Former dex, now your least favourite market makers biggest cheerleader. - Obvs opinions are mine, they are terrible. β©οΈ
Where the wild things are. π΄β
Joined June 2021
- Tweets3.9K
- Following690
- Followers735
- Likes34.9K
"I'm reminded of what Soros once said, which is the way to really make money in the market is figure out what it is the market believes is not true, and more importantly, when they'll realize they've been had.
I think we're in the beginnings of the awakening phase." - @gnoble79
OUT NOW - βCoreWeave, short of a bailout, is a zero, as are many of the neoclouds. And I put my money where my mouth is.β
So says George Noble @gnoble79 who predicts an "almost guaranteed car crash" because the AI Boom, like the Shale Oil Boom before it, requires continual increases in spending (ie capex) before ROI collapses.
George estimates that by 2030, the PROFITS for most hyperscalers will go NEGATIVE as depreciation catches up to revenue growth and reveals just how low he thinks the return on invested capital will be.
George cautions against shorting the companies that are actually making insane amounts of money ($NVDA and $MU etc) but he does strongly think that these companies are overearning and valuing them on current/NTM earnings is a mistake, and he makes a comparison to shipping in the 2000s.
George is honest humble and reflective about his prior mistakes including 2023 ETF. He is on a hot streak now especially on short side ($CRWV $IREN etc), calling 2026 an "unusually good year" especially given the bull market.
I pushed back plenty. We argued about memory oligopoly, the scale of enterprise AI demand, as well as AI Lab revenues and the degree to which it is inflated by early stage companies. Plus also...60% implied default by CoreWeave that sounds excessive etc.
Tons of zingers... if you want George to bring the heat you won't be disappointed.
We recorded this right after a record consecutive 9-day decline in semis (the sell-off is ongoing) so even George despite his views agreed we could be setup for a vicious bounce. But he warned that crashes only ever come from oversold conditions such as the ones we are in now.
$SPCX $TSLA $IREN $NBIS $CRWV
Apple
podcasts.apple.com/us/podcasβ¦
Spotify
open.spotify.com/episode/2Rpβ¦
YouTube
youtube.com/watch?v=dbMBhjCgβ¦
Today's sponsor Sarmaya Thematic ETF (LENS):
sarmayaetf.com/
"So, to be clear, tho I'm discussing this with a straight face, if you think SpaceX is really worth $2.3 trillion, which is where it is today, it's because you think Elon is some kind of anomalous genius.
And if you don't think that, you shouldn't be in the shares" - John Foley
The llm productivity trap is like the twitter info trap.
Can feel like you're more informed than people who don't live here, but really are just bombarded with mis/dis-information until your reality becomes a concoction of predictable reactions to that propaganda.
0xMaka π π³ retweeted
Trading has developed a version of me I never thought Iβd become.
The money is only one part of it.
The patience, discipline, emotional control, self-awareness and accountability I've developed have changed how I approach life itself.
Trading is genuinely life-changing.
Of all the projects winding down this is one that holds a special place.
I have memories attached to every poap, minted in relation to some legendary protocol meeting, or a big interactive event.
A really cool project I organically enjoyed using.
After 5+ years, we're winding down POAP.
Over the years, we:
* minted millions of collectibles across hundreds of communities.
* did incredible collaborations with world-class organizations - Coinbase, Amex, WMG, Bayer, and a long tail of others, including countless in crypto.
* ran one of the biggest activations at Devcon, minting 8000+ collectibles over the course of a week.
* built an Airport Rally to let collectors get POAPs at airports around the world.
POAP meant something to a lot of people.
Unfortunately, crypto's funding cycles and distribution dynamics made it hard to build a sustainable company without cannibalizing the ethos that made POAP mean something. Building on a fragile and quickly evolving stack, in the middle of an incredible hype cycle, only added to the challenges.
Still, there are some big lessons I am carrying from the POAP story.
First: Customer communities are the most undervalued asset a company can have. If you build something with soul, people who use it and love it will take your brand everywhere. Do not underestimate the power of customer communities.
Second: Connection is still the point. It's easy to lose that thread when the conversation is about the latest bubbly tech or price speculation, but the people using POAPs at events, in classrooms, at hackathons were getting something real out of it.
Third: Longevity and brand equity are the only moats left. With the time to building and shipping something new collapsing quickly, and distribution becoming increasingly an engineering problem, customers increasingly want to make sure they can trust that the things they are buying will grow and evolve with the market, without changing radically.
Grateful for everyone who believed in what we were building. It's too soon for me to say what's next, but for now, I'm thinking a lot about what GTM looks like in the age of AI.
"propped up, enriched, & empowered to serve as the face of a ruthless, hyper wealthy elite who want the power to dictate reality and exert control over society.."
- @truthispeaking
Forgot we have options expiring today, so prices will be pinned below concentrations of open interest.
Nice day for a fader, probably would take extra care if a break out trader, and not get greedy with tp's either way.
Am long $sndk today.. me the bearcat?!
But to max 1650, still fading any significant retest ~2000.
Targeting a percent here, a percent ~1575 if we make it, and the last portion ~1650.
Wary of a squeeze maybe to the mid point of the broken bear flag, so wouldn't fade too early.
google really has been kind, only took the one attempt to catch the pivot and was the only trade not to get stopped out today. Can't say sandisk played as nicely.
Did let this breath, was kindest trade last week.
But with expected vol today, will take a chunk here then trail stop for half around 353, leave stop for last just above 4h ema's ~361 for extended targets ~335
"So I've been here for 9 years now and I never experienced in my 9 years as CIO so many scheme members coming back to me with positive comments from a single investment decision, as I experienced in the wake of the exclusion of SpaceX."
- Anders Schelde : CIO at AkademikerPension
0xMaka π π³ retweeted
Proud to share one of our most important research projects to date!
We have discovered that the lack of proper commitment mechanisms makes PBS fundamentally incentive incompatible as a vehicle for Ethereum transactions.
// thread
New post on EthResear.ch!
Builders' Defection and Incentive Compatibility
By:
- @nuconstruct
π ethresear.ch/t/25400
Highlights:
- PBS creates an imperfect-commitment problem: builders observe all bids and full payloads, and nothing cryptographically forces them to honor the auction outcome rather than replicate/profit from bundles.
- In a repeated-game model, honesty is incentive-compatible only if the one-shot defection gain (Ξ) is smaller than the expected discounted loss from being caught: Ξ β€ p * (Ξ΄/(1βΞ΄)) * Ο_h; low detection probability p or low patience Ξ΄ makes defection rational.
- Empirically (libmev panel, Sep 2024βAug 2025), replication-style defection is likely hard to detect on-chain, implying low p (plausibly 0.01β0.1), under which incentive compatibility fails for every major non-TEE builder unless the builder is extremely forward-looking.
- The potential economic scale is large: the paper estimates roughly $26M/year of counterfactual defection surplus available to non-TEE builders in the panel, and ~$49M total exposure when including BuilderNetβs TEE-neutralized share.
- Therefore commitment must be restored architecturally, not reputationally: prevent builders from learning usable bid/payload details (e.g., TEEs like BuilderNet, commitment-based settlement designs, or cryptographic approaches such as threshold-encrypted mempools / commit-reveal / ZK), while noting ePBS (EIP-7732) and FOCIL (EIP-7805) donβt address this replication/confidentiality channel.
ELI5:
Ethereum blocks are often built by a βbuilderβ who gets to see everyoneβs secret trading plans (bids + transaction bundles) before deciding what goes in the next block. Even if the builder runs an auction and promises to include the winning bundle, it can secretly copy the best bundle for itself and keep the profit. The paper asks: can we rely on the builderβs reputation (fear of losing future business) to stop this? The answer is usually no, because copying is hard to detect and the one-time profit can be big. So the fix should be technical: design the system so the builder canβt reuse what it sees (using TEEs or cryptography), instead of hoping it behaves.
