Note: This article is meant for educational and entertainment purposes only and is not to be taken as investment advice.
After spending a few months in early-stage venture investing – listening to lots of pitches and investing in several companies, I am sharing my observations and learning so far. With vibe-coding and AI assisted work in pretty much every aspect of value creation at a company – the way I think about moats is completely changing.
The traditional way I think about competitive advantages – what Warren Buffett calls “moats” – is being completely rewritten by AI. And I’m not talking about some gradual shift here. I’m talking about complete annihilation of advantages that took decades to build.
Let me paint you a picture of what’s happening out there.
The AI Gold Rush is Real
We’re in unprecedented territory – 2025 is the first year in history where more than half of all global VC money is going into AI startups. We’re talking 52.5% of all venture capital (Source: Best Brokers). The $192.7 billion that’s flowed into AI companies in 2025 alone? Naturally, that’s more than any previous full year. Ever.
But here’s what’s surprising and terrifying: it’s not just about the money flowing in. It’s about what’s being destroyed in the process.
Network Effects: Remember Chegg?
Network effects were supposed to be the ultimate moat, right? More users create more value, which attracts more users. Facebook’s social graph. Google’s dominance. Seemed unbreakable.
Then look at what happened to Chegg.
May 2nd, 2023 – the CEO gets on an earnings call and admits “ChatGPT is having an impact on our growth.” Stock crashes 48% in a single day CNBC.
But wait, it gets worse. Today? Chegg’s stock is down 99% from its peak. They’ve lost half a million paying subscribers. $14.5 billion in market value – poof, gone Gizmodo.
Think about this for a second. Chegg spent years building a network of 79 million homework problems. They had 70,000 expert contractors answering questions. Students paid $20 a month for this.
ChatGPT made all of that worthless overnight. Free. Instant. No subscription needed.
And it’s not just education. Google’s search market share has dropped from 80% to 70% in just one year, while ChatGPT went from 1% to 11% Marnoa. For the first time in 20 years, Google’s looking vulnerable.
Brand Power is Evaporating
Adobe – the gold standard of creative software for 40 years. You wanted professional tools? You paid for Adobe.
Their stock is down 36% in the last year while the tech sector is up 24% Nasdaq Finviz. Why? Because Midjourney built an AI tool that’s captured 21 million users and is charging $10/month while Adobe charges $21 for Photoshop alone Ainvest.
Here’s the kicker – Coca-Cola, one of the world’s biggest brands, just used OpenAI and Runway for their ad campaigns instead of Adobe. When your biggest customers are jumping ship to use AI tools that didn’t exist three years ago, your brand moat is gone.
Adobe still has a lot to offer and I know it’ll come back, however so many companies – Canva, Figma, Midjourney, Tik Tok’s Capcut etc. are challenging Adobe’s dominance with much less budget and time.
The Speed of Disruption is Insane
Remember when we used to talk about the “10-year disruption cycle”? Forget that. AI is compressing it to months.
Switching costs? AI automates migrations now. What used to take months and millions of dollars can happen over a weekend.
Scale advantages? Late-stage funding for AI companies went from $48 million average in 2023 to $327 million in 2024 NatLawReview. Why? Because a 5-person team with AI can now do what a 500-person company did two years ago.
Data moats? Everyone thought data would be the new oil. Turns out synthetic data and transfer learning mean you need way less proprietary data than we thought.
Here’s Where It Gets Really Interesting
The top 1% of AI startups are getting valued at 3-10x normal multiples. Seed rounds hitting $161 million valuations – that’s 123% higher than the 95th percentile SaaStr.
Thinking Machines Lab just raised $2 billion at a $10 billion valuation. They haven’t even launched a product yet Best Brokers.
This isn’t rational by traditional standards. But maybe the traditional standards are what’s broken.
The New Moats: Marketing, Sales, and Customer Success
Here’s what I’m seeing that nobody’s talking about enough: The new moats are Marketing, Sales, and Customer Success.
Think about it. When anyone can build anything with AI in days, when switching costs disappear, when brand premiums evaporate – what’s left?
The human stuff. The ability to:
- Cut through the noise and tell a story that matters (Marketing)
- Actually understand and solve customer problems (Sales)
- Make customers so happy they never want to leave (Customer Success)
As one investor told me, with AI making it easier to build and copy products, “tons of companies will be doing the same thing” – so you need capital for distribution, marketing, and standing out Fortune.
The irony is beautiful – in the age of artificial intelligence, the most defensible advantages are profoundly human.
What Actually Works Now?
From what I’m seeing in the trenches, here’s what creates real defensibility today:
1. Compound Learning Systems – AI that gets smarter with every user interaction. These “systems of intelligence” create advantages so deep that competition becomes nearly impossible Greylock Partners.
2. Workflow Lock-in – Not the old “it’s painful to switch” lock-in. The new “we’re so embedded in how you work that leaving would be like cutting off your arm” lock-in.
3. Speed of Iteration – Forget annual product cycles. The winners ship improvements daily, sometimes hourly.
4. Human-AI Hybrid Networks – Combining AI efficiency with human creativity and judgment in ways pure AI can’t replicate.
5. Go-to-Market Excellence – And I can’t stress this enough – the ability to sell, market, and retain customers is now THE differentiator.
The Uncomfortable Truth for Investors
In 2024, VCs poured $209 billion into 15,260 deals Inc. But here’s the thing – the number of deals is actually declining while round sizes explode. Investors are concentrating bets on proven winners rather than spreading them around Best Brokers.
AI now represents 35.7% of global VC deal value, up from 24.7% in 2023 FTI. This train isn’t slowing down.
For those of us in early-stage investing, this creates a weird paradox. Starting a company has never been easier. Building a defensible one has never been harder.
The Bottom Line
The frameworks we learned from Buffett, from business school, from decades of pattern recognition – they’re not wrong. They’re just incomplete for this new world.
Traditional moats made sense when advantages eroded over decades. Now they can evaporate in quarters. Sometimes weeks.
The companies that will win aren’t those trying to build deeper moats around their castles. They’re the ones who’ve accepted that the moats are gone and are building something entirely different – businesses that thrive on change rather than resist it.
The ultimate lesson? The more artificial our intelligence becomes, the more valuable authentic human connection becomes.
Marketing that resonates. Sales that solve real problems. Customer success that creates genuine delight.
The moats haven’t just been crossed – they’re being filled in, paved over, and turned into highways. And the winners aren’t the ones trying to rebuild them. They’re the ones learning to thrive without them.
P.S. – For those asking about specific investment implications, remember this isn’t investment advice. Just observations from someone trying to make sense of this crazy new world we’re building.
Here’s Warren Buffett’s framework:
Warren Buffett’s “economic moat” refers to a sustainable competitive advantage that protects a company’s long-term profitability from rivals
. While different sources may categorize them slightly differently, the primary types of moats identified by Buffett and related investment principles are:
- Intangible Assets: These include valuable, non-physical assets that are difficult to replicate, such as strong brand names (like Coca-Cola or Apple), patents, trademarks, and regulatory licenses. A powerful brand creates customer loyalty that is hard for competitors to overcome.
- Network Effects: This type of moat is created when the value of a service increases as more people use it. Platforms like Google and Facebook benefit from this, as their large user base makes the service more valuable to both new and existing users, creating a “winner-take-all” dynamic.
- Cost Advantage: A company with a cost advantage can produce goods or services at a lower cost than its competitors. This allows them to either charge lower prices to attract customers or maintain higher profit margins. GEICO and Costco are examples of companies with cost advantages.
- Switching Costs: High switching costs make it difficult and expensive for customers to switch to a competitor’s product or service. This is often seen with complex enterprise software, where the time, money, and effort required to move to a new provider deter customers from switching.
- Efficient Scale: This type of moat exists in niche markets or industries where there are a limited number of competitors, making it inefficient for new players to enter. Utility companies are a common example.
A great business has an “enduring” moat that protects high returns on invested capital. Buffett advises managers to focus on “widening the moat” by improving customer satisfaction, reducing costs, and innovating, as competitors will always try to attack a profitable “castle”.