Nov 24, 2025 · 1h 18m · 20vc
Base44’s Founder, Maor Shlomo on How Vibe Coding Will Kill SaaS · 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 the 20VC podcast, host Harry Stebbings interviews Maor Shlomo, founder of Base44, to explore the disruptive power of vibe coding, the technical economics of running AI startups, and the personal sacrifices behind building and acquiring multi-million dollar software platforms.
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 13.6% of the talking time here. How this is scored →
Harry as informed peer 4.2 Guest teaching 4.8 Guest disagreement 2.9 Harry pushing back 3.6
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 6:01 Rejection of Traditional SaaS Dominance
When Harry flatly rejects the idea that non-technical users will build custom software, Maor strongly defends his thesis that liquid software will displace traditional SaaS giants.
Hardest push from Harry ▶ 6:01 Host Challenges Custom CRM PremiseHarry forcefully interrupts Maor to tell him his core belief about users building custom CRMs is wrong and asks if he is serious.
Biggest teaching moment ▶ 25:46 Instant Volatility in LLM SpendMaor educates Harry on modern AI operational reality, explaining how changing a single code string shifts hundreds of thousands of dollars in spend between model vendors overnight.
Harry holds his own ▶ 11:56 Citing Tech Founders on Vibe Coding MoatsHarry demonstrates strong market awareness by citing explicit pushback from Marc Benioff, Mike Cannon-Brooks, and Dylan Field regarding the lack of defensible moats in AI app builders.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Episode Hook and Base44's Success Story | 2 | 2 | 1 | 1 | Harry introduces Maor and sets up the Base44 acquisition story with Wix. Maor explains his reasoning for selling rather than staying bootstrapped or raising venture capital. The dynamic is welcoming and narrative-focused with minimal challenge. | |
| Why Vibe Coding Will Displace Traditional SaaS Giants | 6 | 5 | 5 | 8 | Harry aggressively challenges Maor's thesis that vibe coding will replace traditional SaaS giants like Salesforce and Intercom, telling him directly that he is wrong. Maor holds his ground firmly, arguing that software is becoming liquid and customizable for small businesses. | |
| Building Real Moats in the Vibe Coding Era | 7 | 6 | 3 | 6 | Harry leverages insights from industry leaders like Marc Benioff, Mike Cannon-Brooks, and Dylan Field to question whether vibe coding platforms have any real moats. Maor responds with an architectural breakdown differentiating superficial front-end builders from complex cloud platform integrations. | |
| The Future of the Vibe Coding Market and the Threat of Model Providers | 6 | 7 | 4 | 5 | Harry maps out the competitive landscape across developer tiers and asks Maor to rank direct competitors like Lovable and Replit. Maor dismisses direct competitor threats, educating Harry on how the real existential threat lies with model providers like Google and describing the fluid spend shifting across frontier LLMs. | |
| SaaS Margins, Model Routing, and Real-Time Sentiment Analysis | 6 | 6 | 4 | 7 | Harry pushes hard on the weak gross margin profile of AI wrapper applications, expressing skepticism over promises of model routing. Maor defends the unit economics by detailing how Base44 measures prompt sentiment in real time and routes low-complexity queries to cheaper open-source models. | |
| An Angel Investor's Playbook: Revenue, Moats, and Vertical Integration | 5 | 6 | 2 | 3 | Harry asks how early-stage investors should evaluate rapid ARR growth in AI startups given transient customer spend. Maor breaks down Cursor's vertical integration playbook and explains why investors must look past vanity top-line numbers to assess infrastructure moats. | |
| How Base44 Beats Cursor and the Fallacy of Winner-Take-All | 6 | 6 | 6 | 7 | Harry asks Maor to choose between investing in Cursor or Cognition and pushes back when Maor rejects the venture capital winner-take-all thesis. Maor insists AI markets are expansive enough to support multiple massive niche winners. | |
| Silicon Valley vs. Global Tech Ecosystems | 3 | 4 | 2 | 2 | Harry asks whether operating outside Silicon Valley hurts Base44 compared to US rivals. Maor responds that global tech dynamics have shifted and that Wix provides all the backing he needs. | |
| Advice for Non-Technical Builders | 3 | 5 | 1 | 1 | Harry asks for practical advice for non-technical builders looking to create software. Maor explains the mental shift required to rapidly iterate and throw away prompt generations without traditional development attachment. | |
| Does Money Buy Happiness? | 2 | 3 | 1 | 1 | Harry asks Maor if selling his company for $80 million made him happy. Maor reflects on the utility of wealth for buying time and supporting family, lightheartedly recalling a friend's advice on jet skis. | |
| OpenAI vs. Anthropic | 3 | 4 | 2 | 2 | In a quickfire round, Harry asks Maor to pick between OpenAI and Anthropic, add a dream board member, and share startup regrets. Maor favors Anthropic for its rapid execution and B2B focus. | |
| The Delusions of AI Frontier Labs | 4 | 6 | 5 | 3 | Harry asks what frontier AI labs are most delusional about, and Maor argues they overestimate their moats given how easily customers can switch models overnight. He also shares advice on choosing a supportive life partner. | |
| Advice to My Younger Self on Base 44 | 1 | 3 | 1 | 1 | Harry asks Maor what advice he would give his past self before founding Base44. Maor reflects on learning to navigate early server crashes and volatile consumer feedback emotionally. |
Statements from this episode (32)
Shlomo: Profit margins are the least concerning issue in AI software
“From all things in the industry, I think the margins are the least thing that I'm worried about.”
Shlomo: Existing AI models have barely unlocked their economic value
“Even with existing models, we're only Scratching the surface of economic value that you can create.”
Shlomo: AI enables tiny teams to create immense value with fewer people
“If one person team can get sold for eighty million dollars without people, it means that you'll be able to do a lot more stuff with less people.”
Shlomo: Base44 was highly profitable before its acquisition by Wix
“And then, and like the business was very profitable.”
Shlomo: Joining Wix tripled Base44's odds of success over VC funding
“It was either I raise a lot of money and a lot of funding, or I go and join Wix that basically triples my chances at actually building something special.”
Shlomo: Vibe coding will become software's largest category, swallowing traditional SaaS
“I felt like in my eyes, this is going to be the largest category in, in software industry. Cause a lot of like existing categories that we're familiar with. And CRM and support systems and task management and project management and stuff like that are going to…”
Shlomo: Building custom CRMs via AI will soon beat buying Salesforce licenses
“And there would be a time, it might be in a few months, might be in a few years, where it will be easier to build your own Salesforce type of CRM than to actually buy a license to something off the shelf.”
Shlomo: SaaS companies requiring implementers will not survive the decade
“I think every software company that needs implementers, right, there's like a Salesforce implementer or whatever maybe in a decade from now is not going to make it.”
Shlomo: AI will eliminate moatless CRM companies, but Salesforce can adapt
“Smaller CRM companies, every software that's basically front end on top of a database with some team features and some integrations and does not have a moat. And I think Salesforce is so big and so massive that this is like, this is one of their moats. Like th…”
Shlomo: AI tools building simple front-ends and landing pages will commoditize
“I think early on, the vibe coding category was focused a lot and still is, to some degree, on Like front-end and landing pages where many use cases were on, like websites or e-commerce websites and stuff like that. I think this will get commoditized for the ca…”
Shlomo: Platforms building full SaaS apps will resist AI commoditization
“But for the functional stuff and the complex applications, the organizational applications, the personal applications, the full fully fledged SaaS platforms that people are building, I think they won't get commoditized.”
Shlomo: Existing SaaS tools will integrate built-in vibe coding customization
“I can see a future where a lot of the existing SaaS tools and SaaS products have a built-in vibe coding customizes the software to your needs capability.”
Shlomo: Base44 is not worried about rivals Lovable, Replit, or Bolt
“I'm not worried about those three.”
Shlomo: Google is most likely to win the LLM race
“Google. By far. They have the compute. They're moving very fast. If Gemini at some point wins the race, like literally wins the race then they have the entire stack. Including like Google Cloud and whatever, whatever they need and they have the data and whatev…”
Shlomo: AI foundation model prices will drop toward zero
“We're taking into consideration that the prices of model will go down towards zero.”
Shlomo: Margins for AI coding platforms will improve drastically
“I don't think current margins are bad, but I think they're going to get way better.”
Shlomo: Competitors copy Base44 features within weeks or months
“Every Feature that we put out. We know that's going to take either a few weeks or a few months for competitor to copy, and we see that a lot.”
Shlomo: Base44 processes tens of thousands of prompts per minute
“Nowadays we get, like, tens of thousands of prompts, like, a minute or whatever in tandem, and so we measure the sentiment”
Shlomo: AI apps can fix low margins by training proprietary models post-distribution
“Kerser came out, I think it was two weeks ago, or maybe a few weeks ago with their own model, right? Composer. And this is like an, a nice way to like think about strategy of those types of businesses. So think about Kelsa, right? I don't know how the margins …”
Shlomo: Base44 will fine-tune open-source models on billions of code lines
“And I think this is a really interesting strategy that we're also thinking internally is, like, go as fast as you can with the frontier models. At some point, open source becomes really a viable option to maybe fine tune on, like, the billions, almost trillion…”
Shlomo: AI startups relying primarily on prompt engineering lack defensible moats
“If most of the value Is on the way they prompt the application for a certain thing. Like they've built, let's say a photo editor. And so there's not a lot of features in a photo, like an AI based, like you feed a picture and say, okay, make me, I don't know, m…”
Shlomo: Passed on AI agent startups like Manus over commoditization fears
“I had the opportunity to invest in like this agent companies, right? Like the Manus, the, those folks. I don't like, I felt like, I don't know the industry good enough to understand if at some point this would get commoditized by the model companies. And so th…”
Shlomo: AI software building is not a winner-take-all market
“Software building as a category is gonna grow insanely fast. I think our notion from the previous age of venture capital where a winner takes it all is wrong. And I think this category is going to be so big that everybody will take a niche at some point.”
Shlomo: The AI industry will not hit a progress speed bump
“Me personally, I don't believe we'll hit a speed bump. I think even with existing models, I don't think also we'll hit like a wall with like the improvements of LLMs, but I think even with existing models, we're only scratching the surface of economic value th…”
Shlomo: Operating outside Silicon Valley does not hurt tech startups
“I don't think in any way, no business is hurting. If anything I think the war on talent is really tough in Silicon Valley.”
Shlomo: 95% to 100% of code will be AI-written in two years
“95%. Or a hundred percent, for sure. It's like, ah, you'll be prompting agents.”
Shlomo: Nearly 90% of Base44's code is already AI-generated
“For us, base 44 was built in the age of LLMs, and so I structured the repository in a way that's going to be very easy for LLMs to navigate through and to read the code. I think for us, it's already closer to 90%.”
Shlomo: Planning ahead to save developer time is obsolete with AI
“Iterations is everything. It's, like, it's not anymore about, like, planning ahead so that you save developer time”
Shlomo: Anthropic has better models and growth trajectory than OpenAI
“Moving faster. They correct something in coding. They're very focused on delivering value to B to B customers of, they don't have the first mover advantage, but I think they have better models. And when they look at their trajectory they're better positioned f…”
Shlomo: AI model market will not be winner-take-all
“It's like, either it's not going to be one winner takes it all. And there's going to be like more and more companies getting into this market. It's going to be like the Chinese players that are now like changing the entire market dynamics.”
Shlomo: Developers will stick with Claude Code over superior models
“Like, even if now Gemini three comes out and it's a better model than Cloud and so on, then still there's gonna be a huge number of people staying with CloudCode because they're used to it and they built some things up the stack.”
Shlomo: Consumer feedback spans a wider emotional spectrum than B2B SaaS
“Also consumer, or the consumer market is so much different. Like the emotional spectrum of the feedback that you get is so much wider than building a B to B SaaS, that you get like some, ah, very, ah, not, not, not too positive, not too negative, ah, comments.”