May Habib, Co-founder and CEO, Writer Sujay Jaswa, Co-founder and Managing Partner, WndrCo Jeffrey Katzenberg, Founding Partner, WndrCo, Andrew Nusca, Fortune
Category
🤖
TechTranscript
00:00Hey folks, thanks for being here with us.
00:03Okay, Mae, I want to start with you.
00:06I saw that your company is all about business writing.
00:10I'm a business writer.
00:11So, help?
00:15You gonna replace me?
00:16Well, that's where we got our start.
00:18We are the only generative AI company
00:21that is productized on top of every step function change
00:25in the capabilities of the transformer.
00:27We were founded four years ago.
00:29And that's what models were good at then.
00:31Now, LLMs and natural language
00:35really are the gateway to intelligence.
00:37And now we say, if you can write it,
00:39if you can describe it, you can build it.
00:41And so we have pioneered
00:42the enterprise generative AI category
00:45by combining LLMs with retrieval
00:49and guardrails and orchestration.
00:50And it's that full end-to-end solution
00:53that enterprises need.
00:54And that's what we've built over the last few years.
00:56So I'm safe, is what you're telling me.
00:59I don't know.
01:00I'm right here, May.
01:02All right, Sujay, I wanna come to you next.
01:06You've invested in this company.
01:08Tell us, remind us what WonderCo's thesis is
01:12and how Ryder fits it.
01:13Yeah, of course.
01:14All right, so first off,
01:15we were incredibly lucky to invest in May.
01:18She's crushed it since we invested.
01:20She raised it like 4X the valuation a year later.
01:23I mean, it's kind of wild what she's done.
01:24I'll talk about that in a second.
01:26What WonderCo does is we create a company every year,
01:29about one company a year.
01:31We make about five to seven venture investments a year,
01:34and we make 15 seed investments every year.
01:37The focus of the firm from the beginning,
01:39this is kind of like an unknown thing
01:40we were talking about earlier,
01:41is from the beginning,
01:43two-thirds of the money that we had went into software.
01:46At that point in time,
01:47one-third went into digital media.
01:49In 2020, we said, you know what?
01:51Let's just move entirely to software.
01:53And so since 2020, we've been 100% software,
01:55but we've been doing software since the get-go.
01:57Yeah, for sure.
01:59Jeffrey, you have your pick of what you want to invest in.
02:04Why AI stuff right now?
02:06What are the opportunities and risks
02:08in going in that direction for you?
02:11Well, I mean, I think that, again, as Sujay says,
02:14our focus has really been productivity tools.
02:19How do you take AI and create applications
02:25that make people more effective, more efficient,
02:28more productive, more creative at work?
02:33And there's certainly the big guys that are out there
02:40who are, I think, doing pretty world-changing things
02:45in terms of what's the future look like
02:48and sort of the larger future of AI.
02:51But what we've tried to do
02:52is just sort of be boots on the ground
02:54on a very practical basis.
02:55And you have companies like Rider today
02:58who can instantly deliver productivity
03:02and real ROI to companies of things
03:05that are essential to their day-to-day business
03:08and hers around storytelling,
03:11which is in particular why I think we were excited
03:14and why we related so much to Rider
03:19and to her mission from the outset.
03:21Yeah, yeah.
03:22May, I wanna press you a little bit on this
03:23may it may be I'm replaced or not thing.
03:26We've been talking a great deal over this event
03:29as to the role of humanity and creativity
03:33vis-a-vis the tools.
03:34So how do you look at that, especially given what you do?
03:38Where do we fit in, us humans?
03:40We're tools also.
03:41Humans are tools.
03:42And I know that sounds nuts,
03:44but we are moving to a world of autonomous action for AI
03:50where we're going to be nudged when we're needed.
03:54And across the street,
03:56we've got our customer conference going on
03:58where Uber, Salesforce, Victoria's Secret
04:02are walking folks through what is truly possible
04:05with a Gentic AI using our tooling.
04:08But the guardrails are so important.
04:10And folks are so obsessed with LLMs,
04:14and that's one piece.
04:16The part that folks really miss is,
04:20and this is part of our storytelling.
04:20This is now very meta,
04:21but this is really what's happening in the enterprise.
04:24To make AI work,
04:26you absolutely have to get your business process,
04:29your business logic, your data in order, right?
04:32And marry that with the LLMs
04:34whose reasoning capabilities really are increasing.
04:36We build our own LLMs.
04:37We are pushing the envelope on the architecture
04:40of the transformer to enable the kinds of complex
04:44orchestration of work that truly is possible.
04:48But it is no question so confusing
04:50for the enterprise right now.
04:52And it's just mass delusion
04:54that you can get to transformational generative AI
04:56working just with the hyperscalers.
04:58Yeah.
04:58Jeffrey, I want to come back to you.
05:00Same question essentially around where the humans fit in.
05:04I think about DreamWorks.
05:06You started it in 94.
05:08And would you start it the same way now
05:12given the preponderance of all these tools?
05:14No, but I've lived through these transformations before.
05:19I was born in the Jurassic era
05:21and we used to make animated movies
05:24by hand-drawn cell animation
05:27and went through that transformation
05:30into the resource and assets of technology
05:36as the assets to artists, to animators.
05:41And then obviously into what was the most transformative
05:45part, which came in the sort of mid 90s,
05:48which was CGI animation, which obviously has become,
05:53I don't know, four out of the top 10 movies this year.
05:56So I've seen this story before.
06:00It gets retold time and again.
06:03And each time there are these new sets of tools.
06:07It's a new paintbrush.
06:09It's a new, does anybody know what a typewriter is?
06:14I've heard of it.
06:16And so I actually think these are phenomenal tools.
06:20I think they are gonna be extraordinarily empowering.
06:24I said, I don't know now, it's a year ago
06:27where when I left the business now nine years ago,
06:33it took 500 artists five years
06:37and $200 million to make a movie.
06:40And that is already on this extraordinary curve down.
06:45It doesn't mean it's less creative
06:47and it doesn't mean that it doesn't require
06:50a great storyteller and a great vision.
06:53It does.
06:55It's just, we're giving a new set of tools to them
06:57to actually realize and make these dreams come true.
07:02Right, right.
07:03One thing to think about just to contextualize it
07:06because I think this feels scary
07:08because of how most of the people that we all interact with
07:12have the application of creativity as part of their job.
07:15But if you look back at like,
07:17if you just go back to the time of the American Revolution,
07:20the industry that 75% of American workers worked in
07:24was agriculture.
07:26And you look today and it's like two to 3%.
07:29And it's not because we're producing
07:31or consuming less food per capita.
07:32In fact, probably the opposite, right?
07:34It's because tools allowed us to kind of supercharge things
07:38and humans could go away from work that they would,
07:41you know, kind of the parts of the work that were drudgery
07:43and move to things where we could get
07:45to higher and higher orders.
07:46And I think likely it's gonna be the same thing
07:48with AI tools.
07:50Sushi, do you think that, I mean,
07:52you're looking at the whole landscape
07:54and trying to choose, make your bets, right?
07:57Do you think that we're going to lose entire industries
08:00or entire companies because of these dramatic changes?
08:03We've articulated it at the individual level now
08:05in this conversation.
08:06How do you look at it at a bigger scale?
08:08Look, the big debate is whether AGI is gonna happen
08:11in the near term or whether it's gonna take a long time.
08:13And, you know, there are people smarter than me
08:17that would tell you that AGI is around the corner.
08:19I think where we're betting as WonderCo
08:21is we think we're gonna see, you know, scaling limits
08:24to the development of the current generation
08:27of models and algorithms.
08:28And we think AGI is a long way away.
08:30And if we're right,
08:31then I think what you're gonna see is
08:34industries are just gonna get better
08:36because they have data, they can make things better,
08:38drudgery can go down, but all of us are gonna be fine.
08:41If we're wrong and AGI is around the corner,
08:43all bets are off.
08:44Got it.
08:45I'm gonna come to the audience for a question
08:46in just a minute, but I have one more for you, Jeffrey.
08:49You spent a great deal of the last year
08:51being the co-chair of the Biden campaign.
08:53I would love to know what you learned
08:56about how Silicon Valley and technologists
08:59are thinking about AI in that context.
09:01What did you learn campaigning for President Biden?
09:04Not that much.
09:05Next.
09:13Fair enough, to the audience then.
09:15They won, we lost, it's their bat, their ball.
09:17And I'm rooting for them to play great.
09:21Seriously, like, you know, this is the way
09:23I've been doing this for 50 years.
09:25And, you know, I've been on the winning team
09:28and I've been on the losing team.
09:30I used to be on the winning team,
09:31but it is their turn at bat.
09:34And honestly, for the future of this country
09:37and for the things that I think mean so much
09:40to many of us, I am rooting for their success
09:46in every single respect.
09:48I hope the changes that they bring
09:52make us a better, stronger, more successful country.
09:55Right on.
09:59All right, a question from the audience.
10:03Yes, right over here.
10:04Hi, Steven.
10:06Steven Wolf Boneta from Alpha & Co.
10:08May, huge congrats, love everything
10:10that you've been doing with Rider.
10:12Your issue about data governance is so important
10:15and so many of these companies are just not AI ready
10:17because of the data issue.
10:18Where do you see some of the challenges?
10:20Is it more on the CTO technical side
10:22where they're kind of not made here
10:24and they're kind of being protective?
10:26Or is it on the business side where maybe it's the CMO,
10:28but they really don't understand everything
10:30that you're talking about?
10:31Yeah, so maybe two challenges,
10:33both on the business side and on the tech side.
10:35So on the business side,
10:38it is a real interesting challenge.
10:42If you think about the Industrial Revolution,
10:44all those technologies existed,
10:46but we didn't get that big, big change in productivity
10:50till we got the assembly line.
10:52And so much of what is required
10:56is this assembly, this guard railing of the work.
10:59Give you an example, large Fortune 10 asset manager.
11:03How do you do investor reporting?
11:06You've got five mil on a credit fund
11:08and you've got 10 mil on a private equity fund.
11:11Literally, how do we do reporting?
11:12Nobody knows.
11:14Someone outsources it to a SI for the tune
11:17of 200 mil a year, but there's no process.
11:19And to get generative AI against that process,
11:22the data needs to follow the business logic.
11:26And so the human effort to actually go in
11:30and diagram out all of this stuff, that's data too.
11:34The examples of how we've done it before,
11:36how we don't like to do it, how we wanna do it, that's data.
11:39And how you make that into continuous pipelines,
11:42that's a technology problem right now.
11:44And so folks paint a very wide brush
11:47of data is challenging for generative AI,
11:49but let's get specific and it's all very use case driven.
11:53Yeah, that's well said.
11:54Thank you, May.
11:55Well, let's go for one more, please.
11:57Yes, way over here.
11:59Hey there.
12:00Good to see you.
12:05Hi, Laurie Yoler from Playground.
12:07A number of, I know you've been talking
12:09about enterprise productivity and I think we agree
12:12that is wonderful with the appropriate guard rails,
12:15but a number of artists and writers
12:17who are writing nonfiction and more creative works
12:21have had successful pieces and then have been approached
12:24by AI companies to write the sequels
12:27and use AI to generate those based on the characters
12:31and the initial work.
12:33And they have been told by other friends who are creatives
12:36that that would be terrible, it would dilute their brand
12:39and using AI for that would make it look
12:42like they didn't have that creativity going forward.
12:44What do you advise folks who are being given offers
12:48to develop AI co-generated content?
12:52I'm actually not, I'm not sure, so I don't wanna,
12:56but what I will say is that I have many friends of mine
13:04are among the leading writers, creators,
13:08and showrunners in Hollywood today.
13:11And I would say almost across the board,
13:17they have all talked about how the AI tools today
13:23have been helpful to them.
13:25They have seen them as a resource and as an asset,
13:30they have made them more productive,
13:34they've been able to widen the,
13:41diversity of their work, the quality of their work.
13:44So again, I'm only telling you what I hear from,
13:50I would say among the most talented showrunner creators
13:53today, how they find these tools an amazing resource
13:59for them and it's not constraining them,
14:01it's inspiring them.
14:03Can I build on that, Jeffrey?
14:04Go for it.
14:05We've got-
14:06I'm sorry to interrupt you,
14:07but we're about to wrap it, yes.
14:08The answer for Jeffrey is no.
14:11Forgive me, May.
14:13Please, a round of applause.
14:14Thank you for joining us, folks.