20VC: Why Model Providers Will Kill Many Startups Moving into the Application Layer | Why Deepseek is not a Threat to OpenAI & Why OpenAI Beats Anthropic | Apps vs Models vs Infrastructure: Where is Value in AI with Sridhar Ramaswamy, Snowflake CEO

The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch

The Twenty Minute VC

Business, Investing, Tech News, News

4.4 • 693 Ratings

🗓️ 10 February 2025

⏱️ 55 minutes

🔗️ Recording | iTunes | RSS

🧾️ Download transcript

Summary

Sridhar Ramaswamy is the CEO @ Snowflake, the $60BN public company with $3.5BN in revenue growing 30% per year. Sridhar joined Snowflake following his company, Neeva, being acquired by them for $150M. Prior to founding Neeva, Ramaswamy spent 15 years at Google where he had an integral part in the growth of AdWords and Google's advertising business from $1.5 billion to over $100 billion.

In Today's Episode We Discuss:

1. OpenAI vs Deepseek vs Anthropic:

Why will OpenAI beat Deepseek? What does no one see with Deepseek that they should see?
Why has OpenAI beaten Anthropic? What elements turn a model from a commodity into a sustaining product suite?
Will model providers become application providers?
Will OpenAI be the biggest killer of startups in the next 10 years?

2. Snowflake vs Nvidia & Databricks:

To what extent is Sridhar concerned NVIDIA will move into the data layer and compete with Snowflake?
How does Sridhar view the competition from Databricks? What have they done better than them? What have they done worse than them and lost on?
Does being private hurt or help Databricks in their fight against Snowflake?
If Sridhar could, would he take Snowflake private today?

3. Leadership, Parenting, Money:

Do richer leaders make better leaders? How does being rich change the mindset of a leader?
What are Sridhar's biggest lessons when it comes to parenting?
What about the way that Sridhar was brought up, did he do deliberately differently with his kids?

Transcript

Click on a timestamp to play from that location

0:00.0	I think it is terrifying to be a startup building on top of Open AI.
0:03.4	The line between an infrastructure provider and application provider today is super blurry.
0:11.3	There's basically no guarantee that OpenAI or Anthropic or Microsoft or Google are not going
0:16.4	to go create.
0:17.2	You're not going to switch from chat GPT to Deepseek.
0:19.8	It's a product.
0:20.7	It's not a model. There's a big difference deep seek. It's a product. It's not a model.
0:21.6	There's a big difference, Harry. There's a reason why Anthropic hasn't done as well, because it operates
0:27.6	sort of at the model level. It's all of the additional things that go into chat GPD that I think
0:34.8	has staying power. This is 20 VC with me, Harry Stebbings.
0:38.2	Now, today we answer the core question.
0:40.2	Where will value be generated in the next generation of AI?
0:43.9	Will model providers become application providers and so much more?
0:48.0	Joining me for this incredible discussion, we have Shredha Ramoswamy, CEO at Snowflake.
0:53.2	Snowflake is a $60 billion public company
0:56.4	with $3.5 billion in revenue growing 30% per year. Before Snowflake, Shridhar spent 15 years at
1:03.6	Google playing a central role in the growth of AdWords and Google's advertising business
1:08.3	from $1.5 billion to over $100 billion in revenue.
1:12.6	But before we dive in today, here are two fun facts about our newest brand sponsor, Kajabi.
1:17.6	First, their customers just crossed a collective $8 billion in total revenue. Wow.
1:23.6	Second, Kajabi's users keep 100% of their earnings, with the average Kajabi creator
1:29.1	bringing in over $30,000 per year. In case you didn't know, Kajabi is the leading creator
	...

Please login to see the full transcript.

Previous episode | Next episode

Disclaimer: The podcast and artwork embedded on this page are from The Twenty Minute VC, and are the property of its owner and not affiliated with or endorsed by Tapesearch.

Generated transcripts are the property of The Twenty Minute VC and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.