?
- 24:32Speaker 4: That's what AI is. That's what we used to design
- 24:35Speaker 4: it for when I was very young at OpenAI. We
- 24:38Speaker 4: were essentially just nerds. We wanted to ask questions. We
- 24:41Speaker 4: cared about what was in front of us, and we
- 24:43Speaker 4: cared about what was behind us. And so much of
- 24:45Speaker 4: that now is in understanding that AI is not productivity.
- 24:51Speaker 1: It's perception.
- 24:53Speaker 4: It may let humanity see patterns in climate, in aging,
- 24:58Speaker 4: in reefs, in things that were always there, but we're
- 25:02Speaker 4: waiting for a mind large enough to notice.
- 25:05Speaker 1: And that's what I like to do. That's what I'm
- 25:06Speaker 1: doing now.
- 25:08Speaker 3: Wonderful.
- 25:09Speaker 2: There is so much to learn about all this from you, Andrew,
- 25:12Speaker 2: and I'm sure a lot of leaders, a lot of
- 25:14Speaker 2: companies would want to connect with you.
- 25:16Speaker 3: What's the best way for them to do so?
- 25:19Speaker 1: I appreciate that.
- 25:20Speaker 4: The best way to get in touch with me is
- 25:22Speaker 4: actually through my company website.
- 25:25Speaker 1: It's just www.anmaverick.com.
- 25:30Speaker 4: And from there, you can find just about any way
- 25:32Speaker 4: that you could ever think to reach me to reach me.
- 25:35Speaker 3: Wonderful.
- 25:36Speaker 2: With this, it's set up on this very special edition
- 25:39Speaker 2: of the KJ Business and Money Live.
- 25:41Speaker 3: Thank you so much indeed for joining us.
- 25:45Speaker 1: Thank you.