A research-driven firm dedicated to making crypto happen sooner and better than it would without us.
🌍
Joined September 2018
- Tweets12.5K
- Following1K
- Followers1.6M
- Likes9.7K
Pinned Tweet
A new episode of Hivemind is out now!
This week Hypercall founder @trajan0x joins the team to discuss what the Anthropic IPO means for markets, how cyberattacks are exposing the internet’s weak points, and whether onchain options can finally break out.
Timestamps
00:00 Intro
00:45 Patience Is a Position
03:49 Chop Until the Midterms
09:33 Can Efficiency Crush the Neoclouds?
13:10 What if Compute Gets 10x Cheaper?
17:09 Is the Internet Broken?
23:20 Could You Cash Out Satoshi’s Coins?
27:16 Is Quantum Theft Really Consensus?
35:31 What Led Synapse to Options?
42:54 What Onchain Options Do Better
46:41 Who Wins Options Distribution?
51:31 Credit Is the Next Frontier
55:09 Would You Buy the Anthropic IPO?
57:36 The Biggest Liquidity Drain in History?
58:35 Why Early Stage Venture Looks Best
1:03:22 Elon Comes for the Neoclouds
1:07:32 SpaceX and the Kessler Syndrome
1:08:53 What Are You Underexposed To?
José explains why falling compute prices are a risk for AI labs.
“If there’s an industry-wide slowdown in demand, Anthropic has unused capacity and the spot price crashes, so it’s hard to resell, which is kind of what happened with telecoms in 2001."
A new episode of Hivemind is out now!
This week Hypercall founder @trajan0x joins the team to discuss what the Anthropic IPO means for markets, how cyberattacks are exposing the internet’s weak points, and whether onchain options can finally break out.
Timestamps
00:00 Intro
00:45 Patience Is a Position
03:49 Chop Until the Midterms
09:33 Can Efficiency Crush the Neoclouds?
13:10 What if Compute Gets 10x Cheaper?
17:09 Is the Internet Broken?
23:20 Could You Cash Out Satoshi’s Coins?
27:16 Is Quantum Theft Really Consensus?
35:31 What Led Synapse to Options?
42:54 What Onchain Options Do Better
46:41 Who Wins Options Distribution?
51:31 Credit Is the Next Frontier
55:09 Would You Buy the Anthropic IPO?
57:36 The Biggest Liquidity Drain in History?
58:35 Why Early Stage Venture Looks Best
1:03:22 Elon Comes for the Neoclouds
1:07:32 SpaceX and the Kessler Syndrome
1:08:53 What Are You Underexposed To?
The seven most-traded HYPE options contracts are all on Derive.
Together they did $5.45M in notional over the last 24 hours, about 43% of HYPE options volume across Derive and Deribit.
The top four are all puts expiring in November.
• $2.20M in the Nov 20 $75 put
• $1.11M in the Nov 13 $80 put
• $621K in the Nov 6 $84 put
• $450K in the Nov 27 $85 put
View the crypto options dashboard here for free.
delphi.link/CryptoOptions
Delphi Digital retweeted
We’re bringing Power Perps to Hyperliquid, starting with HYPE².
If HYPE doubles, the HYPE² index quadruples. Options-like payoffs. No strikes. No expiry.
Power Perps bring options flow to Hyperliquid by pooling demand for convexity in one order book.
This work builds on “Power Perpetuals” by @danrobinson,
@_Dave__White_, @andrewjleone and their coauthors.
Why do the biggest startups keep getting more expensive?
José: “I think the mega fund bid and the size of these funds, it just acts almost like indexes in public markets...
The equivalent of index inclusion is when you get to a big enough size that a16z, Sequoia and these mega funds can invest in you.
And there’s only a certain limited number of names that you can plow nine figures or multi nine figures into.
And so those things just keep going up.”
A new episode of Hivemind is out now!
This week Hypercall founder @trajan0x joins the team to discuss what the Anthropic IPO means for markets, how cyberattacks are exposing the internet’s weak points, and whether onchain options can finally break out.
Timestamps
00:00 Intro
00:45 Patience Is a Position
03:49 Chop Until the Midterms
09:33 Can Efficiency Crush the Neoclouds?
13:10 What if Compute Gets 10x Cheaper?
17:09 Is the Internet Broken?
23:20 Could You Cash Out Satoshi’s Coins?
27:16 Is Quantum Theft Really Consensus?
35:31 What Led Synapse to Options?
42:54 What Onchain Options Do Better
46:41 Who Wins Options Distribution?
51:31 Credit Is the Next Frontier
55:09 Would You Buy the Anthropic IPO?
57:36 The Biggest Liquidity Drain in History?
58:35 Why Early Stage Venture Looks Best
1:03:22 Elon Comes for the Neoclouds
1:07:32 SpaceX and the Kessler Syndrome
1:08:53 What Are You Underexposed To?
Delphi Digital retweeted
Why did Chinese labs beat crypto at training AI models?
“There’s 10,000 possible tricks you can do, and you can only pick a hundred in your model... The Chinese ones, their hundred was just better. That’s why they got to building world-class models faster.”
How can small teams compete with the incumbents?
Alex Wilson: “When they offer products like that, they purposely make it sort of generic and broad so it works for a lot of different types of customers.
It’s hard for them to go all in on, let’s say, the nonprofit market and really specialize on that.
We just learned what nonprofits actually want and need and said, we’re going to build it just for them.
If anyone besides a nonprofit wants to use this, they can’t. It won’t actually work or make sense. It was that specific.”
New Delphi Podcast with Alex Wilson (@AlexWilsonTGB), co-founder of Cyclops and The Giving Block
He sold The Giving Block to Shift4 a year after raising. Now he's building stablecoin rails for payments companies
Shift4 is a customer and an investor. Out of stealth in March, Series A in July, more than 10 customers this year when they hoped for 3
His take: most of the world's 200 currencies won't make it. Fewer than 25 left in five to ten years
00:00 Intro
04:39 The Door-to-Door eBay Hustle
15:59 The $50M Anonymous Donation
17:41 Beating Coinbase by Going Narrow
19:55 Sold to Shift4 a Year Later
25:23 From Shift4 to NASA
27:31 Lessons From Jared
29:33 Hiring to Scare People Away
31:08 Why Start Cyclops
36:22 A Ticket to Space in Stablecoins
37:54 Who Stablecoins Cut Out
40:15 Agents Need New Checkouts
43:04 200 Currencies Down to 25
46:12 Seed to Series A in a Year
49:38 Swinging for a $10B Exit
Full episode
Many of the power plants being built for AI do not show up in official data.
Official energy statistics only track plants tied to the grid, so much of what AI campuses build on site goes uncounted.
Plants like these are expected to power more than half of the new U.S. data center capacity coming online in 2028.
Builders are bringing their own power because the grid cannot connect new load fast enough. One off-grid campus in Texas alone plans more than 500 gas engines.
Engine makers’ order books show the buildout more clearly than the official data does. INNIO booked $6.1 billion of orders in the first nine months of 2026, nearly double a year earlier.
Caterpillar’s backlog of large engines has more than tripled since early 2024, and Wärtsilä’s energy orders have more than doubled since early 2025.
These plants are likely to stay after the grid arrives. Some providers now build and run them for data centers under contracts that last up to 18 years, including extensions.
A growing share of AI’s power now comes from plants beside the grid, many of them outside the official numbers.
Read the full report here.
delphi.link/Power
How can onchain options compete with traditional exchanges?
@trajan0x: “If you want to do an at-the-money S&P 500 call a week or two out, your minimum trade size is potentially a couple thousand bucks...
All these options protocols allow fractional options, which makes it so you could trade options a year out, and your minimum contract size is still going to be a few bucks which is huge.”
A new episode of Hivemind is out now!
This week Hypercall founder @trajan0x joins the team to discuss what the Anthropic IPO means for markets, how cyberattacks are exposing the internet’s weak points, and whether onchain options can finally break out.
Timestamps
00:00 Intro
00:45 Patience Is a Position
03:49 Chop Until the Midterms
09:33 Can Efficiency Crush the Neoclouds?
13:10 What if Compute Gets 10x Cheaper?
17:09 Is the Internet Broken?
23:20 Could You Cash Out Satoshi’s Coins?
27:16 Is Quantum Theft Really Consensus?
35:31 What Led Synapse to Options?
42:54 What Onchain Options Do Better
46:41 Who Wins Options Distribution?
51:31 Credit Is the Next Frontier
55:09 Would You Buy the Anthropic IPO?
57:36 The Biggest Liquidity Drain in History?
58:35 Why Early Stage Venture Looks Best
1:03:22 Elon Comes for the Neoclouds
1:07:32 SpaceX and the Kessler Syndrome
1:08:53 What Are You Underexposed To?
Delphi Digital retweeted
A new episode of Hivemind is out now!
This week Hypercall founder @trajan0x joins the team to discuss what the Anthropic IPO means for markets, how cyberattacks are exposing the internet’s weak points, and whether onchain options can finally break out.
Timestamps
00:00 Intro
00:45 Patience Is a Position
03:49 Chop Until the Midterms
09:33 Can Efficiency Crush the Neoclouds?
13:10 What if Compute Gets 10x Cheaper?
17:09 Is the Internet Broken?
23:20 Could You Cash Out Satoshi’s Coins?
27:16 Is Quantum Theft Really Consensus?
35:31 What Led Synapse to Options?
42:54 What Onchain Options Do Better
46:41 Who Wins Options Distribution?
51:31 Credit Is the Next Frontier
55:09 Would You Buy the Anthropic IPO?
57:36 The Biggest Liquidity Drain in History?
58:35 Why Early Stage Venture Looks Best
1:03:22 Elon Comes for the Neoclouds
1:07:32 SpaceX and the Kessler Syndrome
1:08:53 What Are You Underexposed To?
Tarun Chitra argues a single AI model may not be enough to run a robot.
“A lot of the evals have shown that segmenting to different models for different subtasks and then putting it together is oftentimes giving you much better performance than a single model.”
New Delphi Podcast with Tarun Chitra (@tarunchitra), founder of Gauntlet and managing partner at Robot Ventures
Robot just led the $90M round in @NousResearch. Tarun spent three to four months on diligence, from OpenRouter data to GitHub forks, before backing Nous and Hermes Agent
We get into why crypto's first run at AI went wrong, why the crypto people who dropped the token won, why OpenClaw is the worst developer experience he's used, and why Hermes is closer to Ubuntu than to Android
He also found Chinese robot makers building on Hermes, and some inference providers gaming cache pricing
His market structure take: all the money goes to the top and the bottom. Harnesses on top, inference underneath. Hermes is like investing in Phantom or FOMO
00:00 Intro
06:34 Robot Leads Nous's $90M Round
09:37 Crypto AI Was Meme Coins With More Steps
14:40 Why Nous Stood Out Early
28:14 Hermes vs OpenClaw
35:39 The End of AI Lock-In
46:33 The Ubuntu of AI
56:17 Microsoft and Nvidia Join the Round
59:42 OpenRouter Is a DEX for AI
1:07:02 All the Money Goes to the Top
Disc. Delphi Ventures co-led the seed of Nous Research and invested again in the Series A and B
Full episode
Delphi Digital retweeted
America needs new power lines far faster than it can build them.
Department of Energy studies imply the U.S. needs about 5,000 miles of new high-capacity transmission a year. The country built fewer than 900 miles in 2024. Most new data centers need a connection to that grid, so the pace of grid construction will set the pace of AI this decade.
Money alone cannot close that gap. Record utility spending has not shortened the 5 to 15 years it takes to permit and build a high-voltage line. A journeyman lineworker still takes about four years to train however much the job pays.
Existing lines can carry much more in the meantime. Restringing old towers with advanced wire can raise a line’s capacity by half or more within a few years. A Duke University study found today’s grid could absorb 76 GW of new load if data centers cut back just 0.25% of the time.
Until new lines catch up, the fastest gains will come from getting more out of the grid that already exists.
How many national currencies will survive a world of stablecoins?
@AlexWilsonTGB: “Stablecoins are dollarizing all these other economies. People around the world that don’t have access to a major currency want U.S. dollars.
I don’t think emerging countries are going to go down without a fight. But we have nearly 200 currencies around the world. I think it’s probably less than 25, even five, ten years from now.
So much of international business is done in U.S. dollars already. And if it just keeps getting easier and easier, there’s going to be less and less reason to use the local currency.”
New Delphi Podcast with Alex Wilson (@AlexWilsonTGB), co-founder of Cyclops and The Giving Block
He sold The Giving Block to Shift4 a year after raising. Now he's building stablecoin rails for payments companies
Shift4 is a customer and an investor. Out of stealth in March, Series A in July, more than 10 customers this year when they hoped for 3
His take: most of the world's 200 currencies won't make it. Fewer than 25 left in five to ten years
00:00 Intro
04:39 The Door-to-Door eBay Hustle
15:59 The $50M Anonymous Donation
17:41 Beating Coinbase by Going Narrow
19:55 Sold to Shift4 a Year Later
25:23 From Shift4 to NASA
27:31 Lessons From Jared
29:33 Hiring to Scare People Away
31:08 Why Start Cyclops
36:22 A Ticket to Space in Stablecoins
37:54 Who Stablecoins Cut Out
40:15 Agents Need New Checkouts
43:04 200 Currencies Down to 25
46:12 Seed to Series A in a Year
49:38 Swinging for a $10B Exit
Full episode
America needs new power lines far faster than it can build them.
Department of Energy studies imply the U.S. needs about 5,000 miles of new high-capacity transmission a year. The country built fewer than 900 miles in 2024. Most new data centers need a connection to that grid, so the pace of grid construction will set the pace of AI this decade.
Money alone cannot close that gap. Record utility spending has not shortened the 5 to 15 years it takes to permit and build a high-voltage line. A journeyman lineworker still takes about four years to train however much the job pays.
Existing lines can carry much more in the meantime. Restringing old towers with advanced wire can raise a line’s capacity by half or more within a few years. A Duke University study found today’s grid could absorb 76 GW of new load if data centers cut back just 0.25% of the time.
Until new lines catch up, the fastest gains will come from getting more out of the grid that already exists.
Read the full report here. First 30 people get it for free.
delphi.link/AIboom1
