Video · 2026-07-13 · 1h 5m · 6 moments

Wix Founder: Will Base44 Win the Vibe-Coding Wars? | The Truth About the Economics of Vibe-Coding

✦ AI generated

timeline · colored by role

Claim◆

Salesforce's core value isn't its CRM software—it's the deep trust that huge banks and enterprises place in it to hold their customer data, and that trust can't be replicated by a vibe-coded app.

Avishai argues Salesforce's real moat is trust, not its CRM features, so AI coding tools that can replicate features can't replicate that trust and won't displace it.

transcript

Avishai: Okay, Salesforce, what what is the biggest value Salesforce provide? It's not the CRM. It's the fact that you trust a SAS providers with your data, right? You have huge banks, huge companies that actually allow Salesforce to hold all of the customer data because they trust Salesforce.

Prediction

Wix's future customer base will split into three groups: small businesses that stay with Wix's no-code platform, those who fully switch to vibe-coding via Base44, and a middle group that uses both.

Rather than vibe-coding wiping out Wix's market, Avishai sees a three-way split among small-business customers, with Wix owning both ends via Wix and Base44.

transcript

Avishai: Essentially the way I look at it today is that there are three different options. One, the guy that built the pizza place. He said he want to build the website and the management tools for that. Okay, he's going to be staying a witch customer because he don't want to put their effort into that.

Data

Wix's self-trained AI model for Base44 only saves about 1% to 30% versus using frontier models for complex tasks, far less dramatic than the 14-16x savings others claim from open-source models on simpler tasks.

Avishai reveals that despite training their own model, cost savings versus frontier models are modest (roughly 5-10% for complex use cases), undercutting the narrative of massive open-source cost advantages.

transcript

Avishai: It depends, but I would say between 1% to 30%... if you're using open source models, okay, then and and you can use the really basic ones for a lot of different things, they work great, okay, then it would cost you even less. But when you want to use a complicated model, for example, what you need for something like B 44, okay, it's not going to get to that much of a level, but you can achieve 5% or 10%.

Claim◆

AI is given too much credit for what it can actually do—when you look closely at the details, it fails at a lot of things despite being amazing and powerful overall.

Avishai pushes back on AI hype, saying that despite being genuinely powerful, AI still fails at many tasks when examined closely, and people overestimate its current capability.

transcript

Avishai: I think that we all give too much credit for AI and what it can do, right? And and and and when you start looking at the details, it's not really doing that well for a lot of the things. It's amazing. It's incredible. It's super powerful, but and the butt is pretty big.

rebuts · 1supports · 4

Claim

The real value of making money isn't nicer service or status—it's freedom: freedom from fear of not having food tomorrow, and freedom to know you're here because you chose to be.

Avishai reflects that money's true value is the freedom it grants—both from fear of scarcity and to be somewhere entirely by choice, which he says fuels total commitment.

transcript

Avishai: money give you one thing beyond everything else. Okay? And that's freedom. Okay? Freedom from fear that you won't have food tomorrow, right? And then freedom also to know that you are here because you've chosen to be here, right?

extends · 4

Prediction

AI will take much longer than expected to replace humans, partly because the industry quietly redefined AGI to exclude the hard problems it still can't solve, once it became clear models weren't reaching the original bar.

Avishai says he's changed his mind in the last year: he used to fear rapid AI-driven human replacement but now believes it will take far longer, noting the goalposts for 'AGI' have shifted.

transcript

Avishai: I was very concerned very concerned with how quickly we're going to find humans being replaced with AI. And I think I'm now looking looking at the same thing. I think it's going to take much longer than I anticipated.

extends · 1provides context · 1supports · 3

Highlight slides

Salesforce's Real Moat Isn't the CRM✦ from: Salesforce's core value isn't its CRM software—it's the deep trust that huge banks and enterprises place in it to hold their customer data, and that trust can't be replicated by a vibe-coded app.Why Vibe-Coded Apps Can't Compete✦ from: Salesforce's core value isn't its CRM software—it's the deep trust that huge banks and enterprises place in it to hold their customer data, and that trust can't be replicated by a vibe-coded app.AI Gets Too Much Credit✦ from: AI is given too much credit for what it can actually do—when you look closely at the details, it fails at a lot of things despite being amazing and powerful overall.Hype vs. Reality✦ from: AI is given too much credit for what it can actually do—when you look closely at the details, it fails at a lot of things despite being amazing and powerful overall.

Related episodes

Audio · All-In Podcast

SpaceX-Cursor Deal, SaaS Debt Bomb, New Apple CEO, SPLC Indictment, Colon Cancer Spike

→

Video · Machine Learning Street Talk

Intelligence is collective, not artificial — Prof. Michael I. Jordan (UC Berkeley / Inria)

→

Audio · No Priors

Scaling Global Organizations in the Age of AI with ServiceNow Chairman and CEO Bill McDermott

→