I got married last month, and spent the furthest time away from work during that period and the honeymoon
Unplugging while we broke ATHs really helped me reflect on what I’ve learned up until this milestone and what’s important to me in life
I wanted to share this publicly mostly to document my thoughts. I hope it’s helpful but it’s not intended to be life advice or anything like that. Just my own musings which include some stories, quotes, and pictures here and there.
A new episode of Hivemind is out now!
This week the team discusses whether markets are near a bottom, what Jensen's exit means for Nvidia, how Anthropic's IPO could move the market and more.
Timestamps
00:00 Intro
00:43 Are We Near a Market Bottom?
04:23 Why the Risk Model Flipped to Risk-Off
10:45 Was the Weekend Washout the Bottom?
17:28 What Jensen's Exit Means for Nvidia
20:37 Still Long the AI Trade
29:23 The AI Cyberattack Wake-Up Call
31:51 Why Crypto Is a Honeypot Right Now
38:17 Dip-Buying Opportunity or Trap?
44:43 Bitcoin's Security Budget Problem
46:52 Could Something Flip Bitcoin?
49:44 Why Chains Should Go Bunker Mode
56:12 Anthropic's IPO and Its Unlocks
1:01:57 Would You Back Leopold Again?
1:12:04 One Asset for the Next 18 Months
1:15:09 Is Nvidia a Buy at These Levels?
A new episode of Hivemind is out now!
This week the team discusses whether the worst of the selloff is over, how the Anthropic IPO could trade, why stocks onchain have stalled and more.
Timestamps
00:00 Intro
00:45 Is the Worst Behind Us?
07:48 Why the Macro Models Are Cautious
10:25 The Nvidia Wild Card
16:11 Never Sell the Singularity
19:23 AI Stocks at COVID-Low Valuations
24:33 Meta's AI Bet Is Priced at Zero
28:01 How the Anthropic IPO Could Trade
34:00 Will Anthropic Wipe Out SPV Buyers?
38:04 The Mystery Morning Seller
39:42 ZEC in the Wall Street Journal
41:14 Why Stocks OnChain Stalled
44:28 Can Elysium Skip the Launchpad Trap?
50:48 Is ETH's Store of Value a Hail Mary?
56:14 MegaETH Pivots to AI Inference
In terms of work-hours, AI finishes a task about 12x longer every year. Task length doubles every 3.4 months. AI can do 55 hours of human work today. In a year from now, it should be able to do about 4 months of human work.
Open source models catch up to frontier models faster than ever before – and the catch up time is collapsing. Today, it takes 31 days for open models to catch up to frontier models.
A conversation with @jvisserlabs on why everyone giving up on crypto is doing so at exactly the wrong time.
We discuss the dot-com bubble comparisons, why Bitcoin is not a simple liquidity trade anymore, using China's ghost cities as a model, how the crypto & Ai trade are linked, and what this next phase will look like...
00:00 Intro
00:43 When Believers Give Up
03:07 Wall Street Calling
04:41 Bitcoin’s 7 Sigma Reversal
06:23 When Sellers Run Out
07:47 Price Moves Before Stories
09:17 Biggest Move Loading
12:01 Beyond the Money-Printing Trade
14:32 Ghost Cities
15:43 Does the Fed Still Matter?
19:39 Markets Shrug Off Bad News
21:23 Sponsored by Kraken
22:12 Agents As Economic Players
23:30 AI’s Acceleration Point
27:39 Why Agents Choose Crypto
32:06 Missing the Move
35:56 Who Profits From AI Agents?
38:06 Trapped Wealth vs Money
41:05 Robinhood's Role
48:44 Unpacking Tokenization
51:17 AI x Crypto Link
Our new report “Kinetiq: Building the Hyperliquid Economy” is live!
Trading on Hyperliquid has grown faster than the block space its apps run on. Kinetiq is building Elysium to give those apps 100 times the capacity.
Read the report here.
https://t.co/tMA09zdYja
Delphi Media is opening up sponsorships for Convergence, Hivemind and the Delphi Podcast
Crypto, AI, robotics, space, energy, biotech, markets. If it's shaping the next decade, we're covering it
If you want to sponsor one of the shows or learn more, DM @cashtro
Had a member ask me what I think about Blockworks going down the data route and why Delphi never went down that path
To be completely honest, we tried to build a data product a few years ago. Was really fucking hard. Getting the data was one thing, but standardizing and maintaining it was much harder than we had anticipated. Post FTX, we decided to not burn more resources chasing that direction even though we thought it could be valuable.
Its why I give @JasonYanowitz, @0xngmi, and @Adam_Tehc a lot of props. Got to spend time with each of them irl this summer (yes even @0xngmi) and made sure they knew how proud they should be for pulling off what they’ve done so far.
Often you don’t know how much work goes into something until you do it yourself. Over the years at Delphi, we’ve done a lot - Research, Consulting, Incubating, Investing, Media. All I’ve learned is that everything is hard and takes real effort. So you really have to be excited and passionate about what you choose to do, and when you commit: go all in.
As for @Delphi_Digital, we’re doubling down on Research. It’s been a core part of our DNA since the very beginning, and we feel it will only be more important as a signal in an accelerating world with more and more noise. Excited to share more about this in a few weeks.
Had a member ask me what I think about Blockworks going down the data route and why Delphi never went down that path
To be completely honest, we tried to build a data product a few years ago. Was really fucking hard. Getting the data was one thing, but standardizing and maintaining it was much harder than we had anticipated. Post FTX, we decided to not burn more resources chasing that direction even though we thought it could be valuable.
Its why I give @JasonYanowitz, @0xngmi, and @Adam_Tehc a lot of props. Got to spend time with each of them irl this summer (yes even @0xngmi) and made sure they knew how proud they should be for pulling off what they’ve done so far.
Often you don’t know how much work goes into something until you do it yourself. Over the years at Delphi, we’ve done a lot - Research, Consulting, Incubating, Investing, Media. All I’ve learned is that everything is hard and takes real effort. So you really have to be excited and passionate about what you choose to do, and when you commit: go all in.
As for @Delphi_Digital, we’re doubling down on Research. It’s been a core part of our DNA since the very beginning, and we feel it will only be more important as a signal in an accelerating world with more and more noise. Excited to share more about this in a few weeks.
I've been at Delphi for close to two years now.
What the team has been cooking up the past few months has gotten me the most excited since I first joined.
There is still a lot to do but looking forward to having everyone see it.
New Emerging Managers episode with Paavan Gami (@hiro_sugimoto_) of Raas Partners is live, our first with a public equities fund manager
Before starting Raas, Paavan spent 5 years at Greenoaks, where he was part of the now legendary Carvana trade (if you read Colossus’ Neil Mehta profile, you know the one). They sold before the crash in ‘21, bought back in after a 90% drop, then kept buying through another 75% drawdown while LPs told Neil the company was going bankrupt. They ended up owning ~4% of Carvana, a stake worth $1B+ by 2025, and the stock went on to 100x off its 2022 low
His line on how you get comfortable doing that stuck with me: “it’s very hard to run into a burning building if you don’t know the structure of that building already.”
Then he left one of the best seats in investing to make a 20-year bet on India. Beyond the macro case everyone knows (1.5B people, half under 30, 6%+ real growth a year for 25 years), his bigger thesis is that stock picking in India has far more alpha than in the US, for structural reasons:
- Private capital is thin, so the best companies go public early. That means you can basically do Series B/C-style growth investing in public markets
- The index is mostly state-owned companies, metals, mining, oil and gas (only 1 internet company in the top 50), so the market still doesn’t know how to price tech
- Most consumer categories are still ~90% mom-and-pops, so the leader at 5% share can compound to 30-40%
We get into all of it, plus whether AI breaks the India story (I’m more worried than he is), the two businesses he’s “married” (Eternal and Meesho), and how he uses AI agents in his process
(Disclosure: I’m an LP in Raas Partners)