Blockchain & Startup Lawyer - I’ve worked with the best projects in the space. Founder Horizons Law; GP at Marin Digital VC; $PDT; Ex Solana GC, SNX, MyCrypto
Why crypto projects need both Delaware AND offshore structures, and how to get it right:
Most crypto projects start with a Delaware C-Corp which is familiar to investors, easy banking, and straightforward equity.
But stopping there is a massive mistake.
After your first meaningful raise ($1-3M), you need to establish your offshore structure BEFORE token valuation or issuance.
The gold standard:
>Cayman Islands Foundation (headless, no equity ownership)
>BVI subsidiary where token minting happens
>Professional directors ($30k/yr is cheap insurance)
Budget reality: Set aside $140-180K for proper setup. Yes, it seems expensive, but fixing structural problems later costs 10x more.
The structure creates crucial separation:
>US entity → development and research
>Cayman foundation → governance and treasury
>BVI entity → token issuance
Common mistakes:
>"I'll do it later" (creates tax nightmares)
>Using cheap jurisdictions (exchange rejection)
>Keeping too much founder control (regulatory red flags)
>US person dominance in offshore entities (IRS problems)
Second-time founders never make these mistakes. They know proper structure is the foundation that enables their project to scale globally.
Don't be penny-wise and pound-foolish. In an industry where you might build a billion-dollar protocol, the difference between proper and improper structuring can represent hundreds of millions in value.
During earlier cycles, we had Bitcoin OGs who were really here for the tech.
They were the best influencers.
Then we had Ethereum bros who were sorta here for the tech.
Now all we have are memecoin guys.
Hope we can get some better influencers soon.
This is pretty insane.
Give me 1 other company that very frequently cuts their prices by 80%.
These frontier models are in an epic race to achieve AGI and find out how to monetize it. In the meantime they need to keep everyones attention with the latest model and make a few bucks doing it.
major price cuts today:
*80% drop for GPT-5.6 Luna, now $0.20 per million input tokens and $1.20 per million output
*20% drop for GPT-5.6 Terra, to $2/$12
*GPT-5.6 Sol gets Fast mode in the API, up to 2.5x the speed for 2x the price, same intelligence
“Top decile” DPI on a 9 year old venture capital fund is 1.18x. That’s terrible
1confirmation net DPI since we started:
Fund I (2017): 7.08x
Fund II (2019): 3.25x
Fund III (2022): 0.19x
Q2 is in, and it was a monster quarter.
→ $116B in merchant sales, up 32% YoY
→ $3.6B revenue, up 34% YoY
→ $654M free cash flow, 18% margin
Five quarters running, GMV has grown more than 30%.
The bigger story is underneath the numbers: every shift AI is creating happens to play to @Shopify's strengths.
BREAKING: Kalshi has removed the New York AG's civil enforcement action to federal court. The case will be assigned to either Judge Torres or Marrero. This should delay things by a few weeks. NYAG will file an emergency motion to remand it back to state court. It will be granted.
If you ask Claude for a legal opinion, it gives truly terrible answers.
It is terrible at just stating the facts without mentioning that you should take the most risk-averse possible route without considering anything else.
I wonder about the long-term consequences of caring too much about the quick money side of their business.
They also launching Robinhood banking around the same time.
I make 5,000 USD/month from referrals. This is me without even trying/sharing referral links socially.
@useTria is gonna revolutinize earnings for ambassadors and also for users. The more you refer the more you earn.
Launching this feature for all the ambassadors soon.
Cc: @triajapan@triakorea
LLMs are great at a lot of things.
But it's making the dumb people dumber and the smart people more efficient.
People will just write 20-page memos with Claude and not proofread them.
Complete AI slop.
It still has trouble really understanding and caring about the right context to write something meaningful.
Very, very good at math. Science. Discrete questions. Anything with a definitive answer that requires collating a lot of disparate data.
But as of now, it's really best at being the world's best search engine.
new feature observed on hl testnet: stars
afaict stars let deloyers make a hip-3 dex address-allowlisted for trading. testnet allowlist size currently is max 10k addresses. non-allowlisted addresses can still submit reduce-only orders / fund accounts.
see ktob ("BTC Star DEX")
some relevant txs:
0x851e0bc62988fe9e8697042529220801080023abc48c1d7028e6b718e88cd889 (register ktob:BTC)
0x54c1a9c94451ac7e563b0425292214010100c1aedf54cb50f88a551c03558668 (activate star)
0x875b29d555ffdba188d4042529221c01030041baf0f2fa732b23d52814f3b58c (approve trader address)
0x9dbce81f1a86dc269f360425292224010c000004b589faf841859371d98ab611 (unhalt ktob:BTC)
0x3130533baff0a23a32aa0425292ce1010a006b214af3c10cd4f8fe8e6ef47c24 (unapproved address funded)
0x8923485ccfdea73b8a9d0425292d2c01010060426ad1c60d2cebf3af8ed28126 (unapproved address attempts to place order and fails)
0x326fcd720ba84bd833e90425292e06010400e557a6ab6aaad63878c4caac25c2 (approved trader address funded)
0xe070d0c9f5aca692e1ea0425292e33010300e8af90afc56484397c1cb4a0807d (approved trader address places order successfully)
can think of a few reasons why this might be useful, but will refrain from speculation for the time being.