20VC Newsletter - 13th April 2025
Here are the transcripts and top takeaways from 20VC episodes released this last week.
Monday’s episode with Victor Lazarte, GP @ Benchmark:
Download the full transcript:
My 7 key takeaways:
AI Is Replacing People and Big Tech Is Lying to You
Big company CEOs love to say “AI isn’t replacing people, AI is ‘augmenting people’s abilities’”
That is bullshit.
AI is fully replacing people.
The Most Important Question to Ask AI Companies Pre-investing
If models get a lot better, is your company worse or is your company better?
If the company gets worse, it is very hard to justify investing in it.
If the company gets better, they are in a very good position.
Is This a Bubble and Why Buying the Winners Always Works
US labor spend is 10x larger than software spend – and AI is going after it.
Someone will likely start a trillion dollar company in the next 3 years
We need to buy the early winners, even if they’ll seem overpriced in retrospect.
Two Traits That All the Best Founders Have
They are very open-minded, but very disagreeable.
They are interested in what you have to say but have no problem disagreeing with you.
Pedro @ Brex & Brendan @ Mercor have these traits.
Three Different Categories of Companies in a World of AI
Three categories:
Incumbents that will adapt.
Incumbents that won’t adapt & get recreated by startups on an agent-first framework.
Businesses that weren’t possible before that are possible now because of the agent interface.
Why China Is a Stabilising Force for the US
They are an external threat to the US.
The US cannot get distracted with internal conflict.
We are in an important AI arms race and the US needs to win.
If You Start a Company on the West Coast You Are 1,000X More Likely to Succeed
70% of all companies in the world are started on the West Coast.
The big difference is that there is so much knowledge about how companies are built.
When someone starts a company there, they are able to access all of that knowledge.
Thursday’s episode with Rory O'Driscoll, GP @ Scale & Jason Lemkin, Founder @ SaaStr:
Download the full transcript:
My 6 key takeaways:
Why Seed Investing Is for Suckers
Why try to 8x a seed fund over 20 years when you can write one big check into a winner & achieve liquidity in a quarter of the time?
The multiples will be lower, but the returns & will be higher.
When the outcomes today are $100BN+, seed is for suckers.
All the Best Companies Work 7 Days per Week
In 2021, your competition was working 10 hours a week from home.
Now they are working 100 hours a week.
If you haven’t evolved, you are going to die.
Why Thrive Are Masters of the Game Playing Monopoly
They have a brilliant strategy: Buy the best house on every block:
Fintech block: Stripe
AI block: OpenAI
Infrastructure block: Databricks
Now they can sit back & wait for the checks to arrive.
The Issues With Deploying $20BN
Two issues:
You see every good deal, but you also see every deal: Picking is more important than ever.
There may not be enough room for all that capital. It will only stop if the LPs’ bosses stop allocating capital into venture.
Why Fund Size Scaling Is Right and Series A and B Funds Have to Be 700-800M?
A $20M check is not enough to get you in the room now.
Your fund size is your strategy – You need 20-25 deals per fund, each worth $30M.
This means your fund size needs to be at least $700-800M.
Will We Have More or Less Money in Venture in the Next 5 Years?
The best time to invest is when no one wants to; the worst is when everyone does.
We will likely see a significant change in LP capital flows in the next five years.
It will be a better time to invest & a harder time to raise.
Friday’s episode with Richard Socher, CEO @ You.com:
Download the full transcript:
My 4 key takeaways:
Why Infra Layer Companies are in Trouble
High costs and low value capture make them the telcos of AI.
Commoditization pressures more than ever on thin infrastructure layers.
The future is shaky for those solely focused on AI infrastructure.
We’re in the “Valley of Disillusionment” for AI Action Agents
Horizontal agents are not good enough yet.
They do not understand the user enough to make nuanced decisions.
Enterprises want specificity, not generality, in their AI tools.
Why Google Pays Apple $20BN per Year
80% of iPhone users never change their default settings.
It is very hard to move away from that kind of lock-in.
Why Everyone Should Study Computer Science
It’s a different way of thinking & helps you understand the world better.
Even if you’re studying law or medicine you should combine it with computer science.
All of these fields will drastically change in the next couple of decades.
Let us know what your big takeaways from this week’s shows were in the comments below!
Thank you for reading, and don’t miss the great guests we have next week:
Monday episode: Jason Wilk, CEO @ Dave
Thursday episode: Jason Lemkin & Guest
Friday episode: Mayur Gupta, CMO & Growth GM @ Kraken
Thank you for reading 20VC.
This post is public so feel free to share it.
Thanks for reading 20VC! Subscribe for free to receive new posts and support my work.