Vai al lettorePassa al contenuto principale
  • 42 minuti fa

Categoria

📺
TV
Trascrizione
00:00always
00:01meantime, AI voice startup
00:0311 Labs talking about AI
00:04announcing it has surpassed 500 million dollars
00:07now in annual reoccurring revenue
00:09the company also bringing on some new investors
00:11including BlackRock, Nvidia and celebrities
00:13like Jamie Foxx and Eva Longoria
00:15the milestones coming as artists like Taylor Swift
00:18move to protect their voices
00:19through trademarks and
00:21growing concerns over AI deepfakes
00:23joining us right now is 11 Labs CEO
00:25Matty, I always mispronounce
00:27your last name so I'm not even going to try
00:29because I know I'm going to screw it up and then you're going to laugh at me
00:32it's great to see you though
00:35help us understand
00:36where this
00:38where we really are in this
00:39technology journey
00:41obviously from a financial perspective
00:43you guys have been killing it
00:45the question I have is sort of where do you think all of this goes
00:48and for those who are uninitiated
00:50this is a model that could take
00:52even the conversation we have right now
00:54and effectively generate
00:56you know extraordinarily
00:58voice that's almost identical
01:00if not more so
01:03Andrew, Melissa
01:04great to be here to talk about what's happening at 11 Labs
01:07and across our customers
01:09so maybe as a quick introduction
01:10quick context
01:11we launched our first product
01:12in 2023
01:13and now three years later
01:15just crossed
01:17500 million ARR milestone
01:19and are bringing an incredible set of investors on board
01:22to help us in that next phase of growth
01:24and as a company
01:25what we do
01:26we combine research, product, and deployment
01:28to help businesses
01:29transform how they communicate
01:31and that really
01:32started with building
01:34the transformational AI voice
01:36and audio models
01:37and now delivering them
01:38across the platform
01:40that helps power
01:41entirety of interactions
01:42across the company
01:43so whether that's sales
01:44support, business operations
01:46with AI agents
01:47whether that's marketing
01:49and media
01:49with creative tooling
01:50and maybe some of the examples
01:52of the work we do
01:53will be helping Deutsche Telekom
01:54Klarna
01:55better
01:55deploy those conversational agents
01:58bringing the voice
01:59bringing the knowledge
02:00of that organization
02:01bringing the context
02:01of the user
02:02to really elevate
02:04how they can serve
02:05their customers
02:06all the way to NVIDIA
02:07and RAM
02:08helping
02:09across the creative work
02:10and to answer your question
02:12Andrew
02:12I think we are just
02:13at the beginning
02:13of the adoption
02:14of how the businesses
02:15communicate across
02:17Maddy
02:17explain this though
02:18to me
02:19just so I understand
02:19you know we talk
02:20about your company
02:22in the context of voice
02:23and you guys have
02:25some extraordinary
02:25voice models
02:26and then we often
02:28talk about
02:28enterprise companies
02:29that are trying
02:30to support
02:32telephone agents
02:33and other things
02:34I'm thinking of
02:35Brett Taylor's company
02:36for example
02:36and others
02:38how do you
02:40you guys are increasingly
02:41doing a lot of
02:42enterprise work
02:43so where do we think
02:44of you relative
02:45to some of these
02:45other players
02:48you know I think
02:50the biggest piece
02:51that we see
02:51is AI
02:52becomes smarter
02:53and one
02:55leader piece
02:56is that communication
02:57becomes the bottleneck
02:58we can build
02:59intelligent system
03:00but if your AI
03:01sounds robotic
03:02if it interacts
03:03strangely
03:04it just won't
03:04be trusted
03:05or widely adopted
03:07so as a company
03:08we take the two
03:08together
03:09we solve the research
03:10as one of the
03:11few companies
03:12in the space
03:12building those
03:13foundational models
03:14for interaction
03:15how you make it
03:15seamless
03:16how you make it
03:16sound great
03:17alongside
03:18exactly what you
03:20mentioned
03:20how do you bring
03:21the knowledge
03:21of the organization
03:22how you bring
03:23the context
03:23of the user
03:25how you bring
03:25the deployment
03:26capacity
03:26to evaluate
03:27monitor
03:27the agents
03:28in the production
03:29I think the biggest
03:30difference that we
03:31took as a company
03:31is that we built
03:32a configurable
03:34open platform
03:34that can be used
03:36in support
03:36of some of the
03:37other companies
03:38but can also be
03:39used in sales
03:40in proactive
03:41outbound
03:41in business operation
03:42in training
03:43and hiring
03:44so maybe a good
03:45example of that
03:46is Deutsche Telekom
03:47Deutsche Telekom
03:47started our work
03:48by helping
03:50on the marketing
03:51side
03:51so creating
03:52better materials
03:53for their audience
03:54that then extended
03:55to customer support
03:56where everybody
03:57can reach the
03:58T-Mobile network
03:59and have conversational
04:01agents that serve
04:01that better
04:02all the way to now
04:03deploying the first
04:05of its kind
04:05agent inside
04:06of the network
04:07so when you are
04:07calling through phone
04:09as a T-Mobile
04:10subscriber
04:11you can ask an agent
04:12to help you
04:14get additional
04:15information
04:15all the way
04:16to real-time
04:17translate your
04:18conversation
04:18to another language
04:20and that's what
04:21we are seeing
04:21across the last
04:226-12 months
04:23of growth
04:23which is that
04:24enterprise adoption
04:25across those
04:26domains
04:26Mani what do you
04:28see as the
04:29moat around
04:30your business
04:31the reason I ask
04:32is it seems like
04:33almost every day
04:34the next generation
04:36frontier model
04:38is able
04:39and capable
04:40of doing things
04:40we never thought
04:41was possible
04:42and effectively
04:43entering new
04:44spaces
04:46yeah the
04:47way we look
04:48at it
04:48it's a combination
04:49of research
04:51product
04:51and ecosystem
04:52we create
04:52across
04:53so research
04:54my co-founder
04:55my best friend
04:55is one of the
04:56leading researchers
04:57in the space
04:57of course
04:58is able to
04:59create models
05:00that lead on
05:01benchmarks
05:01against the
05:02biggest AI labs
05:03out there
05:03whether that
05:04started with
05:04text-to-speech
05:05speech-to-text
05:06or that interactive
05:07ability for agents
05:08and we think
05:09that will compound
05:10over time
05:11as you refine
05:12the models
05:13for specific
05:13industries
05:14like healthcare
05:14financial services
05:15with combination
05:16of data
05:17with combination
05:18of how to make
05:18them reliable
05:19in those settings
05:20the product
05:21here you need
05:22to understand
05:23the workflow
05:23you need to
05:24understand the
05:24problems of the
05:24customers
05:25and to your
05:26earlier question
05:26you really need
05:27to go deeper
05:28into each
05:29of the domains
05:30to build
05:30the platform
05:31that's very
05:32subtle
05:33and the last
05:34piece is the
05:35ecosystem
05:35and that's
05:35a little bit
05:36to why our
05:37investor represents
05:39Saturday Verset
05:40where we are
05:41not only creating
05:42the product
05:42but bringing
05:43that in
05:44combination
05:44between tech
05:45and creative
05:46space
05:47by creating
05:47an ecosystem
05:48where people
05:48can contribute
05:49voices
05:49contribute
05:50templates
05:50for agents
05:51well that's
05:51where I wanted
05:52to go with
05:52this
05:52let's just
05:53talk about
05:53the IP
05:53behind these
05:54voices
05:54because the
05:55truth is
05:56people like
05:56Taylor Swift
05:57and others
05:57have brought
05:58lawsuits
05:58everyone's
05:59very worried
05:59about their
06:01IP
06:02I should be
06:03worried about
06:03my own voice
06:04and now I know
06:04you're getting
06:05into business
06:05with people
06:06like Jamie Foxx
06:06and Eva Liguria
06:08so how is this
06:09going to work?
06:11As a company
06:12we took a
06:13very specific
06:14approach
06:14where we want
06:15to work
06:16with the industry
06:17to disrupt
06:18some of the
06:18process
06:19and learn
06:19how the
06:20storytelling
06:20of the future
06:21can be made
06:22that started
06:23all the way
06:24when we launched
06:24the first product
06:25we created
06:25the first
06:26effectively voice
06:27marketplace
06:28where anyone
06:29can create
06:29their voice
06:30we authenticate
06:31it
06:31then you can
06:32share it
06:32and while the
06:33voice is being
06:33used you can
06:34earn possibly
06:35compensation as
06:35a result
06:36today we have
06:3720,000 voices
06:38on the platform
06:39that people can
06:40work and expanded
06:41this model to
06:41music and
06:42recently to
06:43agent templates
06:44where you can
06:44pre-configure
06:45agents so people
06:45can build on
06:46top of
06:46as well
06:48and that's
06:49where we are
06:49bringing a lot
06:50of investors
06:50on board
06:51to be able
06:51to learn
06:52from them
06:52of how they
06:53see the future
06:54of storytelling
06:54how they see
06:55the future
06:55of media
06:55and bring
06:56those learnings
06:57into the platform
06:57maybe a great
06:58example of some
06:59of that work
07:00has been
07:00recent with
07:01Matthew McConaughey
07:02where Matthew
07:03creates his newsletter
07:05that he narrates
07:06in English
07:06but for the first
07:07time ever
07:08we brought that
07:08over to Spanish
07:10to Portuguese
07:11where you can
07:11enjoy with his
07:13incredible voice
07:14intonation and emotion
07:15that same newsletter
07:17with that new
07:18language set
07:19and that's something
07:19we will see
07:20increasingly
07:20across the companies
07:23we work with
07:23do you think
07:24long term
07:26that this means
07:27that the incumbents
07:28that this is going
07:29to benefit
07:30the incumbents
07:31meaning
07:32somebody who's
07:33famous
07:34currently famous
07:35will have a much
07:36better shot
07:37of people
07:37who are
07:38going to
07:39come up
07:40in the future
07:40because it's
07:41going to make it
07:41harder for people
07:42to come up
07:42in the future
07:43because the incumbents
07:44are going to have
07:44a bigger share
07:48there are two
07:49patterns that are
07:50happening
07:50of course
07:50one is
07:51with the
07:53democratization
07:53of creative
07:54tools
07:54everyone can get
07:56started
07:56without the
07:58same access
07:58of expressive
07:59equipment
08:00that was needed
08:01in the past
08:02so I think
08:02that's the first
08:03pattern
08:03where you will see
08:04increasing amount
08:05of people
08:05just being able
08:06to break through
08:06that might have
08:07not had the
08:08resources before
08:09together
08:10now of course
08:11you actually
08:11now need the
08:12talent
08:12you need the
08:13story
08:13you need the
08:14emotions
08:14that you can
08:14expose
08:15and that will be
08:16the differentiating
08:17part
08:17there will be
08:18a second
08:18pattern
08:18to your
08:19question
08:20where people
08:21that already
08:22build the
08:22audience
08:23will be able
08:24to expand
08:25their work
08:25in just
08:26completely new
08:26ways
08:27whether that's
08:27breaking down
08:28language barrier
08:28and speaking
08:29other languages
08:29whether that's
08:30shifting from
08:31static to
08:31interactive content
08:32where in the
08:33past that
08:34delivery just
08:35happened once
08:35now you'll be
08:36able to interact
08:37with those
08:37characters in
08:38a completely
08:38new settings
08:39whether that's
08:40in the gaming
08:40environment
08:41whether that's
08:42through a fun
08:42engagement campaign
08:44all the way
08:45through to new
08:46educational
08:46capacities
08:47that were
08:47impossible
08:48before
08:48recently we had
08:50a pleasure
08:50of working
08:50with master
08:51class
08:51where in
08:52master class
08:52previously you
08:53had a static
08:54content delivery
08:55for educational
08:55lessons
08:56now you can
08:57learn cooking
08:57by speaking
08:58with Gordon
08:59Ramsey
08:59all the way
09:00through to
09:00learning
09:00negotiation
09:01by negotiating
09:02life with
09:03Chris Voss
09:03and I think
09:04we'll see
09:04increasing
09:05amounts of
09:05that
09:06okay final
09:07question
09:07it's a security
09:08question
09:09do you and
09:09your family
09:10now have
09:10what they call
09:11a safe
09:11word
09:12because of
09:14these kind
09:14of models
09:15everybody's so
09:15scared
09:16these models
09:16are so good
09:17that you know
09:18you could have
09:18your you know
09:20your your
09:20your brother
09:21or sister
09:21or wife
09:22or kids
09:23call you up
09:23on the phone
09:24or you think
09:24they're calling
09:25on the phone
09:25because the voice
09:26is so good
09:26and you need
09:27to have a safe
09:28word that they
09:28don't otherwise
09:29know
09:32look it's a
09:33it's a great
09:33question I think
09:33in general
09:34we think you will
09:35see increasing
09:36amounts of AI
09:36content AI
09:37interaction AI
09:38agents in the
09:38future all of
09:39us will have
09:39an AI agent
09:40that interacts
09:40on our behalf
09:41but then she
09:42can call us
09:43so yes
09:43we'll need
09:44a new safety
09:45systems in
09:46place
09:46as a company
09:47we do
09:49provide a lot
09:50of new safeguards
09:51new abilities
09:52of how you can
09:52identify and
09:53authenticate that
09:54across whether
09:55that's provenance
09:56and traceability
09:57whether that's
09:57moderation
09:58but also
09:59collaborating with
10:00the industry
10:00on new
10:01mechanisms
10:01like synth
10:02ID or C2PA
10:03to your pointed
10:04question
10:05we do
10:06we do have
10:07a set of
10:07mechanisms
10:08with my
10:09close friends
10:10to make sure
10:10that as I'm
10:12a public
10:13increasingly a
10:13public figure
10:14they know
10:15how to reach
10:16me
10:17or when I'm
10:18reaching out
10:18in terms of
10:19need
10:20and we are
10:21speaking with
10:21governments
10:21of how to
10:22make sure
10:22that the
10:23same mechanism
10:23exists for
10:24the white space
10:25and Matty
10:26thank you
10:26congratulations
10:27look forward
10:28to talking
10:28again soon
10:29and talking
10:30to the real
10:30version too
10:31I enjoy
10:32talking to
10:32the real
10:32version of
10:33you
Commenti