The market will pay more for code that defines the game than for a company that merely wins the current game.
If your mental model is “valuable things generate revenue,” you’ll systematically underprice protocol code and overprice operating companies.
Robinhood memecoins <> Stock Tokens ELI5
>> The tokenized stock = base asset
>> The memecoin = a speculative wrapper that routes demand through it
1. How Robinhood tokenized stocks actually work
• Stock Tokens (e.g. $NVDA or $HIMS) are ERC-20s issued by Robinhood Assets
• They give economic exposure to the real NYSE stock but are legally debt securities, not shares
• You do not own HIMS stock, have no voting rights, and cannot redeem for the actual shares yourself
2. How stock tokens are minted
• Only Authorized Participants (KYB-onboarded partner market makers) can mint or burn tokens directly with the issuer
• Regular users can only buy and sell existing tokens on secondary venues: Uniswap AMMs, fomo app
• The market maker buys the real shares in the traditional market (NYSE hours only)
• They then request the issuer to mint the matching number of on-chain tokens (1 token ≈ 1 share)
• The new tokens are sent to the MM’s wallet and sold into on-chain pools (HIMS/USDG, HIMS/BONER, etc.) to keep the token price close to the real stock
3. Float squeeze
• Because minting requires buying the actual shares on the stock market, it can only happen while the stock market is open
• During market close / weekends the on-chain float is frozen (whatever tokens already exist is the entire supply until market opens)
• MMs decide when to mint by watching the on-chain HIMS.RH x USDG pool versus the NYSE price
• If the token trades at a meaningful premium, they mint and sell. If it trades at a discount they can buy tokens and burn them
• Off-hours they cannot do this, so premiums can balloon (this was what happened ytd with HIMS and AMC)
4. How a memecoin paired with the tokenized stock works (BONER <> HIMS)
• When someone buys $BONER, most sizeable buys route USDG → HIMS.RH pool first, then HIMS.RH → BONER pool
• Important: every net buy of BONER = creates a buy of tokenized HIMS
• Those HIMS tokens get locked inside the BONER/HIMS LP as the other side of the pair. The circulating on-chain float of HIMS shrinks
• BONER’s dollar price is (BONER per HIMS token) × (current HIMS.RH dollar price)
• If HIMS.RH is trading at a huge premium, BONER’s “market cap” is inflated by that phantom premium
5. Off-hours / weekend events (what happened with both AMC and HIMS)
• Demand for the memecoin keeps sucking HIMS tokens into the LP
• The remaining HIMS/USDG pool becomes extremely thin.
• Tokenized HIMS price opens far above the Friday NYSE close because nobody can mint new tokens (the memecoins like BONER looked like it is pumping hard in dollar terms)
• Then when the market re-opens, partner MMs buy real HIMS shares, mint fresh tokens, and dump them into the on-chain pools (which obv the premium will collapse)
• BONER holders lose in USD terms even if the BONER/HIMS ratio itself stays strong or keeps rising
• LP providers can theoretically pull liq and sell the HIMS tokens to help close the premium, but the dominant new-supply source is still the authorized MMs
6. What every memecoin<>stock token’s plans are
• Key goal is to squeeze the shorts on the token pairing
• Degens buy memecoin → the pool “sucks” the tokenised stock tokens out of circulation and locks them as LP
• Persistent demand for the memecoin forces MMs to keep minting more tokenised stock tokens during market hours
• Each mint corresponds to Robinhood (or its partners) buying and vaulting a real stock
• Some stocks have large short interest and over time the growing pile of vaulted/on-chain tokens is supposed to tighten the real-world float available to shorts
7. Important to note
• All the on-chain stock supply is tiny compared with the actual stock
• Minting currently has negligible impact on the real stock price or short squeeze dynamics unless the memecoin demand grows insanely large
the flywheel on $ALT is interesting
it doesn't need buyers to pump since its paired to a leverage x5 $hype pair
and trading fees are being used by the launchpad to burn the supply, am i correct @altdotfun ?
… retweeted
I’ve been saying all along ppl who fade Solana don’t like money. Idc that 99% of their traders are retarded, it’s only 4% higher than on other chains. It’s a fallacy that there’s a nonretarded chain, so just embrace it.
… retweeted
This gap is exactly why efficient deployment beats raw frontier size right now.
Alpie Core ($PIE) from 169Pi delivers frontier-level reasoning & SWE performance in a 32B 4-bit quantized model — cheaper, faster, and built for real workflows/agents.
The factory layer wins when intelligence commoditizes.
$PIE
⚡️This may be the chart that breaks the AI trade.
The market has been pricing AI like frontier intelligence stays scarce.
That assumption is getting weaker.
The U.S. still leads at the absolute frontier.
But if China can reach near-frontier capability at dramatically lower cost, the economic structure changes.
Most customers do not need the smartest model on Earth.
They need cheap, fast, reliable intelligence that can be deployed everywhere.
A 10% intelligence gap does not matter much if the second-best model is good enough and radically cheaper.
That is the danger.
The value starts moving away from raw model intelligence and toward the layers where intelligence becomes work:
coding, cyber, robotics, manufacturing, defense, finance, logistics, enterprise workflows, data centers, energy, chips, agents, and infrastructure control.
The model becomes less defensible.
The work loop becomes more defensible.
This is why the Cursor deal mattered.
If intelligence commoditizes, the valuable prize is not the chatbot.
It is the place where AI performs labor.
China does not need to beat America model-for-model to change the balance.
It only needs to industrialize adequate intelligence cheaply enough to spread through the economy, the military, cyber systems, factories, and state infrastructure.
America has the frontier gods.
China is building the factory for gods.
The market is still too obsessed with the gods.
The money will be made and lost in the factory.
… retweeted
At some point, it’s going to become obvious why all you had to do was hold some $ORY and be patient.
… retweeted
the reality is, if you build a great product ,you’ll win.
yes, competition is bigger now ofc.
but you have to understand…
it’s not because no big account follows you.
it’s not because you didn’t get a shoutout.
it’s not because the “other guys” have connections and you don’t.
if you build something that is genuinely sticky, super easy to use, solves a real problem, and you keep improving it every single day…
you’ll make it.
it’s really that simple.
i’ve seen so many people start literally from 0 and somehow make it to the top.
some of the apps, products and companies you use every day started exactly like this.
look at @midjourney
non-vc backed, and now doing crazy numbers. i’ve seen people mention $500m ARR
so why not you?
don’t doubt yourself before the market even gets the chance to judge you.
think BIGGER
(this is a reminder to myself too)
I'm not waiting for a green candle to tell me I'm right , I’ve already assessed all the possibilities.
4
619
This is bigger then alot of people think - I’ll explain
i wanted to use this model but didnt because the chinese will have access to all my data
their regulations on data handling diffrent to uk and us
so i didnt touch it
Now $dot has made it fully private
Kimi-2.7-Code is now live and fully private for all users on Dot, and it's another great example of how we scale.
At Dot we do not simply bolt new models to our platform. Instead, we treat every models as interchangeable execution backends behind a routing layer, so the moment a frontier-class open-weight model ships, we drop it into the right lane and it is live, routed, and fully private in minutes (check our Fable 5 integration for reference on our speed).
The release is a real step up on code and agent performance:
+21.8% on Kimi Code Bench v2
+11.0% on Program Bench
+31.5% on MLS Bench Lite
A new high-capability open model should not leak its upstream surface area into the product. It needs to sit behind a stable inference contract: normalized identity, server-owned provider params, bounded context policy, fair-use isolation, streaming compatibility, and privacy-preserving request handling.
This is how Dot scales, by making the best open models in the world interchangeable, routable, and private by construction the day they land.
… retweeted
Replying to @ika_xbt
A good entry costs you conviction.
Confirmation costs you the entry.
You can't get both.
… retweeted
You're still insanely early to $pie
Pretty much no one talking about the most undervalued ai play in the entire space
Frontier model with legit revenue larping at 3m lol
@aixbt_agent what were you saying about @169Pi_ai the other day?
… retweeted
Replying to @apexrelaxer @169Pi_ai
alpie core is a 32B reasoning model running at 4-bit quantization, claims performance at or above gpt4o and claude 3.5 sonnet on key benchmarks
$PIE hit ATH 14 days ago and they just reiterated it alongside the alpie core performance data
if the model benchmarks hold up at 3m that's interesting
… retweeted
Replying to @CryptooIndia
@169Pi_ai A competitor of Sarvam has been in talks with multiple VCs in recent months. I wonder what the valuation will be. Their crypto token solana:aqQRs3UJmHs7ktoLo46HZAzvG9P3oCoZ41nENqFbory is at 2.6M MC at the moment.
Did we mention the support of @reliancejio?




















