Madz
284 posts


@VictorTaelin @banteg thoughts? I think you manage to do multi hour autonomous coding sessions successfully (after hours of spec design?)
English

Ok, I think my experiment leaving AI working on stuff 24/7 ends here. It doesn't work. Code explodes in complexity, results are not that great, the AI can't get past hard walls (it is still completely unable to even *grasp* SupGen), and it is insanely expensive (spent ~1k over the last 2 days). The best results are on the JS compiler, mostly because it is familiar (compared to inets), but not worth losing control over the codebase.
I think the dream of having AI's working on the background and making real progress on things that matter (i.e., truly new things) isn't here yet. It is still a machine hard-stuck on its own training data, incapable of thinking out of the box. It is great for building things that were already built. But not new things
Also coding normally has the under-appreciated advantage that you're doing two things at the same time: building a codebase *and* learning it. AI's do only half of that. The other half is obviously impossible 🤔
English
Madz retweetledi

@0xLightcycle Do you use a skill to help you build your spec or just use the plan mode?
English
Madz retweetledi

Trading Fast Signals: Multi-Period Trading At Home
I wrote about my biggest W here. there won't be another article like this for the foreseeable future bcs recently I've only been getting Ls.
as usual giving access to some who RT

qm@quantymacro
yo friends I think today I might have had my biggest W at work so far… > *watching the market* > “huh that’s weird it has happened like at least 6 times in the last 2w” > *run some tests* > “damn there might be something here” > “I swear Corey has a podcast that mentions this” > he has > “I swear ceph has some tweets about this as well” > he has > there is definitely something here > *run final checks* > this thing can save us around ~8-15% of tcost, boost our Sharpe by ~15%. if we spend more time on this very likely it can be even more my PM was stunned when he saw it
English

You could chill 90% of the time, then wait for the RWA high volatility musical chairs, and you'd probably save yourself all of the drawdown that comes with overtrading anything and everything CT clowns spew at you. #freealpha
English
Madz retweetledi

The Hyper Foundation has been established to support the growth of the Hyperliquid blockchain and ecosystem. Today, the flagship application on Hyperliquid is the premier decentralised exchange, part of the foundation for a new financial system. Hyperliquid's order books already provide the deepest and most robust onchain liquidity for a wide spectrum of assets.
With the launch of the HyperEVM, any application built on Hyperliquid will also have access to this deep liquidity and other native financial primitives. For the first time, finance will be performant, liquid, and accessible.
The next step of Hyperliquid's journey is the native token, HYPE. A native token is essential for the HyperBFT proof-of-stake consensus, the HyperEVM, and further developments on the roadmap. As part of the genesis distribution, eligible users can choose to receive HYPE and optionally a Hypurr NFT to commemorate the upcoming launch of the HyperEVM on mainnet. Users must review and accept the Genesis Event Terms on hyperfoundation.org/genesis before November 11, 2024 23:59 UTC to participate.
Further information about HYPE and the Genesis Event will be shared at a later date.

English
Madz retweetledi

@PendleIntern If USDe market is implying 40% APY in ENA airdrop, and sUSDe earns the same amount of sats, doesn't it mean the current sUSDe implied APY is 'underpriced'?
sUSDe yield = underlying yield + airdrop yield = 54%
am i missing something?
English

@grandbase_fi deployer has removed their $GB liquidity and dumped it all before bridging out of base. Has the deployer been hacked?
Transfer of LP:
basescan.org/tx/0x105cd4051…
The selling:
basescan.org/token/0x2af864…
English
















