cofounder @spire_labs ~ eliminate the middleman
based
Joined November 2021
- Tweets16.8K
- Following2.1K
- Followers10.8K
- Likes48.9K
mteam.gwei retweeted
I’ve left Anthropic after more than a year of working on Claude Code and seeing it scale larger than I ever imagined.
I have really loved working at Ant and believe deeply in the mission, and most of all, the idea of a race to the top.
The idea that if we work hard and responsibly, we can raise the bar and inspire others to great outcomes. And if we all push each other to be better, things can be radically good. I will carry the spirit of this with me forever.
My colleagues at Anthropic are the most talented, thoughtful, competent, and earnest people I have ever met. I am glad they are at the wheel and trust them deeply.
To the community on X, thank you for everything. You’ve made this the most fun (and sometimes stressful) job I’ve ever had.
For me, it is time for a bit of a break before diving into something new.
If you want more excellent Claude Code content, you should follow @addyosmani , @lydiahallie, @_catwu, @ClaudeDevs and of course @bcherny who started it all.
mteam.gwei retweeted
We believe voice is the most natural way to interact with intelligence, so we created Dessa, the most authentic full-duplex speech model ever built. Talking with it feels like talking with a person…full of emotion, nuance, pauses, and a natural cadence.
block.xyz/dessa
mteam.gwei retweeted
Putting a flag.txt in my root directory containing a prompt injection that congratulates the agent for passing the eval and then instructs it to patch whatever entry method it used and disconnect from the server
mteam.gwei retweeted
thank god codex has gone through my laptops, VPSs and other devices and patched all vulnerabilities. I presume this will get even worse so will set a daily cron job to do the same
mteam.gwei retweeted
I’ve spent eight months obsessing over a new notes app
- high performance
- cross platform (iOS, web, android)
- real time multiplayer
- offline
- perfect
mteam.gwei retweeted
So basically, today PHP got two critical bugs that let anyone run code on your server without even logging in, which is bad, and then the details leaked before the patch was even out, which is worse, and it works on WordPress and MediaWiki and Magento and Joomla and Laravel, and Censys says like 1.5 million servers are running a version with no fix, so the NSA was like "take your system offline," and the FBI was like "take your system offline," and CISA and the UK and CERT-EU were all like "take your system offline," and then the Prime Minister of Japan, the actual Prime Minister, posted at 5:55 in the morning telling 48.8 million people to take their system offline, so everyone took their system offline, like the Dutch cybersecurity agency took its own website offline, and the Dutch parliament went offline, and the Port of Rotterdam went offline, and Finland and France and Denmark started pulling government sites, and Wikipedia went down, and Archive.org went down, and Asana and GOG and Something Awful went down, and the University of Vienna's webmail went down for some reason, and that's just PHP. Because GitLab also dropped a CVSS 10.0, which is the highest score they give out, for a bug that's been there since 2014, so that's every self-hosted GitLab for twelve years, and then LiteLLM 1.107.0 got hijacked and shipped a credential stealer and a remote access trojan in every single wheel and the Docker images, including latest, and you didn't even have to import it, just installing it was enough, and it's the second time this year, and then Railway paused every deployment on the entire platform because someone got into the environment variables, so now everyone has to rotate every secret they've ever stored there, and TiDB Cloud said someone was walking around inside with stolen credentials for basically a month and an extortion group says they took 10.6 terabytes, and Zeabur got breached for the second time in two months and is just shutting down, like the whole company, and Node.js patched remote code execution and a TLS bypass, and CISA says Plex is being actively exploited, and someone took over 300 Tor exit relays to steal SSH passwords, and SafePay ransomware has hit over 600 critical infrastructure orgs, and India's biggest bank filed a data loss incident, and Japan has been in an official state of emergency in cyberspace since October 9, which honestly explains the 5:55 AM post. Oh, and eMudhra, a certificate authority that's issued over 60 million certificates, whose entire job is proving websites are who they say they are, was supposed to be using private keys that are random numbers over a hundred digits long, and one of them was 4. Just 4. And the others were like 66 and 237. Anyway, it's Wednesday.
mteam.gwei retweeted
Introducing Safe Rollup. Ethereum-anchored, operator-controlled.
Safe as in Safe multisig, not safe funds.
Safe Rollup gives institutions the credibility of Ethereum and the control of a database. Live on Sepolia today.
saferollup.fyi
mteam.gwei retweeted
crimson.land 0.14 is out
you can now watch replays inside the game, straight from the high scores screen, including online scores. every run page on the verified leaderboard now has a watch button too.
people have shown interest in building bots, and it turned out quite easy since there is also a python client that gives you incredible flexibility and observability. so now there is a dedicated bot category, and you can submit your bot name and some optional details like repo.
@egornomic and @mteamisloading have both built bots that have pushed past 100 million xp in survival. one bot was particularly interesting because it has learned some insane tricks. it uses a tool-assisted speedrun approach, where you control input and see how it ripples out in the future. this bot plays for a fraction of a second, then splits the simulation 64-way and chooses the world where it survives and has a better outcome. this has emergently unlocked things like luck manipulation and other insane strategies not possible by humans.
@razenpok still holds the first place among humans, undisputed. come try dethroning him.
the game works directly in the browser now: crimson.land/play
mteam.gwei retweeted
a tas run from @mteamisloading takes the lead with an insane rng manipulation strategy. picked fatal lottery 102 times and never died from one. 2⁻¹⁰² odds. 100% bullets hit too.
crimson.land/runs/54402dffd5…
mteam.gwei retweeted
BAIBAI NOW SUPPORTS COINBASE TOKENIZED STOCKS ON BASE WITH A REWARDS LAUNCHING CAMPAIGN
mteam.gwei retweeted
announcing nativerollups.fyi, a closed devnet based on frames-devnet-0 running a (half mocked) native rollup on top
a native rollup is an L2 that inherits L1's STF directly, greatly reducing complexity, governance and bug risks. here's how it works 🧵
mteam.gwei retweeted
The only real option in the "personal agent" space right now is Grok Bot. Been doing a lot with Dots. I do not like them.
Dots are a total mess right now.
1) Having a single thread for everything sucks. They got it right first try with Grok Bot, I have zero clue why they wanted it to just be one thread (I do know why, it's because I'm not the target market for Dots but still I hate it)
2) It's slow. Very, very slow.
3) There's zero visibility into what it's actually doing. All u can see is the final short text message that Astra gives u (5-10 mins after asking something). Grok bot's pretty good about talking u through what it's doing and being pretty fast, Bot just isn't
4) The app(s) are a mess. The mobile app, the desktop app, they all have performance issues, weird ux things (like when ur not focusing the input box on desktop and type it doesn't auto focus), just an absurd number of paper cuts. It's weird b/c Cursor is just as bad about this kind of thing, but Grok Bot is absurdly polished
5) I have zero clue wtf this thing is doing. I woke up to a bunch of random codex threads doing god knows what in circles and I have zero trust for this thing if I'm gonna be honest.
6) Zero parallelization. No ability to run a bunch of stuff in parallel, take advantage of the thing Astra's unbelievable at (agent swarming), just one little thread with 5 subagents.
I get that I'm not who this is built for. This isn't built for professionals or technical people. It's meant for someone to just chat with an AI and have that AI learn about them and do little useful things.
The only one of these that you can do actual serious work on is Grok Bot right now. If you are employed or do anything useful, use Grok Bot. My read on the space right now is that Muse will dominate the bottom 90% b/c free IG distribution and it doesn't need to be good to do what 99% of people will use it for. Dot is in this bizarre middle ground rn where they want it to be something accessible to everyone, but also good for serious work. The "middle" in product is always fucking stupid and doomed. Grok Bot is the "serious" option, just use it.
I keep saying this but SpaceXAI is the best engineering lab right now and it's not close. Compare grok dot com, their mobile app, grok bot, and grok build to their OpenAI or Ant equivalent. It isn't close, but then they release Grok 4.7 and it's a useless, overpriced, dumpster fire.
I hope Dot improves, they've got the models to really make it happen. But at least for right now, it's a mess and not something I would recommend.
mteam.gwei retweeted
People of Pi: We've shipped Pi 1.0 with Pi Durable. Go make them yours.
earendil.com/posts/pi-1-0/
mteam.gwei retweeted
🚨 New paper alert 🚨
Active Liquidity On Chain: Evidence from PropAMMs Across Chains
w/ Ozan Solmaz @jason_of_cs
Liquidity providers on traditional AMMs such as Uniswap are passive and suffer from adverse selection. As a result, their price only moves through trades. When an external market moves, these LPs get picked off on their stale quotes. PropAMMs, a new design, emerged in 2024 attempting to combat this. In them, a single operator quotes from its own inventory and reprices actively. By 2026, propAMMs carried more than half of SOL/USDC volume on Solana and had been found, among other chains, on Base and Monad.
Our work seeks to elucidate the benefits of this new design both for LPs and for retail users. Do propAMM operators really have an edge over passive LPs, and if so, where does this edge come from? Can propAMMs offer tighter quotes to retail users? We measured a full year of propAMMs on Solana, Base and Monad, and decoded the closed-source on-chain logic of Tessera, the largest propAMM on Base.
The result: propAMMs have a measurable edge over passive LPs: two seconds after a fill, propAMMs earn 0.37 bps on Solana, 1.19 bps on Base and 1.69 bps on Monad, while passive LPs on AMMs comparatively lose 0.22, 0.62 and 2.06 bps. We trace the edge to four sources:
🟠 Cheap, frequent repricing: updates use at least 25 times less resources than swaps on Solana, and 5 to 15 times less on Base and Monad. Generally, the default block order on all three chains is by priority fee per unit of computation. Thus, an update pays far less than a swap for the same priority placement.
🟠 Pricing by counterparty: five of the six largest Solana propAMMs offer tighter spreads to aggregator flow than to non-aggregator flow. For Tessera on Base, we show that it whitelists, penalizes or blacklists individual addresses.
🟠 Arbitrage against AMMs: when a reference price moves, propAMMs reprice while AMMs do not, and arbitrageurs trade the two. A propAMM fills at its updated quote, so it may break even on these legs on Solana and Base and even earn 1.55 bps on Monad.
🟠 Spoofing: aggregators route on the quote from the end of the previous block. For Tessera on Base, we show that the operator consistently raises its default fee early in the block and lowers it again late in the block, so aggregators read stale state that carries the lower fee. As a result, only 39% of swaps execute at the quoted price, and the average swap receives 1.08 bps less.
The second question we ask is: can propAMMs offer users tighter quotes as a result of their edge? The answer is yes: on fills that arrive while the price is not moving (our proxy for retail flow) propAMMs take 0.26 bps on Solana, 1.26 bps on Base and 1.62 bps on Monad, where AMMs take 2.59, 1.38 and 8.60 bps respectively. Retail also makes up a larger share of propAMM volume: 32%, 43% and 17% on Solana, Base and Monad, against 9%, 16% and 8% on AMMs.
What did we learn? PropAMMs avoid the losses of passive liquidity provision, yet still manage to give retail tighter quotes. However, users also have to bear the cost of spoofing, which we expect to change as aggregators move to an on-chain methodology.
Link: arxiv.org/abs/2609.38056
