Latest / The Starting Block / 1-on-1 with Aptos co-founder Avery Ching

Transcript

1-on-1 with Aptos co-founder Avery Ching · The Starting Block

• 0:08
  Good morning and good afternoon to
• 0:10
  you, wherever you're joining us from
• 0:11
  today, welcome to The Starting
• 0:12
  Block. If it's your first time here,
• 0:14
  welcome.
• 0:15
  My name is Gareth Jenkinson.
• 0:16
  I'm the host and head of multimedia
• 0:17
  here at The Block.
• 0:19
  We're coming up on month
• 0:21
  number two, being wrapped up of The
• 0:22
  Starting block. It's been a great
• 0:24
  journey so far, many, many fantastic
• 0:26
  conversations, starting all the way
• 0:28
  back with that in-person interview
• 0:29
  with CZ two months ago
• 0:31
  in Abu Dhabi. If you haven't seen
• 0:33
  that, you can go to The Block's
• 0:34
  YouTube channel and you'll find that
• 0:36
  interview along with many other.
• 0:38
  Of our top interviews in the last
• 0:40
  two months of the show.
• 0:41
  We were also up
• 0:43
  at a SALT
• 0:45
  conference this week.
• 0:46
  If you wanna catch some of the
• 0:48
  interviews that my colleague Calvin
• 0:49
  Sparks conducted over the last
• 0:51
  couple of days, we've already
• 0:52
  published his interview with Anthony
• 0:54
  Scaramucci, some really interesting
• 0:57
  takes coming out of the
• 0:59
  SALT Conference up there.
• 1:01
  So please do go ahead and subscribe
• 1:02
  to our YouTube channel if you don't
• 1:04
  already. And if you're watching us
• 1:05
  here on X this afternoon.
• 1:07
  Give us a like and a retweet so we
• 1:08
  can have more people join us for
• 1:10
  our conversation today.
• 1:12
  In a few minutes time, we're gonna
• 1:13
  be joined by Aptos CEO
• 1:15
  and co-founder Avery Ching
• 1:17
  to have a wide ranging conversation.
• 1:19
  We last caught up with him in person
• 1:20
  at Consensus a couple of months ago,
• 1:22
  so plenty to chat about today.
• 1:25
  And of course, if you're joining us,
• 1:26
  first thing this
• 1:28
  afternoon, we're getting close to
• 1:29
  the afternoon in US time, even we're
• 1:32
  broadcasting a little bit later
• 1:33
  today. It's been an interesting day.
• 1:34
  Or 24 hours for the crypto markets
• 1:36
  in general, Bitcoin absolutely
• 1:38
  cruising over the $70,000
• 1:40
  mark and we'll be jumping into that.
• 1:42
  And a couple of other stories in the
• 1:44
  block briefing, which is our rundown
• 1:45
  of the biggest stories you need to
• 1:47
  know in the last 24 hours,
• 1:49
  curated from the block's exceptional
• 1:51
  news team. And we're gonna jump
• 1:53
  straight into that today with our
• 1:54
  first headline, which is of course
• 1:56
  markets related.
• 1:57
  Crypto markets ripped higher on
• 1:59
  Wednesday, Bitcoin touched.
• 2:01
  $69,000 yesterday
• 2:03
  for the first time in two months.
• 2:04
  And as I mentioned already up
• 2:06
  near the $72,000
• 2:08
  mark while Ether reclaimed 2000
• 2:11
  as it jumped 10%.
• 2:13
  And that rally followed two
• 2:14
  catalysts. The treasury department
• 2:16
  said it's at least doubling the
• 2:18
  size of its liquidity buyback
• 2:20
  operations for longer dated
• 2:21
  securities starting on September
• 2:23
  9th and the SEC floated
• 2:26
  a fresh crypto offering proposal.
• 2:29
  That surge wasn't painless though.
• 2:30
  With a move trigger.
• 2:31
  Nearly $2 billion in liquidations.
• 2:34
  And my DJ and producer Jordan tells
• 2:36
  me that he was reading a few stories
• 2:38
  that said that might've been the
• 2:39
  worst we've seen in history and
• 2:41
  CryptoLink stocks caught that
• 2:42
  updraft too with strategy and
• 2:44
  Bitmine each gaining around
• 2:46
  10%. So really, really interesting
• 2:48
  last 24 hours in the crypto markets.
• 2:50
  And we'll be watching that very
• 2:51
  closely. If you want any more market
• 2:53
  updates or insights, just go and
• 2:55
  check out the block's website.
• 2:56
  That's theblock.co.
• 2:59
  And staying on that vein,
• 3:00
  Hyperliquid's hype token surged
• 3:02
  after President Trump said the CFTC
• 3:05
  is working to bring the
• 3:07
  Perpetuals platform into the United
• 3:09
  States in what he called a fully
• 3:10
  compliant and legal fashion,
• 3:13
  speaking at a Wednesday press
• 3:14
  conference alongside technology
• 3:16
  leaders and federal agency chairs,
• 3:18
  as well as many big names in the
• 3:19
  cryptocurrency industry.
• 3:21
  Trump said CFTC chair
• 3:23
  Michael Selig is working to
• 3:25
  bring perpetual futures markets
• 3:27
  onshore.
• 3:28
  Hyperliquid has gained real traction
• 3:30
  among traders and a path to
• 3:31
  operating legally in the US would
• 3:33
  open the door to a far larger
• 3:35
  American user base.
• 3:37
  Watch this one, clarity on perps
• 3:39
  could reshape derivatives trading
• 3:41
  here. And if you're in it just for
• 3:42
  the memes, plenty of fantastic
• 3:44
  memes on X in the last 24
• 3:46
  hours around hype and
• 3:48
  the dimension of hyperliquids,
• 3:50
  specifically by the
• 3:52
  United States president.
• 3:53
  Then the ETF money is flowing back
• 3:55
  in, U.S.
• 3:56
  Spot Bitcoin exchanges.
• 3:58
  Exchange Traded Funds, rather,
• 3:59
  pulled in $517 million
• 4:01
  in net inflows on Wednesday,
• 4:04
  right in the broader market rebound.
• 4:05
  According to SoSoValue, 8
• 4:07
  of the 12 Bitcoin ETFs saw
• 4:09
  inflows led by BlackRock's
• 4:11
  iBit with $285 million
• 4:14
  worth.
• 4:15
  ARK in 21 shares, ARKB
• 4:16
  added $78 million
• 4:18
  and Fidelity's FBTC
• 4:20
  brought in $62 million.
• 4:23
  It was the biggest single day haul
• 4:24
  since early May, three and
• 4:26
  a half months ago.
• 4:27
  Analyst tied that rush straight back
• 4:29
  to that surprise rally sparked by
• 4:31
  the treasury's buyback expansion.
• 4:33
  Then regulators are stepping on the
• 4:35
  gas. The Office of the Comptroller
• 4:36
  of the Currency is racing to
• 4:38
  finalize rules by November to
• 4:40
  implement the Genius Act stablecoin
• 4:41
  law signed by President Trump.
• 4:44
  Speaking at the Seoul Conference in
• 4:45
  Wyoming on Wednesday, Comptrollers
• 4:47
  Jonathan Gould said the agency
• 4:49
  has taken in public comments and
• 4:51
  revised its initial 376
• 4:53
  page proposal.
• 4:55
  Which clarifies OCC jurisdiction
• 4:57
  over stable coins, including capital
• 4:59
  and liquidity requirements.
• 5:01
  Gold said the goal is a final rule
• 5:03
  by November, so the agency
• 5:05
  can start processing applications
• 5:07
  within the new year for
• 5:09
  issuers.
• 5:10
  A real timeline is finally coming
• 5:12
  into view.
• 5:14
  Then a notable move in the
• 5:15
  interoperability space, Nethamind,
• 5:17
  a core contributor to the Ethereum
• 5:19
  ecosystem is leaving its role.
• 5:21
  As a layer zero verifier
• 5:23
  and shifting its cross-chain
• 5:25
  operations to Chainlink after
• 5:27
  what it called an extensive review.
• 5:29
  Netamain said Wednesday it had
• 5:31
  migrated away from running a
• 5:33
  decentralized verify network on
• 5:35
  layer zero and joined the Chainlink
• 5:36
  network as a node operator and
• 5:38
  strategic technology provider,
• 5:40
  helping secure Chainlink's CCIP
• 5:43
  and data feeds.
• 5:44
  It will also contribute engineering
• 5:46
  tools and integration support for
• 5:48
  teams building on chain.
• 5:49
  It's a meaningful realignment among
• 5:52
  the providers competing to secure
• 5:54
  cross-chain messaging.
• 5:56
  And really impressive to see Sergei
• 5:58
  Nazarov, the founder of
• 6:00
  Chainlink speaking right
• 6:02
  next to Donald Trump yesterday as
• 6:03
  well.
• 6:04
  Moving on, there's still a pulse on
• 6:07
  crypto market structure legislation.
• 6:09
  Democratic Senator Ruben Gallego,
• 6:11
  a key negotiator on the Clarity
• 6:12
  Axis, Congress can still...
• 6:19
  Paul's talks.
• 6:20
  Speaking at SALT once again, we had
• 6:22
  plenty of news coming out of the
• 6:23
  conference.
• 6:24
  Gallego said multiple versions of an
• 6:25
  ethics provision have been sent to
• 6:27
  President Trump with nothing to show
• 6:29
  for it. In his words, we've been
• 6:31
  sending offers over and over again
• 6:33
  to the White House, and they've been
• 6:35
  coming back either blank or even
• 6:37
  further back, or we've heard
• 6:38
  nothing.
• 6:39
  The path forward is narrow, but
• 6:41
  negotiators say that
• 6:43
  it is enclosed.
• 6:43
  If you're looking for any of that
• 6:45
  news coming out SALT conference,
• 6:47
  You can check that out on the Blocks
• 6:49
  website.
• 6:50
  Lots and lots of great
• 6:52
  interviews and great anecdotes
• 6:54
  coming from different industry
• 6:56
  leaders out of the US.
• 6:57
  It's been a big day for
• 6:59
  crypto on Wednesday in
• 7:01
  the US and that's about
• 7:03
  it for the Block Briefing this
• 7:05
  morning or afternoon, wherever
• 7:06
  you're joining us.
• 7:07
  We're gonna be moving into the hot
• 7:09
  seat with
• 7:11
  Avery Ching, who's of course the
• 7:12
  co-founder and CEO of
• 7:14
  Aptos. Avery Great to have you
• 7:16
  back with us.
• 7:17
  We were last sitting next to each
• 7:18
  other at
• 7:20
  Consensus in Miami.
• 7:21
  So first things first, how are you
• 7:22
  doing?
• 7:23
  I'm doing great.
• 7:24
  It's good to see you again since
• 7:25
  Consensus.
• 7:28
  Yeah, let's jump straight into the
• 7:30
  conversation, Avery.
• 7:31
  In the last two months since we last
• 7:33
  saw each other, I'm sure it's been
• 7:35
  full steam ahead.
• 7:35
  Can you give us a rundown of what
• 7:37
  you've been really devoting most
• 7:39
  of your time and energy to in recent
• 7:41
  weeks?
• 7:42
  I think the main focus
• 7:44
  for us has been looking at the
• 7:46
  applicability of AI technology and
• 7:48
  how crypto can play a big role in
• 7:50
  that. And I know a lot of people
• 7:51
  have been trying to crack this nut
• 7:52
  for a long time.
• 7:54
  And we see so many different trends
• 7:56
  in the space that are making
• 7:58
  tremendous headway.
• 8:00
  Recently this new Quen 3.8,
• 8:02
  27 billion parameter model was
• 8:03
  released that can perform at
• 8:05
  level of Opus 46,
• 8:08
  a frontier model just six months
• 8:09
  ago. And you can do it on
• 8:11
  your Mac, your M5
• 8:14
  Mac, sorry.
• 8:15
  And that's incredible, that's an
• 8:16
  incredible change, right?
• 8:17
  Taking these massive clusters of
• 8:19
  GPUs and getting the performance of
• 8:22
  the same kind of effect six months
• 8:23
  ago in
• 8:25
  a high-end prosumer
• 8:27
  like CPU,
• 8:29
  that's amazing.
• 8:30
  And that kind of shift.
• 8:32
  It doesn't just affect crypto, it
• 8:34
  affects every single industry out
• 8:35
  there.
• 8:36
  So taking those things and applying
• 8:38
  it to where the best way for
• 8:39
  blockchain and AI to collaborate,
• 8:41
  I think we've got a good handle on
• 8:42
  it and we're really excited for the
• 8:44
  future in that space.
• 8:48
  Oh, I think you're muted, Garrett.
• 8:52
  Sorry about that.
• 8:53
  Let's pull on that thread a little
• 8:54
  bit. We've had some great guests and
• 8:57
  speakers from the industry joining
• 8:58
  us on the show in recent weeks, you
• 8:59
  know, the likes of Niaz Ilya
• 9:02
  Polisukhin and a few other
• 9:04
  players that are really delving deep
• 9:06
  into the intersection of blockchain
• 9:07
  technology and AI agents.
• 9:10
  And you've been speaking quite
• 9:11
  passionately about this intersection
• 9:13
  as well. Can you give us your kind
• 9:15
  of thesis on why
• 9:17
  blockchain technology in this stack
• 9:19
  is so well suited for.
• 9:22
  AI and perhaps agentic
• 9:24
  commerce in general and where you
• 9:26
  see this going.
• 9:27
  A lot of people talk about agented
• 9:29
  commerce. I think that
• 9:30
  to me is a bit of a
• 9:32
  longer cycle, to be honest, because
• 9:35
  the first thing agents will do is
• 9:36
  actually leverage our credit cards
• 9:37
  and existing financial rails to pay
• 9:39
  for services, for goods.
• 9:42
  If you wanna book an appointment on
• 9:43
  a reservation for you and
• 9:45
  your spouse the next week, you're
• 9:47
  gonna use a credit card for that.
• 9:49
  However, you know the future is
• 9:50
  going to be when agents start to get
• 9:52
  more prolific, when you and I each
• 9:54
  have 10 to 30
• 9:56
  to 100 agents all working on our
• 9:57
  behalf and corporations have the
• 9:59
  same number working on each
• 10:00
  employee's behalf, then it becomes a
• 10:02
  much more interesting situation
• 10:04
  where these agents will sort of
• 10:05
  negotiate with each other, use
• 10:07
  currencies that they both can
• 10:08
  operate with very easily with and I
• 10:10
  think that's where the digital rails
• 10:11
  come into play. So it's more of a
• 10:13
  later cycle, but I think overall
• 10:15
  you know it's important and you know
• 10:17
  people are trying very hard to
• 10:18
  increase distribution of digital
• 10:20
  payments immediately i i do think
• 10:22
  that there are interesting unique
• 10:23
  aspects here you can imagine
• 10:25
  even things like a P2P
• 10:27
  networks, right, were something that
• 10:28
  were pretty big a while back.
• 10:30
  They all died out. And one of the
• 10:31
  big things I think that was missing
• 10:33
  from that was the ability for
• 10:35
  people to get paid in those
• 10:36
  networks, right, if you're providing
• 10:37
  a service, you're setting up
• 10:40
  file infrastructure for goods and
• 10:42
  for people that download things,
• 10:43
  you know, you probably need to be
• 10:44
  compensated for that. And I think
• 10:45
  with those things in mind,
• 10:48
  the next generation P2 P networks
• 10:49
  can be powered by kind of more
• 10:50
  genetic workflows, by more
• 10:52
  protocols, as well as the
• 10:53
  ability to pay for services.
• 10:55
  So it's a cross between maybe
• 10:58
  DeFi and
• 11:00
  AI
• 11:02
  and physical infrastructure as
• 11:05
  well.
• 11:06
  But those kind of things, I think,
• 11:07
  are much more interesting from agent
• 11:09
  to agent world than the kind of
• 11:10
  agents like negotiating and paying
• 11:11
  for stuff. Because that's a problem
• 11:13
  that has to be solved much more
• 11:15
  by deeper, deeper issues.
• 11:17
  So today, we all kind of use agents
• 11:19
  to just I don't know,
• 11:21
  we're scratching the surface here.
• 11:24
  A lot of us are using agents for
• 11:26
  chat, maybe for some specific tasks.
• 11:28
  The biggest use case of agents today
• 11:30
  is coders.
• 11:31
  Myself and the whole team and all
• 11:33
  the technical teams will build
• 11:35
  things with agents, and that's
• 11:36
  fantastic.
• 11:36
  We'll search for bugs with agents
• 11:38
  and that fantastic.
• 11:39
  But the rest of the space,
• 11:41
  like legal, media,
• 11:44
  HR, all this other stuff is just
• 11:45
  starting to grow, just starting
• 11:46
  right to take off.
• 11:47
  And, you know, we're just hitting
• 11:49
  probably like.
• 11:50
  Not even a percent of what it's
• 11:51
  going to be in the future.
• 11:52
  Inference costs are coming down,
• 11:54
  usage is growing.
• 11:55
  And so this is like a pretty clear
• 11:57
  use case for why we
• 11:59
  need to be confident with AI.
• 12:00
  We need to have the ability for us
• 12:01
  to trust these systems.
• 12:03
  A lot of those reasons is because of
• 12:04
  trust, a lot of reason because of
• 12:05
  inference is not cheap, and
• 12:07
  also the usability aspect,
• 12:09
  right? It's not easy to use these
• 12:10
  agents.
• 12:11
  So, when you solve all three,
• 12:14
  and I think blockchain is a big
• 12:15
  piece of this, we're going to see a
• 12:16
  tremendous, tremendous change, not
• 12:18
  just in crypto, but across the
• 12:20
  entire world where crypto is a part
• 12:22
  of it, blockchain technology is a
• 12:23
  important part of this.
• 12:25
  I stand to be corrected here, but
• 12:27
  some anecdotes online say that
• 12:29
  Aptos Foundation and AptoS Labs has
• 12:31
  committed something around $50
• 12:32
  million to back network
• 12:34
  infrastructure for various
• 12:36
  parts of the
• 12:38
  AI space. Can you talk to us a
• 12:39
  little bit about what you think is
• 12:41
  the most important kind of
• 12:44
  development area and where
• 12:46
  you would like to see that money
• 12:47
  best spent?
• 12:49
  Yes, I think I love it.
• 12:50
  Those funds are going towards, you
• 12:52
  know, commentary I as well as the
• 12:53
  trading aspects. So markets and
• 12:55
  machines has been something that
• 12:56
  we've been focused on for well
• 12:57
  over a year now.
• 12:58
  It's something that continues to
• 13:00
  think to be something we believe
• 13:02
  in.
• 13:03
  On the market side, it's
• 13:05
  really about a conglomeration of
• 13:07
  services here, right? So we have
• 13:07
  decibel, which is providing perps
• 13:10
  and spot at unprecedented
• 13:11
  performance ever seen before
• 13:13
  as a decentralized product.
• 13:15
  Spot just launched a couple days
• 13:16
  ago. It is doing great.
• 13:18
  We'll start to see all kinds of
• 13:19
  assets coming on board spot, but the
• 13:21
  next range of that is going to be
• 13:22
  products on top, vaults,
• 13:24
  yield. All these things that,
• 13:26
  you know, once you have the core
• 13:27
  basics, you can sort of build a
• 13:29
  foundation on top of.
• 13:31
  And we think that that's exciting.
• 13:32
  We also see tokenizing workstreams
• 13:34
  popping off these days.
• 13:36
  Many of these, even if, you now,
• 13:38
  whatever happens to Clarity, I know
• 13:39
  it's in a very interesting state
• 13:40
  right now.
• 13:42
  Making progress one day,
• 13:44
  setbacks the next.
• 13:46
  But the reality is every single
• 13:47
  large financial player is thinking
• 13:49
  about tokenization very seriously in
• 13:51
  the US.
• 13:53
  That's a shift we haven't seen in
• 13:55
  an entire like eight plus years of
• 13:56
  crypto, people are thinking about
• 13:58
  experimenting with it, but now
• 13:59
  they're really taking it to product
• 14:01
  and commercialization and
• 14:03
  it's very positive.
• 14:04
  The other thing that's interesting
• 14:05
  is going to be along the payments
• 14:06
  work streams that we've seen as
• 14:07
  well.
• 14:09
  Payments as a product is growing
• 14:10
  tremendously.
• 14:11
  We see that there's
• 14:13
  new organizations coming out like
• 14:14
  OUSD, which we're part of,
• 14:16
  but. You know, others like Tether
• 14:18
  and Circle continue to
• 14:20
  make inroads and grow.
• 14:22
  And their use cases and adoption are
• 14:23
  starting to take off in different
• 14:25
  ways. And others are starting
• 14:26
  recognize that value.
• 14:27
  So from the market side, you know,
• 14:29
  the tokenization, the payments
• 14:30
  workstreams, yield workstream,
• 14:32
  trading infrastructure, spot,
• 14:35
  perhaps, and then also kind
• 14:37
  of AI playing alongside that we
• 14:38
  feel is a very, very strong pillar
• 14:40
  for the use of blockchain
• 14:41
  technology. But more importantly,
• 14:43
  it's going to be an important pillar
• 14:44
  of the new financial infrastructure
• 14:45
  for the world.
• 14:46
  The second piece that's very
• 14:47
  interesting is AI as a technology
• 14:49
  itself, besides enhancing the
• 14:51
  space of finances, is going to
• 14:53
  be very transformational when it
• 14:54
  comes to the growth
• 14:56
  of all software development.
• 14:57
  I think we're all seeing transitions
• 14:58
  in there.
• 14:59
  Our blockchain is protected by AI,
• 15:01
  it is developed by AI.
• 15:03
  It is something, a tool that is
• 15:05
  indispensable to engineers that
• 15:06
  operate on it.
• 15:07
  We have very sophisticated
• 15:09
  systems where AI will triage and
• 15:11
  debug issues before they become
• 15:12
  major issues.
• 15:14
  All autonomously without our
• 15:16
  intervention, and then humans take
• 15:17
  the intervention when it comes to
• 15:18
  pushing buttons today.
• 15:20
  In the future, when the AI
• 15:22
  demonstrates it can do those
• 15:24
  things properly time and time and
• 15:25
  time again, humans won't be in the
• 15:27
  loop. But today, we definitely, for
• 15:28
  the sake of the blockchain and for
• 15:30
  the billions of dollars of value
• 15:31
  stored out, we of course make sure
• 15:33
  humans are in the loo.
• 15:34
  But we see the promise.
• 15:36
  We see that in five or ten years,
• 15:38
  that system will be fully automated.
• 15:39
  So it's very natural for
• 15:41
  every software company to take a
• 15:42
  strong investment.
• 15:44
  Into AI infrastructure, but for us
• 15:46
  we think that AI plus crypto
• 15:47
  provides a unique value prop
• 15:49
  that hasn't been seen before that's
• 15:50
  why we developed Shelby with Jump
• 15:52
  Crypto to build out
• 15:54
  a new system
• 15:56
  for AI storage and alongside that
• 15:58
  infrastructure. We really compute
• 15:59
  services and we're building
• 16:01
  application services that
• 16:03
  all leverage the full combination
• 16:04
  layer of it and I think it's gonna
• 16:06
  be a very transformational thing for
• 16:07
  people because the thing holding
• 16:09
  most people back is that
• 16:11
  trust of services is the ability to
• 16:12
  kind of compartmentalize
• 16:13
  information. It is really, AI is an
• 16:15
  extension of my thinking.
• 16:17
  It's an extension your thinking, it's
• 16:18
  extension ourselves. And today, our
• 16:20
  brains really protect that
• 16:20
  information. And yes, we can extract
• 16:22
  some of the information through
• 16:23
  these interviews or through
• 16:24
  discussions, but we control what
• 16:26
  the information flow is.
• 16:27
  You know, when you have a system
• 16:28
  like, you know, chat CPT or
• 16:31
  others that are kind of ingesting
• 16:32
  the data and using it for training,
• 16:33
  you know it's really extracting a
• 16:35
  lot of the value from.
• 16:36
  From what we think about, how we
• 16:38
  store and process information,
• 16:40
  and so that confidentiality aspect
• 16:41
  and the ability for blockchain to be
• 16:43
  an immutable ledger of record for
• 16:45
  all this information is, I think, a
• 16:47
  game changer for this to really take
• 16:48
  off as technology to enable a
• 16:50
  thousand the use of today's AI.
• 16:53
  Yeah, beautiful.
• 16:54
  There were a couple of threads I
• 16:55
  wanted to pull on as you spoke
• 16:56
  there. And I think the first one you
• 16:58
  mentioned is Decibel, and it might
• 16:59
  take us in a little bit of a
• 17:00
  different direction of the
• 17:02
  conversation now.
• 17:03
  But there's been a huge amount of
• 17:05
  interest in hyperliquid as
• 17:08
  a decentralized kind of trading
• 17:10
  venue.
• 17:11
  And obviously we saw a lot of
• 17:13
  interest yesterday with President
• 17:14
  Donald Trump talking about that.
• 17:16
  Do you see Decibe as a sort of
• 17:18
  competitor in the same
• 17:20
  mold as hyperliqid in.
• 17:22
  Can you talk to us a little bit
• 17:23
  about the differences between the
• 17:24
  two and perhaps the opportunities at
• 17:26
  hand?
• 17:27
  Yeah, I mean, I think it's great to
• 17:29
  hear the president, you know,
• 17:30
  talking about how they're going to
• 17:31
  bring
• 17:33
  decentralized products into the US.
• 17:35
  That's been something that's a real,
• 17:36
  real challenge for everyone building
• 17:37
  the space, you know, since forever.
• 17:40
  So that's very positive development.
• 17:42
  And I see some of the key differences
• 17:44
  between hyper liquid decibel is
• 17:46
  hyper liquid infrastructure is kind
• 17:47
  of built on more close,
• 17:49
  you know, it's close source code is
• 17:51
  It is not the same decibel
• 17:53
  is the kind of only open source,
• 17:56
  you know, a base product where it
• 17:58
  relies completely on smart contract
• 17:59
  infrastructure.
• 18:00
  It is running on a fully
• 18:01
  decentralized platform.
• 18:03
  It is all kind of more compliant
• 18:05
  with the way that the CFTC and SEC
• 18:06
  have given guidance around what
• 18:09
  infrastructure should be no central
• 18:10
  centralization, closed source as
• 18:13
  a point of decentralization and
• 18:15
  also kind of no biasing towards
• 18:17
  particular components and open
• 18:18
  source as being very important for
• 18:20
  that.
• 18:23
  And so I think that's one
• 18:24
  key differentiator.
• 18:26
  But the other one is going to just
• 18:27
  be, look, there's so many things you
• 18:28
  can do once you have it built into
• 18:30
  the smart contracts layer.
• 18:31
  So HyperLiquid is built as kind of a
• 18:32
  standalone product.
• 18:35
  And they have an interface layer
• 18:37
  to hybrid VM, but it's obviously
• 18:39
  different performance
• 18:40
  characteristics, different
• 18:41
  accessibility and access to the
• 18:43
  product features.
• 18:44
  Decibel is completely built on smart
• 18:45
  contracts and infrastructure, which
• 18:47
  means everyone can plug it.
• 18:48
  But everyone can kind of see the
• 18:49
  different mechanics and the entire
• 18:50
  order book is completely visible.
• 18:53
  The code is visible through move.
• 18:55
  Everything is there.
• 18:58
  And that's, I think, a very
• 19:00
  different kind of infrastructure
• 19:00
  composability aspect.
• 19:02
  And one that's much more
• 19:05
  A longer term would be safer,
• 19:07
  like we've seen with all open source
• 19:08
  projects in the future, they become
• 19:09
  safer because people can peer into
• 19:10
  them, inspect them, find holes and
• 19:12
  bugs, close source code,
• 19:14
  it's security by obscurity, and
• 19:16
  that's just a different trade-off.
• 19:19
  Additionally, I think just the
• 19:20
  on-chain aspect of it and the
• 19:21
  composability of it will also do
• 19:23
  very novel things.
• 19:24
  For example, making a delta-neutral
• 19:25
  trade in Decibel is going
• 19:27
  to be much far easier because
• 19:29
  everything's a smart contract line.
• 19:30
  I can initiate these two
• 19:31
  transactions and make them compose
• 19:33
  like a long Bitcoin on
• 19:35
  the spot and short on purpose.
• 19:36
  At the same time, close that trade
• 19:38
  in a single topic.
• 19:39
  Those are the things that are,
• 19:41
  I think, very interesting for
• 19:43
  composability and aspects of
• 19:44
  infrastructure of trading, and then
• 19:46
  exposing those features as
• 19:48
  vaults, exposing those features into
• 19:50
  a consumer base that can look for
• 19:51
  yield above the 3% to
• 19:53
  4% of the treasury bonds yield.
• 19:56
  You can get 68% on a Bitcoin
• 19:58
  product or more very safely,
• 20:00
  leveraging some of the same
• 20:02
  properties of what Athena does.
• 20:04
  That's a very exciting, I think,
• 20:05
  future for where this product
• 20:07
  goes. So you can think of Exchange
• 20:08
  as not just a front-end product,
• 20:10
  but as a base layer in which you can
• 20:12
  build a whole suite of products on
• 20:13
  top of that we're very excited
• 20:16
  And when it comes to OUSD,
• 20:19
  you mentioned that briefly in
• 20:21
  the previous question that I'd asked
• 20:23
  you.
• 20:23
  Can you talk to us a little bit
• 20:24
  about the Open Standard Consortium?
• 20:27
  How Aptos got to be
• 20:29
  involved? What that's going to look
• 20:31
  like going forward?
• 20:32
  And how powerful it is having this
• 20:35
  collection of the world's biggest
• 20:37
  financial institutions coming
• 20:39
  together to explore a
• 20:42
  different way of doing stablecoins?
• 20:43
  Because obviously we've got two
• 20:44
  giants. In the space in
• 20:46
  Tether and Circle.
• 20:48
  And this is everyone else saying,
• 20:49
  hey, we're gonna join this race
• 20:52
  and start something special here.
• 20:55
  Yeah, we're definitely excited to
• 20:56
  see it.
• 20:57
  And I think we've been, of course,
• 20:59
  very happy with our partners in
• 21:01
  Tether and Circle.
• 21:03
  But the thing about USD that's
• 21:05
  interesting, I think, is just it
• 21:06
  brings me more back to the ideas
• 21:07
  where we had in the neighborhood in
• 21:09
  2018, where it
• 21:11
  was a consortium-based approach.
• 21:12
  It's more about the
• 21:15
  ability for the distribution
• 21:17
  and usage of the token, as opposed
• 21:19
  to just necessarily one party,
• 21:21
  kind of driving everything forward.
• 21:24
  And the especially interesting thing
• 21:25
  is the pass back of
• 21:27
  the revenue back to
• 21:29
  the distribution partners completely
• 21:31
  or nearly completely is a game
• 21:33
  changer when it comes to
• 21:35
  what we think is the growth aspect
• 21:36
  of it. And so that piece
• 21:38
  of it is going to be interesting.
• 21:39
  I think it will hopefully be
• 21:42
  very well received in the market and
• 21:44
  become something that people start
• 21:46
  to rally around when it comes to
• 21:47
  adoption. They don't have to think,
• 21:48
  well, I need to kind of create my
• 21:50
  own stable coin because I need get
• 21:51
  my own yields out of it, it's more
• 21:53
  like, well, there's already a
• 21:54
  product that exists, it's
• 21:56
  backed by a consortium, it's
• 21:57
  something that provides all the
• 21:59
  functionality that you need.
• 22:00
  And then the question is, why do you
• 22:02
  need your own stable in
• 22:04
  this mix?
• 22:05
  Obviously, Tether and Circle have
• 22:06
  great use cases out there, and
• 22:07
  they're going to continue to do
• 22:08
  that, I'm sure.
• 22:09
  And so I don't think there's space
• 22:11
  just for one stable that
• 22:13
  kind of rules them all.
• 22:14
  There are different aspects to these
• 22:15
  products. It's just like there's no
• 22:17
  one deposit product that rules them
• 22:19
  all, there's one yield product that
• 22:20
  owns them all it's kind of the same
• 22:22
  idea.
• 22:23
  And you can almost think of these
• 22:24
  stablecoins more as yield products
• 22:25
  than it's actually just purely
• 22:27
  fiat cash rolling around your
• 22:29
  pocket.
• 22:30
  It's an interesting proposition as
• 22:32
  well, especially if you just
• 22:33
  consider that Tether, at the end
• 22:35
  of the day, takes those deposits
• 22:37
  and invest
• 22:40
  that money in different things.
• 22:41
  Obviously, they hold cash reserves,
• 22:43
  but they're one of the biggest
• 22:45
  US Treasury bondholders
• 22:48
  in the world as well.
• 22:49
  That's become a billion-dollar
• 22:52
  revenue generating business
• 22:55
  every quarter, which is crazy to
• 22:57
  think about.
• 22:58
  It does suggest
• 23:00
  that there is a space to
• 23:03
  compete with that if
• 23:05
  you can actually give the
• 23:07
  depositors something back because at
• 23:09
  the end of the day, we're all giving
• 23:11
  to the dollar deposits to use
• 23:13
  the utility of USDT
• 23:15
  across the protocols
• 23:17
  that it is sent
• 23:19
  across, but they're the ones
• 23:21
  that make the money at the of the
• 23:22
  end the day. And there is something
• 23:24
  to be said about sharing that value
• 23:25
  back with the people that are.
• 23:27
  Depositing their money just for the
• 23:29
  utility alone.
• 23:30
  Yeah, I think it carves out a new
• 23:31
  space, right? Tethers can do
• 23:33
  that because they've established
• 23:34
  such a great corridor of payments
• 23:35
  infrastructure.
• 23:37
  They've got the PSPs and they've got
• 23:38
  the regional providers and they got
• 23:40
  a huge network that essentially
• 23:42
  people pay for, right.
• 23:43
  That's what they're willing to pay
• 23:44
  for. And it's proven its value
• 23:46
  over the years.
• 23:47
  Same thing with Circle.
• 23:50
  And I think OECD will carve its own
• 23:52
  path based on its ability
• 23:53
  to grow distribution for its use
• 23:55
  cases. And I don't think it's
• 23:57
  a win-win-win that takes all the
• 23:58
  situation, to be honest.
• 24:00
  No, I don't think so at all.
• 24:02
  And as you can, I mean, like we even
• 24:03
  saw Ripple a couple of years ago
• 24:04
  move into the stable coin space.
• 24:07
  There's still plenty of competition
• 24:09
  and room for new
• 24:11
  players.
• 24:12
  I wanted to talk to you a little bit
• 24:13
  about the sort of
• 24:15
  the development of AI in
• 24:17
  the last year and its
• 24:19
  impact on both security
• 24:21
  incidents and code
• 24:23
  reviewing. So, you know, in the DeFi
• 24:25
  ecosystem, especially in Ethereum,
• 24:27
  we've seen some pretty large
• 24:29
  exploits. Albeit not being
• 24:31
  smart contracts exploits,
• 24:33
  largely social engineering attacks,
• 24:34
  those kind of things.
• 24:35
  In the Bitcoin ecosystem in the last
• 24:37
  four weeks, we've seen the cold
• 24:39
  card exploits.
• 24:40
  And then the Bitcoin Red team going
• 24:42
  about using frontier
• 24:44
  models to go and find exploits
• 24:46
  and disclose them to
• 24:48
  teams that are building part
• 24:50
  of Bitcoin's open source code so
• 24:52
  that they can go and fix them and
• 24:53
  make those
• 24:55
  exploits go away or the potential
• 24:57
  exploits be nullified.
• 25:00
  From your perspective, how much
• 25:02
  have you been leaning on AI
• 25:05
  tools to review everything that
• 25:07
  Aptos has built till now and perhaps
• 25:08
  some of the tools and
• 25:10
  systems that have been built in your
• 25:12
  ecosystem and how powerful has that
• 25:13
  become?
• 25:15
  So AI is, I think, used at pretty
• 25:17
  much every level of defense in the
• 25:18
  infrastructure today. It's used for
• 25:20
  some code review.
• 25:22
  We used to have two code reviewers
• 25:24
  for every single PR that landed.
• 25:26
  We're a little bit paranoid.
• 25:27
  I think we're only able to add the
• 25:29
  policy in the blockchain-based
• 25:30
  space.
• 25:31
  And now we feel like with AI
• 25:33
  being able to take on the reviewing
• 25:35
  of large amounts of code, we have
• 25:37
  multiple bots that do that,
• 25:39
  and agents.
• 25:40
  And then we now can get down to
• 25:42
  one, which is great. So that already
• 25:43
  saves half the review time for a
• 25:44
  human.
• 25:46
  And the codes seem to be safer.
• 25:48
  We've seen less outcomes of
• 25:50
  negativity from that because the eye
• 25:52
  reviewers are actually quite good.
• 25:54
  So that's one of the first lines of
• 25:55
  defense even before code screen
• 25:56
  system.
• 25:58
  Then, of course, we still employ
• 26:00
  human security
• 26:02
  agencies to go and review the code.
• 26:04
  They're using significant amount of
• 26:05
  AI tools, building tools on their
• 26:06
  behalf.
• 26:08
  And we start to actually build very
• 26:09
  custom tools within the Optos
• 26:11
  ecosystem internally for
• 26:13
  doing kind of deep audit searches as
• 26:15
  well as variant searches that,
• 26:17
  you know, if you see a bug that we
• 26:19
  looked at in the past, we can look
• 26:19
  for all instances of that bug
• 26:22
  in the code base.
• 26:24
  And that's allowing us to
• 26:26
  be very proactive when it comes to
• 26:28
  attacking these kind of bugs that
• 26:30
  may cause network
• 26:31
  to halt, processes
• 26:34
  not behaving correctly, maybe in the
• 26:35
  worst case, loss of funds.
• 26:37
  And the final step, of course, is if
• 26:39
  there is something that gets through
• 26:41
  somehow to mainnet, which we haven't
• 26:43
  seen ever in the Atlas ecosystem,
• 26:45
  you have AI agents
• 26:47
  deployed at the ready to look
• 26:49
  for anomalies within logs.
• 26:50
  And once you find the numbers and
• 26:52
  logs...
• 26:52
  They immediately notify operators.
• 26:54
  Operators can go ahead and triage
• 26:55
  that using AI tools to
• 26:57
  automatically kind of piece through
• 26:58
  the logs. You know, there are so
• 26:59
  many different complex logs in the
• 27:00
  system, they're validator logs.
• 27:02
  There are gonna be RPC logs.
• 27:04
  There's gonna be logs around
• 27:06
  our indexing services.
• 27:09
  So many different logs, right?
• 27:10
  So piecing them together is
• 27:11
  something that AI can do really,
• 27:12
  really well. Pulling from a
• 27:13
  different place is using NCP.
• 27:15
  And then kind of triaging incident
• 27:17
  quickly, having instant response
• 27:18
  very, very fast to it, and then
• 27:20
  deploying fixes. So.
• 27:22
  I'd say throughout the entire
• 27:23
  pipeline, you're leveraging AI,
• 27:25
  but I think the important
• 27:27
  thing to notice, there are different
• 27:28
  industries.
• 27:29
  There's a blockchain industry that
• 27:31
  secures billions of dollars of
• 27:32
  funds.
• 27:33
  There are also more products that
• 27:35
  have less loss if there are bugs.
• 27:36
  I used to work at Facebook.
• 27:38
  If we made a bug, we'd probably just
• 27:39
  roll it forward. Not a big deal.
• 27:41
  If it's blockchain,
• 27:43
  you can't just roll forward a bug
• 27:45
  that fixes a loss of a
• 27:47
  billion dollars.
• 27:49
  That's definitely a different kind
• 27:50
  of issue and we have to take it very
• 27:52
  seriously.
• 27:53
  So we're very, you know, very
• 27:55
  much using AI in the flow of
• 27:56
  everything when it comes to
• 27:58
  blockchain. And I think in many
• 27:59
  ways, we're like the first kind of
• 28:03
  AI protected blockchain and
• 28:06
  we have the tooling for it too.
• 28:07
  So Move has been a big piece of this
• 28:09
  where we built Move because
• 28:11
  we wanted to be more secure.
• 28:12
  We just started developing it more
• 28:13
  than eight years ago at Facebook.
• 28:16
  Today, I think that
• 28:18
  bet has paid off in spades, in the
• 28:20
  sense that it's got some lower
• 28:22
  layers of security above Ethereum,
• 28:23
  besides the static compiler checker,
• 28:25
  you've got the dynamic checker at
• 28:26
  runtime, so while your
• 28:28
  code is running, you have a check
• 28:30
  that makes sure that you can't have
• 28:31
  these invariants that are broken,
• 28:32
  and you also have this move prover,
• 28:34
  which is constantly running to
• 28:36
  make sure the code you deployed as a
• 28:38
  smart contract is going to behave
• 28:39
  the way you think it's going to be.
• 28:41
  And now these tools can be developed
• 28:42
  also much quicker, like we can build
• 28:43
  them and verify even more powerfully
• 28:45
  with AI.
• 28:47
  And so those are, I think, some of
• 28:48
  the things that has been very
• 28:50
  transformational to the way we
• 28:51
  develop, the way we test
• 28:54
  and find bugs, and the way that we
• 28:55
  also productionize our
• 28:56
  infrastructure.
• 28:58
  It's funny how much of a focus has
• 29:00
  come on to sort of AI assisted
• 29:02
  hacking and
• 29:04
  that kind of thing when I
• 29:06
  think not even six months ago,
• 29:08
  the big focus in Bitcoin and
• 29:10
  just the wider cryptocurrency
• 29:11
  industry was being quantum safe.
• 29:15
  Can we talk a little bit about post
• 29:16
  quantum signatures, where Aptos
• 29:18
  stands with that in comparison to
• 29:20
  some of the other players in the
• 29:21
  space because I stand to be
• 29:23
  corrected, But I think Nier is
• 29:25
  is one of the only chains that has
• 29:26
  full quantum-resistant
• 29:29
  cryptography.
• 29:31
  But I know that Aptos has
• 29:33
  an opt-in for these post-quantum
• 29:34
  signatures. Can you talk to us a bit
• 29:36
  about quantum readiness and if you
• 29:38
  think people are neglecting this,
• 29:40
  given the state of AI exploits
• 29:42
  going on?
• 29:43
  So quantum readiness is something
• 29:45
  that I think every chain needs to be
• 29:46
  prepared for.
• 29:47
  I think that there's a, there is,
• 29:49
  you know, so we were kind of
• 29:50
  announcing, I think, CoinMe has put
• 29:52
  a report several months ago,
• 29:54
  kind of indicating that, you know,
• 29:55
  there are only two chains ready for
• 29:57
  quantum. That was, you know, Aptos
• 29:59
  and I think another one.
• 30:00
  I can't remember which one now,
• 30:02
  though. But That
• 30:04
  report, I think, was probably
• 30:06
  the first report to kind of indicate
• 30:08
  how serious quantum threat was
• 30:10
  when it comes to a broad set of
• 30:11
  chains. I know I was worried about
• 30:12
  Bitcoin and Ethereum primarily
• 30:14
  because Bitcoin moves so slowly.
• 30:16
  It'd be a big change to Bitcoin to
• 30:17
  introduce these new signatures and
• 30:20
  that's something developers need to
• 30:21
  prepare probably years in advance
• 30:23
  for.
• 30:24
  When it comes after this, though, we
• 30:25
  were very fast moving and our
• 30:26
  cryptography team is second to none.
• 30:28
  So we've had that change ready
• 30:31
  for a long time.
• 30:32
  That being said, the
• 30:34
  cost of paying quantum-resistant
• 30:36
  signatures is non-trivial.
• 30:38
  It probably introduces multiples
• 30:40
  of costs over a non-quantum
• 30:42
  signature, and so we don't
• 30:44
  force users to pay that cost now.
• 30:46
  We think that as the threat gets
• 30:48
  closer, it's easy to flip a switch.
• 30:50
  People can upgrade
• 30:52
  their accounts whenever they feel
• 30:53
  like it, and now we're
• 30:55
  great for quantum.
• 30:57
  It's something that we're excited
• 30:59
  about because it
• 31:01
  also will help secure.
• 31:02
  I mean, once you're quantum resistant,
• 31:04
  you'll have kind of like very long
• 31:05
  life, probably forward secrecy.
• 31:08
  Forward secrecy is important.
• 31:09
  So your secrets are revealed during
• 31:10
  a time period in which you get
• 31:11
  comfortable with that.
• 31:13
  But, you know, eventually everything
• 31:15
  is meant to be broken.
• 31:16
  Like that's just kind of the
• 31:17
  reality. So it's a never ending
• 31:18
  battle of making sure we stay ahead
• 31:19
  of the latest and greatest in terms
• 31:21
  of cryptography.
• 31:23
  But, you know, I feel very
• 31:24
  comfortable with that.
• 31:25
  With what we have in place and I
• 31:26
  think we're ahead of pretty much
• 31:28
  most places at this point.
• 31:31
  Avery, when it comes to sort of like
• 31:33
  on-chain usage and the
• 31:35
  most important metrics measuring
• 31:36
  that, obviously the block is very
• 31:37
  well known for its data dashboards
• 31:39
  and I would implore anyone
• 31:41
  who's watching the show today to go
• 31:43
  ahead and take a look at our data
• 31:45
  section of our website and all
• 31:47
  of our on-chain metrics.
• 31:49
  But you'll look at a lot of these
• 31:51
  different metrics for different
• 31:52
  chains and a lot of people will tell
• 31:53
  you the revenue
• 31:56
  of the chain is what matters most.
• 31:57
  Monthly active users is
• 31:59
  what matters most.
• 32:02
  In your mind, what are you
• 32:04
  defining as success for Aptos right
• 32:06
  now, and where do you see that focus
• 32:08
  moving in the next year?
• 32:11
  I think...
• 32:12
  All teams find to take a little hard
• 32:13
  look at themselves and think about
• 32:15
  the metrics, like you said,
• 32:18
  look, they're there.
• 32:19
  Those metrics you mentioned have
• 32:20
  their usefulness, but by
• 32:22
  and large, the way the industry is
• 32:24
  thought about them completely
• 32:24
  useless, right?
• 32:25
  Like, it's pretty easy to to
• 32:27
  get a bunch of accounts to use your
• 32:29
  chain for various purposes.
• 32:30
  And I think that's just a
• 32:32
  misnomer.
• 32:34
  What we look at is kind of like what
• 32:35
  we call real users, and it's very
• 32:36
  hard to figure out what real users
• 32:38
  are on chain.
• 32:39
  Um, but,
• 32:41
  uh, We want to
• 32:43
  see like who's
• 32:45
  not just trying to, you know, farm
• 32:47
  awards. I think that's very
• 32:48
  important. Farming awards is a
• 32:50
  temporary process.
• 32:51
  We believe that's just dying out
• 32:53
  now because we're
• 32:55
  looking for the products that
• 32:56
  generate kind of sustainable
• 32:57
  long-term yield.
• 32:58
  And then the ones I mentioned earlier
• 32:59
  around the vaults that can generate
• 33:01
  sustainable yield because,
• 33:03
  you know, Bitcoin is inherently long
• 33:05
  as a position. So it's this delta
• 33:07
  neutral trade where it kind of.
• 33:09
  Generously yielding the funding
• 33:10
  rate, I think, is a very good trade
• 33:12
  unless people have psychological
• 33:13
  changes around Bitcoin.
• 33:14
  I think that would be an interesting
• 33:16
  shift.
• 33:17
  But if that psychology changes, then
• 33:19
  you just get less yield
• 33:20
  or you just stop
• 33:22
  the product.
• 33:24
  That's a straightforward thing to
• 33:25
  do.
• 33:26
  But the idea of a
• 33:28
  chain paying
• 33:29
  rewards to customers
• 33:31
  with that outpace the kind of
• 33:33
  revenue that is generated by those
• 33:35
  products. I think.
• 33:37
  It's hard to discern from just pure
• 33:38
  metrics, but if you can discern
• 33:40
  that, that to me is the kind of real
• 33:42
  on-chain yield,
• 33:44
  where if you see revenue coming in,
• 33:46
  like in Decibel, for example, we're
• 33:47
  not putting out tremendous
• 33:49
  amounts of rewards, and as
• 33:51
  a product it's earning money in
• 33:53
  revenue, it's profitable,
• 33:55
  not profitable, but rather
• 33:56
  generating that revenue of the
• 33:58
  incentives being put out.
• 33:59
  Like that's a good sign, like that
• 34:01
  is a really good metric
• 34:03
  that we look at.
• 34:04
  We also look a lot at these
• 34:06
  commercial deals that are becoming
• 34:08
  much more common
• 34:10
  these days in which two parties
• 34:12
  partner together and instead of the
• 34:13
  partnership being one party paying
• 34:14
  another for a marketing
• 34:16
  announcement, it is really about how
• 34:18
  do you have value creation and then
• 34:20
  the value creation being shared
• 34:21
  amongst those parties.
• 34:23
  Those are the things that are not
• 34:25
  easy to put on the top of a
• 34:26
  dashboard, but.
• 34:29
  I think they're the most important
• 34:30
  metrics like real customer adoption
• 34:32
  usage as well as kind of revenue
• 34:34
  generation using the chain as a
• 34:36
  product and where the revolution can
• 34:37
  go back to the chain in a real
• 34:39
  meaningful fashion is very important
• 34:41
  because even you take say app
• 34:42
  revenue app revenue on
• 34:44
  a chain you know except for burning
• 34:46
  transaction fees which some apps
• 34:47
  made a very low transaction fees has
• 34:50
  a very muted effect on top
• 34:52
  of like what the chain is doing
• 34:53
  rights so
• 34:56
  Those kind of things are
• 34:58
  kind of where we see the
• 35:00
  metrics going forward.
• 35:00
  Kind of real adoption, real usage
• 35:02
  and very hard to measure honestly on
• 35:04
  a dashboard.
• 35:05
  But you can use a lot of new AI
• 35:06
  tools to go and track kind of
• 35:08
  customer behavior and then determine
• 35:09
  is this a real user?
• 35:11
  Is this like someone who is just a
• 35:12
  temporary user or is this someone
• 35:14
  like...
• 35:14
  Maybe using 100 accounts to kind of,
• 35:16
  you know, game a system, like those
• 35:18
  are, those are the things that I
• 35:19
  think you need to look at and then
• 35:21
  kind of figuring out like what is
• 35:22
  the actual user account?
• 35:23
  What is the, you know, the type of
• 35:25
  user that's coming in here,
• 35:27
  using a product generating yield,
• 35:29
  trading on a product, you're doing a
• 35:31
  payments operation, and
• 35:33
  try to identify this is an actual
• 35:34
  customer versus like this is someone
• 35:36
  just trying to rewards
• 35:38
  in the system.
• 35:39
  Yeah, and I think that obviously the
• 35:41
  consistency or longevity of some
• 35:43
  of these things matters as well.
• 35:45
  Like Robinhood chain launched
• 35:47
  basically less than a month ago,
• 35:49
  saw a huge spike in sort of
• 35:51
  revenue fees, and it'll be
• 35:53
  interesting to see if they can
• 35:55
  continue that because it was
• 35:57
  a sort of mixed bag between
• 35:59
  memecoin launches on the
• 36:01
  chain and then RWA
• 36:03
  tokenization you know, security
• 36:05
  is on chain.
• 36:06
  I think all of us in the industry
• 36:07
  would love to see that be the bulk
• 36:09
  of securities on chain and sort
• 36:11
  of a new era
• 36:13
  for traditional finance being on
• 36:15
  the rails that we've built.
• 36:18
  But that needs to be proven over
• 36:19
  time and I'm not quite sure we're
• 36:21
  there yet.
• 36:22
  But this is also something that
• 36:23
  Aptos is leaning into I would
• 36:25
  imagine is very much the
• 36:27
  tokenization route and whether that
• 36:29
  security is all sorts of different
• 36:30
  RWAs.
• 36:31
  Do you see that as being a primary
• 36:33
  driver of the use?
• 36:34
  Of the protocols that we've built
• 36:36
  because we've got
• 36:38
  tens of very good blockchain
• 36:40
  protocols that have great
• 36:42
  fundamental metrics and use cases.
• 36:45
  But is there a winner takes all
• 36:47
  situation here in your mind?
• 36:49
  I don't think there's a vertex-all
• 36:50
  situation here either, to be honest,
• 36:51
  but I think what...
• 36:52
  What is exciting is that when
• 36:54
  we're having discussions now with
• 36:55
  potential partners, the large
• 36:58
  transfer agents that can
• 37:00
  have access to trillions of dollars
• 37:02
  of assets, literally, or
• 37:03
  the large kind of financial
• 37:05
  structures in the US,
• 37:07
  what we're seeing is that tend to
• 37:09
  tokenization and trading,
• 37:11
  they're taking a strong interest in
• 37:12
  and that's very positive
• 37:14
  for us because when they start to
• 37:16
  look at the product and less about
• 37:17
  the hype cycle, you know, I said in
• 37:18
  the previous cycle was kind of more,
• 37:20
  how much can you pay me to kind of,
• 37:22
  you to do an experiment with you.
• 37:24
  I think now that's
• 37:25
  just a matter of like, how much do
• 37:27
  you want to pay for that, right?
• 37:27
  And does it help you in your
• 37:29
  endeavors?
• 37:30
  Today it's more about how do we
• 37:32
  build value and create value?
• 37:33
  And when we create value, it means
• 37:34
  we need to look at the product very
• 37:36
  strongly.
• 37:37
  That product needs several aspects
• 37:38
  from what we hear.
• 37:39
  It needs to be really secure, right,
• 37:41
  because anyone who manages anywhere
• 37:43
  near a trillion dollars of assets,
• 37:44
  like the first thing that they're
• 37:46
  really selling is the in their
• 37:47
  platform and their product.
• 37:49
  One hack we just talked about
• 37:51
  earlier can change the whole thing.
• 37:52
  One downtime can change whole thing
• 37:53
  and we've seen your app does
• 37:55
  is pretty much the only chain that
• 37:57
  can provide high
• 37:59
  security, amazing speed,
• 38:02
  high scalability, as well
• 38:04
  as customizability through move
• 38:06
  in an entire market.
• 38:08
  Everything else has seen tremendous
• 38:09
  amounts of downtime
• 38:12
  or loss of funds or
• 38:15
  just not the flexibility in the
• 38:17
  smart contracts platform to do
• 38:18
  things you need to get done.
• 38:19
  And of course, that is the most
• 38:20
  affordable chain out there.
• 38:21
  So these aspects,
• 38:23
  we're built for consumer adoption,
• 38:26
  for product adoption, for
• 38:27
  enterprise-grade adoption.
• 38:28
  And so when it becomes less about
• 38:31
  the let's run an experiment together
• 38:33
  and more about let's build a great
• 38:34
  product that generates revenue and
• 38:36
  value and creation for people, that
• 38:38
  I think is a tremendous game changer
• 38:40
  in the market.
• 38:42
  My final question, Avery,
• 38:44
  when it comes to
• 38:46
  characteristics of chains or
• 38:48
  protocols that are setting them
• 38:50
  apart and will be really important
• 38:51
  in the future, do you think privacy
• 38:53
  still remains really high up
• 38:55
  there? Because it seems like most
• 38:57
  people accept that
• 38:59
  security is pretty good across these
• 39:01
  chains and I don't think,
• 39:03
  unless you're really like a
• 39:05
  Bitcoin maxi or really a
• 39:07
  big believer in Ethereum and its
• 39:08
  level of decentralization, it's
• 39:10
  really kind of Bye.
• 39:11
  Security is there,
• 39:13
  the speed of the chains are there,
• 39:17
  but privacy might be
• 39:18
  a missing piece of the puzzle.
• 39:20
  Do you think that that is true?
• 39:22
  I think privacy is something that is
• 39:24
  still a bit flexible.
• 39:26
  We have a great privacy standard.
• 39:28
  We were the first ones to launch
• 39:29
  this in main net confidentiality
• 39:30
  assets.
• 39:32
  Some others are trying to, I think,
• 39:33
  are following those footsteps now.
• 39:35
  It is a very good step forward
• 39:37
  for compliant privacy.
• 39:39
  I think that as we see these
• 39:41
  institutions coming on board, their
• 39:42
  requirements will be a little bit
• 39:43
  different. And we need to keep
• 39:44
  evolving the product towards those
• 39:46
  requirements going forward.
• 39:47
  It's a differentiator in the sense
• 39:49
  that you kind of have stable stakes
• 39:50
  for having it in the future.
• 39:52
  But it's not going to, I think, be a
• 39:53
  longer term. I mean, I think
• 39:54
  everyone would kind of adopt privacy
• 39:56
  infrastructure and common show
• 39:57
  infrastructure into their
• 39:58
  blockchains.
• 40:00
  That being said, the way you do it
• 40:01
  is nuanced in how it
• 40:03
  affects those other dimensions we
• 40:05
  talked about, security.
• 40:07
  Cost scalability is going to be
• 40:09
  really really matter because some of
• 40:10
  the techniques you use for privacy
• 40:11
  or coming necessarily are very
• 40:12
  expensive they might take seconds to
• 40:14
  complete or or they might cost you
• 40:17
  you know ten cents or a dollar to to
• 40:18
  kind of do that's not very scalable
• 40:20
  right that's not going to work out
• 40:22
  for a lot of transfers going forward
• 40:23
  it might work for a very small
• 40:24
  number of transfers but
• 40:26
  it's not gonna work it like at scale
• 40:27
  so that's that's I think something
• 40:29
  that we we think about a lot and
• 40:30
  going back to the AI side
• 40:32
  This is where I think
• 40:33
  confidentiality is a really big,
• 40:34
  this is a game changer because
• 40:37
  you know I think commentary AI
• 40:39
  inference will over time surpass
• 40:41
  what kind of non-commentary AI
• 40:43
  inference is today in a very
• 40:44
  meaningful way. It'll be a default
• 40:46
  for us the same way if you
• 40:48
  think about how SSL is and
• 40:50
  TLS, like every website you go to
• 40:52
  today is secured.
• 40:53
  That wasn't the case before that
• 40:54
  happened, right? But now it's just,
• 40:56
  if you go to one, Chrome will warn
• 40:57
  you, hey, this website is insecure.
• 41:00
  Do you really want to go here?
• 41:01
  Because it's so uncommon.
• 41:02
  And it's ridiculous in
• 41:04
  the age of information
• 41:06
  sharing, when the internet allows
• 41:08
  us to share the information freely.
• 41:11
  But now, securely, I think the same
• 41:12
  thing for AI. AI will allow us,
• 41:14
  we'll have infrastructure that we
• 41:15
  build that allows us share
• 41:16
  information securely
• 41:19
  between the kind of parts that we
• 41:20
  want to, but not just through the
• 41:21
  general web.
• 41:23
  Beautiful. Avery, thank you so much
• 41:25
  for being so giving of your
• 41:26
  time once again, it's always good to
• 41:28
  catch up with you.
• 41:29
  We'll watch with keen interest as
• 41:31
  the rest of the year plays out and
• 41:33
  how things progress at Aptos.
• 41:35
  But again, thank you so much for
• 41:36
  taking the time.
• 41:37
  Great to be here with you again.
• 41:39
  Thank you very much, Avery, and
• 41:41
  thank you for joining us for The
• 41:42
  Starting Block today.
• 41:43
  We will be back with our final show
• 41:45
  of the week.
• 41:47
  We've got Costas from Mr.
• 41:49
  Labs joining us tomorrow, one of the
• 41:50
  co-founders.
• 41:51
  We're going to have a very, very
• 41:52
  interesting conversation about
• 41:54
  privacy, security, all these things.
• 41:57
  So please set a reminder on your
• 41:58
  calendar tomorrow. We are
• 41:59
  broadcasting at normal time
• 42:00
  tomorrow. That's 8.30 a.m.
• 42:01
  Eastern time if you're in the U.S.,
• 42:03
  2.30 Central European time if you're
• 42:04
  here in Europe.
• 42:05
  Please do join us.
• 42:06
  And once again, if you've missed any
• 42:08
  of the great interviews or episodes.
• 42:09
  That we've had here on the starting
• 42:11
  block. You can find those on
• 42:13
  YouTube. If you go to the live
• 42:14
  section of our YouTube channel, you
• 42:16
  can find all the previous shows that
• 42:17
  we'd done, and then we have
• 42:19
  uploaded select interviews from
• 42:21
  the past two months of our
• 42:23
  broadcasts on our main YouTube
• 42:25
  channel as well.
• 42:26
  So go ahead and check those out.
• 42:27
  If you don't already follow us right
• 42:29
  here on X, if you joined us here
• 42:30
  today, like and retweet this
• 42:31
  broadcast so more people can catch
• 42:33
  the conversation we've just had with
• 42:35
  Avery.
• 42:36
  Until tomorrow, it's been great to
• 42:38
  have you with us. We'll see you
• 42:39
  again next time.