enzo retweeted
Celebrate GTA Online’s colorful cast of supporting characters with 2X GTA$ and RP on Contact Missions (stacking to 4X for GTA+ Members): rsg.ms/1c708ad
enzo retweeted
Crypto memecoin times officially ended today for me when I saw that.
@elonmusk personally QUOTED $Claudia post (together with the ticker btw) and the market cap can not even jump above $7 mln after his post.
Memecoins are done - we are cooked.
3 years ago it would be already at $100 mln.
enzo retweeted
not a single good and respected trader posted their october PNL
the only one's posting are sludge trading (streaming) terrible traders
they're feasting on all the 10-20k-30k toppers, because they dev and bundle the entire floor
it's all getting glorified too, it's fucking over
enzo retweeted
FORBIDDEN STUDIOS PRESENTS: MIGHTY ZION ORANGE RANGER 🍊⚡
Straight from the lost archives of 1995 Saturday morning television.
It's morphin' time.
While observing the universe, @NASAHubble caught a galaxy cluster smiling for the camera. 😊
The smile lines are arcs caused by an effect known as gravitational lensing, where massive structures warp spacetime with their powerful gravitational forces and act as a cosmic lens.
enzo retweeted
Drainer got your wallet? Let it sign all night.
On Bunker your wallet key can't withdraw. That takes a separate one-time key the drainer never saw.
Two keys. One door. Live on Solana.
enzo retweeted
The world’s first room temperature superconductor precursor born from Grok has been created at @thedeardao and @grok
Here, you can check the room-temperature superconductor predictive materials we created using Grok on our GitHub.
github.com/Deardaogit/prelim…
Read our story through the thread.
Elon Musk: I expect Grok to discover new physics by next year. Let that sink in.
“I want to just emphasize this point. With respect to academic questions, Grok 4 is better than PhD-level in every subject. No exceptions.
This doesn't mean that at times it may lack common sense and it has not yet invented new technologies or discovered new physics, but that is just a matter of time. I think it may discover new technologies as soon as later this year and I would be shocked if it has not done so next year.
I would expect Grok to literally discover new technologies that are actually useful no later than next year and maybe end of this year. It might discover new physics next year and within two years, I'd say almost certainly. Just let that sink in.”
Grok 4 presentation by xAI, July 9, 2025
Made and published via superstudio.so
enzo retweeted
Vitalik keeps making the case for privacy. $ENGRAM puts it where it belongs:
The mind that prices your trade has no secrets: 109 public bytes.
Its trainers work sealed, time-locked to drand, so nobody can copy them.
Power in the open. Work in private.
engramv4.xyz
enzo retweeted
imagine this, Elon tweets his new dog
immediately 1000 + coins are deployed , you buy the one on SOL but see another one on RH, BNB, and ETH
we now have a 4way chain pvp but thats not all, there is also a stonk pair, otc pad , agency pad , etc and to add to it there is a image/ ticker/ pairing / og /kol pvp you have to fight through
then on RH we have pons vs LONGS with the same other vamp attacks and im not even gon get started on BNB chain
finally (maybe) we have ETH which will most likely be the PVE runner but could also hard pvp
man ts is genuinely untradeable, just adapt and learn to be quick with your trades
enzo retweeted
Replying to @RealAngryDegen @Microsoft
You tagged me, so here's the real version. No hype. How we'd actually build a Quine explorer, and what it would do.
First, why Quine matters. Biology's bottleneck isn't data. It's the loop: guess, run the experiment, wait weeks, revise, repeat. Every wasted experiment costs months and money.
Microsoft's Quine goes after that loop. One world model trained across genomics, proteins, chemistry, cell state and imaging, plus a harness that plans the steps, pulls in tools and literature, and ranks what to test. With the Broad Institute it ranked compounds to shift pancreatic tumor cells between states in one weekend. Its top picks produced the biggest shifts in the lab.
The catch: it's closed. Access is a selective fellowship. Most labs, grad students and biotech founders can't touch it.
So that's what we build. The same loop, running on public data, no fellowship required. Two parts.
Part 1: the research explorer.
Quine sits on 40+ papers and tools across six research lines: proteins, regulatory genome, tumor microenvironment, pathology, microscopy, biological data. EvoDiff, FoldingDiff, the Dayhoff Atlas, Virchow, PRISM and the rest. We map every one in plain English: what it does, which part of the world model it feeds, and whether you can download and run it yourself.
Part 2: the AI layer.
You ask a biology question in plain English. It pulls live evidence across the same five data types Quine trains on:
genes: Ensembl, gnomAD, ClinVar
proteins: UniProt, PDB, AlphaFold
chemistry: ChEMBL, PubChem
cell state: GTEx, CELLxGENE, Connectivity Map
imaging: Human Protein Atlas, JUMP Cell Painting
Plus Open Targets, PubMed and ClinicalTrials.gov for what's already known and what's already in the clinic.
Then it reasons like a scientist, not a search box. It runs the method I already use:
- breaks the question into testable steps
- tags every claim measured, computed or inferred
- resolves contradictions instead of averaging them away
- kills obvious answers. If a postdoc with PubMed finds it in 30 minutes, it's not a lead
It ends where Quine ends: a ranked list of what to test, each prediction written down before the result, and the cheapest experiment that could prove it wrong. The lab result goes back in, the ranking updates, you go again. That's the loop.
Example. Ask: which approved drugs could push aggressive pancreatic tumor cells back toward the less aggressive classical state? It pulls the genes that define each state, checks which of them are druggable, finds drugs that flip that gene signature in Connectivity Map, drops the obvious ones, and ranks the rest with the experiment that would test each one.
Why this matters: Quine proves the architecture works at Microsoft scale. Nobody outside Microsoft can copy their trained model. But the loop runs on public data today, and the harness keeps a slot for their model the day they open it.
I've already run this loop for drug discovery. A wet lab put one of its leads into testing.
So yeah. Stay tuned.




























