Jan 15, 2025 · 1h 13m · news
Victor Riparbelli, CEO @Synthesia: OpenAI vs Anthropic vs X.ai - Who Wins and Why | E1246 · 20VC with Harry Stebbings
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of 20VC, Synthesia Co-Founder and CEO Victor Riparbelli discusses raising a $100 million Series D, the structural shift from text to video communication, and why owning the end-to-end enterprise workflow is the ultimate defensive moat in the AI hype cycle.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Harry holds 16.6% of the talking time here. How this is scored →
Harry as informed peer 3.7 Guest teaching 3.5 Guest disagreement 1.7 Harry pushing back 2.9
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 30:54 Victor dismisses OpenAI threat as classical VC brain
Victor directly rejects Harry's premise about OpenAI swallowing application startups, calling the investor narrative 'classical VC brain' and forcefully explaining why workflow moats beat raw foundation models.
Hardest push from Harry ▶ 40:37 Harry refutes AI video democratization using his internal team benchmarkHarry explicitly challenges Victor's claim about AI content democratization by citing his own daily media production experience and detailing how AI clipping tools fall far short of human editor judgment.
Biggest teaching moment ▶ 30:54 Victor educates host on enterprise video workflow moats versus model technologyVictor corrects Harry's assumption that Synthesia is vulnerable to model providers by walking through the entire enterprise value chain—from script editing to localized video players—showing why avatars are merely a feature.
Harry holds his own ▶ 40:37 Harry leverages hands-on production expertise to push back on AI timelinesHarry brings concrete domain expertise as a major podcast producer to show Victor why ambiguous editorial decisions in media creation resist automated AI shortcuts.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Opening Pleasantries and Series D Announcement | 2 | 3 | 1 | 1 | Harry congratulates Victor on Synthesia's $100M Series D round at a $2.1B valuation and asks about the company's early funding history. Victor explains how European venture capitalists in 2017 possessed a private equity mindset that couldn't grasp AI content generation, leading to 80-90 rejections before Mark Cuban invested. | |
| The Truth About Product-Market Fit and Early Capital | 4 | 3 | 2 | 3 | Harry asks Victor how he views ultra-high valuation seed rounds for pedigreed founders. Victor argues that raising too much money before product-market fit breeds lack of focus, whereas Harry pushes back by framing PMF as a moving, multi-chapter goal across customer tiers. | |
| The Future of Communication: Video vs. Text | 3 | 4 | 1 | 1 | Victor outlines his vision that text is an inefficient information compression method and predicts video and audio will replace text as the default web communication medium. Harry agrees, sharing his personal habit of analyzing TikTok content transitions every night. | |
| Capital as a Weapon and Navigating Competitors | 4 | 3 | 2 | 4 | Harry asks if raising large capital rounds is meant to signal competitors to stay away and asks who directly copied Synthesia's mission statement. Victor emphasizes that capital is for product development rather than competitor posturing, while acknowledging learning from market feedback. | |
| The Impending Enterprise Churn and the AI Hype Cycle | 4 | 4 | 2 | 3 | Harry asks where we are in the AI hype cycle as enterprise buyers demand clear ROI. Victor explains that enterprise buyers often lack clarity, leading AI startups to optimize for initial contract closing over true renewal value, which will trigger massive churn. | |
| High Expectations and Developer Productivity | 4 | 3 | 2 | 3 | Harry quotes Mark Zuckerberg's statement about AI replacing mid-level software engineers by mid-year. Victor reframes this as typical tech industry hyperbole, explaining that building complex software involves human feedback loops that AI cannot easily automate. | |
| Commoditization and Betting on OpenAI vs. Anthropic vs. X.ai | 4 | 3 | 2 | 3 | Harry asks Victor to choose where he would invest $10 million among OpenAI, Anthropic, or X.ai. Victor chooses X.ai due to asymmetric upside, Elon Musk's track record, direct access to real-time X data, and native distribution. | |
| Differentiators of Success: Compute, Control, and Content Moderation | 5 | 4 | 2 | 5 | Victor explains that control layers and granular editing are harder to build into AI models than slot-machine generation. Harry directly asks if Synthesia acts as the arbiter of justice for acceptable content, and Victor confirms they take a strict enterprise moderation stance. | |
| Workflow over Models: Defensive Moats in AI Application | 4 | 6 | 5 | 4 | Harry relays an investor concern that OpenAI will inevitably move into Synthesia's territory and render them obsolete. Victor forcefully rebuts this as classical VC over-focus on raw model technology, explaining that Synthesia's true moat is end-to-end enterprise video workflow. | |
| Expectations for GPT-5 and the Definition of AGI | 3 | 4 | 1 | 2 | Harry asks about the delays surrounding GPT-5 and whether it will live up to expectations. Victor notes that human perception saturates quickly and argues that current models are already sufficient for solving most business problems. | |
| The Ultimate Democratization of Content Creation | 3 | 4 | 1 | 2 | Victor outlines the democratization of content creation, comparing generative AI video to how digital instruments transformed the music industry. He predicts the technical cost of creating a Hollywood film will drop to zero. | |
| AI Limitations and the Premium on Human Storytelling | 6 | 3 | 3 | 6 | Harry strongly challenges Victor's timeline for AI content creation by pointing out how current AI clipping tools remain vastly inferior to his in-house production team. Victor concedes that technical removal of barriers places a higher premium than ever on human storytelling and judgment. | |
| The Debate on Banning TikTok and Social Media Diets | 5 | 2 | 1 | 3 | Harry asks if TikTok should be banned, and Victor defends the platform. Harry presents a framework comparing content consumption to a personal food diet, arguing individuals bear responsibility for curating positive algorithms. | |
| Professional Creators Over Traditional Journalism | 4 | 4 | 2 | 4 | Harry claims domain practitioner content is vastly superior to traditional journalism. Victor discusses content verification systems akin to Shazam for digital provenance, and Harry questions if flagging synthetic origins inherently devalues AI content. | |
| AI's Impact on the Content Labor Market | 3 | 4 | 1 | 2 | Harry asks how AI will reshape creative labor markets and which specific job roles are threatened. Victor argues that technical tasks like camera operation will decline while creative storytelling skills will expand rapidly. | |
| Building Synthesia in London vs. Silicon Valley | 3 | 4 | 1 | 2 | Victor reveals he founded Synthesia in London primarily because he could not obtain a US visa at the time. He contrasts European employee loyalty with the transactional equity-hopping mindset prevalent in Silicon Valley. | |
| Work Ethic: Silicon Valley Engineers vs. Europe | 4 | 3 | 2 | 4 | Harry asks if Silicon Valley engineers work harder than European developers. Victor rejects this stereotype, asserting that extreme work ethic is an individual trait rather than a geographic one, while Harry pushes on recent decline in UK building sentiment. | |
| Advice for Keir Starmer and the LSE Liquidity Issue | 5 | 3 | 2 | 4 | Harry asks Victor what advice he would give UK Prime Minister Keir Starmer to retain tech founders. Victor advises against over-taxation and regulation, while Harry sharply criticizes the government's persistent negative messaging. | |
| Does IPO Location Matter and Avoiding European Over-Regulation | 4 | 3 | 1 | 3 | Harry discusses the London Stock Exchange's liquidity issues and asks if IPO destination actually matters to tech founders. Victor argues that liquidity matters wherever obtained and transitions into quick-fire topics on gaming as decision simulation. | |
| Quick Fire: The Secret Superpower of a Generalist | 2 | 3 | 1 | 1 | In the quick-fire segment, Victor reflects on his superpower as a generalist who synthesizes diverse, random knowledge domains into effective executive decision-making. | |
| Quick Fire: Mark Cuban and Gaming Advice for Future Kids | 2 | 3 | 1 | 1 | Victor single-out Mark Cuban as the investor he would choose again for taking a chance on their seed round, and strongly advises startup founders to take early secondary liquidity to relieve personal stress. |
Statements from this episode (55)
Riparbelli: AI is in a bubble and capital will burn
“I mean, I think we're definitely in a bubble, right? There's a lot of money that's gonna go up in flames in AI products.”
Riparbelli: Contract renewals are the true signal in enterprise sales
“The real signal is not that you sign a contract. The real signal is renewal.”
Riparbelli: Synthesia raised $100M Series D led by NEA
“As we just raised a series D which is super exciting, big milestone for us. We raised a hundred million dollars led by NEA with participation from all of our existing investors.”
Riparbelli: By 2027, anyone will create Hollywood films from a laptop
“In 10 years, you're gonna be able to make a Hollywood film from your laptop, needing nothing else than your imagination.”
Riparbelli: European VC is dominated by private equity minds, not technologists
“I think in Europe, the VC industry is still dominated by people who used to work in private equity, and that's a very different mindset than technologists, right?”
Riparbelli: Mark Cuban led Synthesia's $1M seed at $5M post-money
“The first round was a million dollars... And we ended up doing it at the five million post with Mark Cuban”
Riparbelli: Capital constraints force founder discipline that large funding rounds ruin
“Operating on the constraints that we had at the time really just focused us on our customers and selling the product... Working under constraints... Forces a lot of discipline that's easily gets lost the more money you have”
Riparbelli: Larger Teams Slow Down Product-Market Fit Discovery
“Having a team of 20 people instead of five people trying to learn that, I think actually slows you down.”
Riparbelli: Founders Shouldn't Hire Sales or PMs Pre-PMF
“Unless you have product market fit, like you shouldn't have product managers and sales people, right? That is your job as a founder.”
Riparbelli: Bootstrapping a Category-Defining Enterprise Company Is a Myth
“And you're not gonna win that by just bootstrapping all the way, right? I think that's a myth.”
Riparbelli: Our kids may be the last generation to read and write
“I actually do think that maybe not us, but maybe our kids, kids are going to be one of the last generations they'll read and write as like the default web communication. I think we'll increasingly just consume everything via video and audio.”
Riparbelli: Synthesia kept its rapid growth secret through Series C
“One of the interesting things about Synthesia is that when we raised our A to B to C round, It was very much a secret how fast we were growing and how much people love the product.”
Riparbelli: A rival AI video company verbatim copied Synthesia's messaging
“We have one particular competitor that, I mean, literally verbatim copied like our mission statement, like everything I say, they like just copy everything that we do.”
Riparbelli: Competitors' marketing spend accidentally reaccelerated Synthesia's growth
“We had a massive reacceleration kind of the second half of the year and a big part of that is actually because our competitors did a whole bunch of marketing for the category. We clearly had the superior product. And so that actually helped us quite, quite sig…”
Riparbelli: Framing software as 'AI employees' is dumb and misleading
“I really, really hate it when people call, like, AI employees. I think it's, like, I think it's so dumb. I think it's not helpful to build useful technologies that people want to adopt. And I just think it's the wrong way of thinking about that. You're going t…”
Riparbelli: AI is nowhere near ready to fully automate complex software
“But if you look at most software out there today, it's very complex and it's not just technology, right? It's a lot of humans, it's customers, it's feedback loops. And I don't think we're anywhere near being able to automate those things.”
Riparbelli: Software engineers will be significantly more productive in 6 to 18 months
“But I definitely think that a great software engineer will be way more productive in the next six, 12 to 18 months.”
Riparbelli: Claiming AI prompt tools will easily kill software incumbents is naive
“And I think that's like a really, really naive view of what it's like to build a business. Technology and lines of code is super important, but as you know, right, you've built your company yourself. It's a, it goes way deeper than that. And I think it really,…”
Riparbelli: Current LLM text generation technology is already good enough for most uses
“I do think we are seeing the commoditization of the text generation layer, right? For most of the use cases that We'll see LLMs transform the world as we know it today. I think the current generation technologies are good enough. It's about, of course, improvi…”
Riparbelli: xAI has more asymmetric upside than OpenAI or Anthropic
“I actually think I would do X. I think they're all great companies. I would do X because I think there, there's the most asymmetric upside if Elon delivers what he usually does. And I mean, we've all heard of some of the, like, that data set that he built, lik…”
Riparbelli: xAI's ownership of X provides unique data and distribution moats
“I think the fact that he owns X is really, really powerful. OpenAI clearly managed to capture, like, the consumer as in, like, the destination that you go to use an LLM. I think we'll see LLMs being a part of, like, many different apps, and I think owning X al…”
Riparbelli: Compute scale alone will not determine AI winners
“I think they will continue, but I don't think it's just going to be linear. Like whoever has the most compute is going to win.”
Riparbelli: AI algorithms will achieve 10x to 100x efficiency gains
“Someone will come up with an algorithm that can, that is like 10, a hundred times as efficient as what it is today.”
Riparbelli: Human content moderation is not the right way forward
“I don't think that like humans moderating content is the right way forward.”
Riparbelli: 99.9% of Synthesia user content is completely uncontroversial
“And it's been a journey for us because the hard thing is you have the, what we call the green content, the content that everyone agrees is great. That's 99.9% of the content.”
Riparbelli: Synthesia has no obligation to uphold free speech for $30 monthly
“We don't, as a company have, we don't see ourselves as having to uphold any kind of right of free speech. And frankly, from a business perspective, having someone pay me 30 dollars a month to create very questionable conspiracy content, it's just not good busi…”
Riparbelli: Winning AI startups in coming years will prioritize workflow over models
“I think that's something we're going to see the next couple of years is that the winning companies are going to be the ones that are not overly obsessed about the AI component of their platforms, but who build for workflow.”
Riparbelli: Synthesia is shifting some workloads from OpenAI to Anthropic and Gemini
“We're most using OpenAI... We've shifted some workloads to Anthropic. We're also using Gemini actually.”
Riparbelli: OpenAI's next GPT release won't match ChatGPT's cultural impact
“I think it'll come out and it'll be like, good. It'll be definitely better than the previous things. But I think most people will not like care that much because the bar for someone to really care that much is incredibly high, right?”
Riparbelli: AI tools will allow YouTube creators to compete directly with Hollywood
“It means that people like yourself, YouTube creators, everyone else can compete with Hollywood in terms of what they can bring to life, right?”
Stebbings: AI clipping tools will not match human editing in two years
“Like, when I look at the clipping tools, They are so far off what my team is able to do upstairs. It's incomparable. I don't see that being two years out.”
Riparbelli: Human-created content will command a higher premium as AI content grows
“The value of like real content is, is going to be higher than ever because we have a lot of air general content, and I think there'll be a genre of itself, but just like we still like to watch people play piano or know that someone that the artists we're watch…”
Riparbelli: AI-Made Content Will Outpace Human Content in Five Years
“Definitely AI made content for sure.”
Riparbelli: TikTok Is a Great Product With Mostly Educational Content
“I think TikTok is actually a really great product, potentially like a bit of an unpopular opinion, but I think what TikTok has done, when I look at my TikTok feed today, I would say it's 90% highly informative, educational, really great content.”
Riparbelli: Future media ecosystem will be driven by citizens and experts, not journalists
“We're kind of starting to find our way into like, how do we build a really awesome media ecosystem that is driven by citizens and experts rather than journalists and politicians.”
Riparbelli: A two-minute TikTok conveys information better than a book
“A two minute long TikTok video is a better way of understanding a topic Then a 20 minute long YouTube video, which is a better way of understanding a topic than a few hundred page book.”
Riparbelli: Most online content and Instagram photos are already partially fake
“We're already in a world where, like, most content that we consume today is in some, to some extent, fake, right? Like, most pictures on Instagram have been, like, touched up in some way, shape, or form.”
Riparbelli: AI video will be indistinguishable from real video within years
“You could make, you know, a video of you talking about something in a couple of years, no one will be able to tell if it's actually real or not, but it won't really matter that much. What matters is that you made it right and that whatever you say in that vide…”
Riparbelli: Technical content creation roles like camera operation will fade away
“I think the, you know, like very technical roles, like camera operation, those types of things, they'll probably like fade away to some extent.”
Riparbelli: Human trust and integration, not technology, are AI's main bottleneck
“I think most of the rate limiting and how quickly AI becomes a part of our daily life is actually not the technology. As I said earlier, I think we have very, very powerful LLM, so they can do a lot of like great things, look at better and better, but the rate…”
Riparbelli: West Coast startups still have higher chances of success
“I still do think that The chance of success is higher if you're in the US on the west coast. I think there's a lot of pros, especially like the last six, seven years got a lot better to build here, but do still think that there is a lot of advantages to be in …”
Riparbelli: European startup employees still view stock options as lottery tickets
“In Europe, I think people, again, for good and for better and for worse, right? I think people are, they think less like that, right? As you of course know, like stock options in Europe, people still view it as like, A random lottery ticket, not really as part…”
Riparbelli: Average Silicon Valley engineers do not work harder than European engineers
“Honestly, I don't really have a strong sense that people work way harder in the Valley. I actually don't like, I know a couple of great friends who've moved over there and opened offices. And I have not heard from any of them that like, wow, people just work s…”
Riparbelli: London is producing global AI leaders, not regional winners
“What we're seeing in the London ecosystem is that we're actually building global leaders, not just, like, regional winners of, like, expense management or something like that. Like, you have 11 Labs who are leading in voice. You have us, which is leading in en…”
Riparbelli: Denmark would have been an impossible place to build Synthesia
“I come from Denmark, right? That would have been a terrible place to build a company. That would have been impossible, basically.”
Riparbelli: Moving Silicon Valley talent to London is easier than New York
“Like I've, in our journey, I've seen that it's easier to move people from the Valley to London than from the Valley to New York”
Riparbelli: One or two $10B+ UK tech companies will trigger ecosystem growth
“If we don't screw it up, we just need one or two companies to become plus ten billion dollar companies headquartered in the UK, then I think we'll really start to see the ecosystem grow.”
Riparbelli: London Stock Exchange liquidity shortage is a disaster
“The London Stock Exchange, like, they need to fix that, like, they need to put some liquidity in the system. I don't know, I'm not a financier, so I don't know what they need to do to fix it, but I mean, that is a disaster, right?”
Riparbelli: Europe has all the AI regulation and none of the companies
“Europe has all the AI regulation and none of the AI companies.”
Riparbelli: Playing computer games is the best preparation for future success
“I think that playing computer games is probably one of the best things a kid can do for like future success in their life.”
Riparbelli: Diverse Inputs Improve Subconscious Pattern Recognition and Decision-Making
“I think, I really truly believe that those companies are subconscious, And the more food and diverse things you put into your brain, the better you get at like making analogies and pattern matching and those kind of things.”
Riparbelli: Studying business at university is mostly useless
“I think studying business controversial opinion is like mostly useless. I think you have to do that by learning and you can, there's basically like 10 ideas in business. And if you learn those, you'll be pretty well covered.”
Riparbelli: Mark Cuban Is the One Investor He'd Choose for Next Venture
“I would take my Cuban again because he truly believed when absolutely no one else believed in us. And I'll be forever grateful to him for doing that.”