
Investor, digital assets. Research first capital. Early-stage & infrastructure. APAC markets | Views my own @Panga_Ventures
Joined August 2011
- Tweets11.5K
- Following793
- Followers38.9K
- Likes19.7K
Andrew Lee retweeted
My favorite thing of the last 12 months was people telling me they max LTV borrowed against crypto to buy $DRAM… the (Leo)^2 derivative
Andrew Lee retweeted
Let this Linux kernel CVE related to IPv4 (unused!) options be a reminder to us all: we didn't have to do this IPv6 thing. The IPv4 header has unused options fields that were already variable length to 40 bytes! We could have just cannibalized a few of those bits way back in 1995, had plenty of address space to work with this whole time, and as a bonus not had this remotely exploitable vulnerability in the Linux TCP stack.
ubuntu.com/security/CVE-2026…
Andrew Lee retweeted
So fun to do DealBook with my friend and partner of 11 years, @gdb. Thrilled the world got to see the brilliant, fiercely determined leader I have the privilege of working with every day.
Greg’s role at OpenAI has evolved from writing code 100 hours a week in the early days to now leading some of the hardest and most consequential parts of our business.
One thing that has not changed is his extraordinary commitment to our mission. I can say with confidence that OpenAI would not exist without Greg, and I believe the world would be on a much worse path toward the AI future without him.
Andrew Lee retweeted
Breakpoint is a wrap.
3 days at Olympia with @solana teams. Wallets, staking, payments, tokenized stocks, launchpads.
Nobody asked how to get more wallets. Everyone asked which users are worth it.
10M Solana wallets covered on @addressableid , targeting live.
Andrew Lee retweeted
YOU CAN SEE EVERYTHING, PRIMETIME, and DIGGER make 2026 the best year for film since 2015 (SICARIO, EX MACHINA, and MAD MAX: FURY ROAD). At long last, movies are back.
Andrew Lee retweeted
He's nothing like this. The tech industry is nothing like this. A freakish 'everyone who's rich is dumb and weird and evil' fantasy for the groundlings.
Andrew Lee retweeted
This is great, and all businesses should run with this kind of messaging. At checkout, companies can break down each additional tax with the name of the politician responsible. A list of Mamdani fees clearly marked on every purchase just might change the country.
Andrew Lee retweeted
I see VCs on X criticizing "Marxism" all the time, but strangely, I think VCs might be the strongest modern advocates of class consciousness.
The advice they give over and over: "don't sell your labor." Which is also the central tenet of Marxism (“alienation”)!
Taken to its logical conclusion, everyone a capitalist~=everyone a communist.
Autonomy shifts spend from labor to compute and liability.
If you wanted to know all the jobs that AI will produce, look no further than cybersecurity. This is going to become one of the biggest investment areas for enterprises going forward.
And even as the work becomes more agentic and autonomous you will need more talent to point the agents at the most critical work, review findings, patch systems, respond to incidents, and more.
Lots of opportunity right now for the systems that protect data and enterprises, as well as the professionals that are deploying them and protecting organizations.
It’s going to only get more wild out there.
313
Andrew Lee retweeted
My kids now have to solve age appropriate math problems before each episode of Bluey, thanks to the miracle of AI reverse engineering and upgrading my TV 🔥🔥🔥
Andrew Lee retweeted
You can play the claw machine yourself using stablecoins (there's a faucet to get $0.25 of OUSD)
The machine auctions off the right to play it, using the Machine Payments Protocol (MPP)
The payments go into an onchain pot, and if you pick up the ball, you win half the pot
The real cost isn't the collision. It's the capital tied up in insurance premiums and legal fees while you learn a basic mechanical function.
381
Andrew Lee retweeted
Introducing: The Claw
While you were setting up your OpenClaw we set up a physical claw machine in our office and hooked it up to the Tempo blockchain
Play it yourself, or train your own robotics model to play it and climb the leaderboard
🧵
Andrew Lee retweeted
Centaur is the apex enterprise cloud agents product - we built it for Paradigm and Tempo’s unique needs, happy to add more providers
This quoted post is unavailable.
Accumulation is only valid if the mandate allows drawdowns.
In times like this, if you're a long term investor, you don't want to be waffling about being in or out, you want to be stacking units systematically. Don't get confused by the traders, it's a separate game.
On unit accumulation, depending on how long you've been in the game, it's helpful to think in "relative units" of the investment you did best with. I'll use $ETH because the math is easy and relates well to younger networks.
ETH has ~120M units. Let's say I once held 10 ETH.
Now I'm looking to stack $SOL, but SOL has ~636M units, or ~5.3x the # of units. Therefore, to have an equivalent position size in network terms, I'd need 53 SOL. If I want to own double the network % of SOL, that I owned of ETH, I'd need 106 SOL.
Or, perhaps I'm considering $TIA, which has ~1.16B total supply, or ~9.7x the # of units of ETH. In that case, I'd need 97 units of TIA to have a network equivalent. If I want to place a bigger network bet on TIA, because the network value is lower than when I entered ETH, then 10x the % ownership would require me to accumulate 970 TIA. 100x the % ownership = 9,700 TIA. Going for significantly more ownership, at cheap prices, if you're right, can lead to asymmetric outcomes.
$ZEC is another good example of this, since its supply mirrors $BTC's. And so it's easy to think of your "dream # of BTC," and aim to accumulate that in ZEC, if you're so inclined.
Pick your networks of choice, and do the math. I have always found thinking in unit terms keeps me grounded, optimistic, and goal oriented in dreary times. I can then dream about future upside in the thick of the bull, but the important thing is to systematically stack the units with a high conviction methodology in the bear or early-bull pullbacks.
In terms of where we are now, which is probably why most of you are still reading :) BTC has been bleeding out the long-tail while remaining resilient, with many names 20-40%+ down from the highs while BTC is down just shy of 10%. You can either say BTC has already bled the risk appetite out of the market and so most of the work is done, or you can say BTC hasn't dropped much so it'll be really terrible if it goes cliff diving. I'm sympathetic to the former view.
I read this as an early-bull opportunity rather than a warning sign that this move was a "bull trap," and so I'm inclined to be a buyer. You have to respect the possibility of lower levels, and so tranche out any dry powder that remains with % cash used corresponding to % probability those levels are the local bottom - yes, you likely end up with some excess cash if not all levels are hit, but that's okay. And don't forget, there's already opportunity in front of you.
380
Andrew Lee retweeted
Once again Yuval resorts to personal attacks on me without fully engaging on a complicated topic. (He also has some weird fetish about mine being a teacher, but that's his business).
So let me elaborate my critique of the utility of tokenized bank deposits like JPMD in simple language that hopefully he can understand.
The biggest difference between the liability of a fractionally-reserved bank and that of a narrow bank (aka stablecoin issuer) is run risk. Levered banks live under constant threat of runs, narrow banks don't.
Due to run risk, a tokenized deposit can never be a bearer instrument. The bank must know who every token holder is at all times to manage the FDIC (or foreign equivalent) insurance limit. Otherwise a single customer could put $1m into 4 wallets and quadruple their insurance, thus Sybil Attacking the FDIC.
Also due to run risk, tokenized deposits cannot be allowed to trade freely in the secondary market. Stablecoins like USDT and USDC have micro-depegs all the time due to liquidity, but nobody cares because they are fully reserved. A tokenized deposit having even the slightest depeg could cause a run on the entire bank.
Run risk is why bank deposits have opt-in ownership (everyone must be KYC'ed) but stablecoins have opt-out (almost nobody does, unless there is a law enforcement action).
And BTW, every single senior bank exec (including at JPM) and bank regulator I've ever discussed this issue with agrees with my logic. The concept of a tokenized deposit moving as freely as a stablecoin is terrifying.
This is also why stablecoin issuers have gone out of their way to promote their liability for agentic use while banks have outright banned it.
Yuval's point about "stablecoin issuers also holding deposits" is irrelevant since they are fully reserved.
He also seems to imply the law will change to force stablecoins issuers to also have to KYC every holder. Genius doesn't touch this, nor does any regulator I know. This change would destroy stablecoin utility.
Now, knowing all of this, why do I think products like JPMD are mostly innovation theater? Because you literally can't do anything with them you couldn't also do with an API and 30 year old database tech inside JPM.
You can't use JPMD in DeFi unless you are only trading with other JPM customers and only against other assets directly on JPM's balance sheet. Why you'd want to pay Ethereum gas fees and deal with 12 second slot times to do the same thing the bank can handle instantly for free is beyond me.
Also, any swap of JPMD against a different asset (say ETH) runs the risk of JPM refusing to honor the deposit swap leg of the transaction. This swap isn't atomic and has high settlement risk.
Can stablecoin issuers also do this and interfere in DeFi? Yes, but they go out of their way not to, because having their coins featured in DeFi is great for business - unlike for banks.
Also, stable issuers don't know who 99% of token holders are, so have no basis to interfering. Banks know who every depositor is and have a long history of censoring, including against the crypto industry.
I'm guessing the reason Yuval is upset about my post is because JPMD is soon coming to Canton. That deployment actually makes sense. They should definitely do it.
Canton is just a database with a button. JPMD is an asset with a button. Both feature a single entity, possibly a single person, who can endogenously void holder rights at any time. Thus both have infinite settlement risk.
Infinity multiplied by infinity is still infinity.
Simple math that even a non-academic could understand.
“Professor” who doesn’t know that so-called regulated stables have bank deposits. Some regulations require them to hold bank deposits. They can depeg.
He uses “regulated,” then says it will be “less restricted,” meaning the current state of regulation. He might realize that regulation is fluid and therefore regulated stablecoins could be more restrictive too?!
He talks about a book transfer. What is a stablecoin if not book transfers with a rule that only certain accounts can create/redeem? Stablecoins can freeze assets.
JP and other tokenized deposits can participate in DeFi if they meet an asset issuer’s requirements no different from stables that cannot if the issuer freezes the wallet.
This is who we let teach our kids. Shameful.
Fail in the corporate world, fall back to “academic,” and you end up as Omid.
Convenience constraints destroy unit economics before technology scales.
Waymo and Zoox first impressions
I use self driving on my Tesla all the time, but I finally got around to trying the commercial autonomous ride hailing services this week in Las Vegas.
The biggest issue by far is the very limited pickup and dropoff points, and only having a couple minutes to get to them resulted in a bit of a rush.
If you are going to have a small-integer number of locations, you should make it super-obvious where those locations are. Waymo included a picture, but it was a rather undistinguished parking garage image and it wasn’t clear which floor it was on. Being able to switch the map to an aerial photography view might be helpful in some cases. Pan-able 360 photos of the target points might also be helpful.
The first Zoox ride was flawless, but we had a lengthy walk to get to our actual destination.
The first Waymo ride was a little rougher. It felt more indecisive with steering, like Teslas did a couple years ago. This was still early-access in Vegas, so it is probably better in the active commercial markets.
The second Zoox ride had a small incident – there was another Zoox ahead of us at a stoplight (and a Waymo and another Zoox next to us!) that repeatedly tried to back up into our Zoox, to the point that ours actually honked at the one ahead of it a couple times. Eventually, after multiple light cycles, ours started to tentatively try to get around it, which left it partway into the next lane and distressingly close to a fast moving semi. The one ahead finally moved, and the ride continued normally.
Overall, the Zoox experience felt a bit more charming than the Waymo, with several nice little touches, but they both get the job done. I still need to try a Cybercab.
237
























