0xtenac

sometimes you have to crashout on the ai

.@ritualnet going live on public testnet is the next major step of creating cryptographically verified autonomous systems what can be done on Ritual is far higher than "agentic payments" larp we've been seeing for 2 years already

@tonbistudio it's not, they are: 1. running out of compute 2. servicing mythos to enterprise clietns - siphoning compute from 4.6/4.7 to serve enterprise users

Idk I've been working with 4.7 all day and it doesn't seem any better

the modern man

@farmer_bocchi add GPU financing, it's all so tiresome

actual state of defi currently

are we playing to not lose? or we playing to win mfers.

@behrin @Neutrl @pendle_fi @strata_markets respect the work you are doing do you have any intention of expanding beyond the current otc + hedge strategy you have built or can this trade be further scaled with minimal diminishing returns

There are a few liquid yield opps in the Neutrl Eco right now. 1. Staking sNUSD 7-8% @Neutrl 2. Neutrl @pendle_fi pools have been seeded with $35M offering >8% on $10M shots 3. Strata who have tranched Pendle 6/9% sNUSD plus Strata x Neutrl points @strata_markets 4. Royco just released a sr/jr tranche and market is scaling up 5-9% @roycoprotocol 5. @monad / @withAUSD incentivised Neutrl vault on @AccountableData targeting +9% These are all short term and sNUSD and Neutrl sNUSD very liquid for >$5-10M.

@BittexXBT @CL207 we're cooked arent we...

@CL207 idk why everyone was yapping about quantum for weeks but not how vulnerable DeFi is when everyone can use AI to mass extract

this is great, there is simply way way way too much money onchain than people realize and it better be safe 🙏

Having problems? You should try winning

yield in on-chain venues are quite mid, not only have they compressed significantly but rational capital managers need to underwrite the risk of: - smart contract risk - economic risk - operator risk just to get 4.5% YTD when you could buy STRC off a primer broker for 11%

gemma-4 is great model i do think using this base model to fine-tune for domain specific heuristics and then leveraging skills in tandem with auto-research and a clear objective function can produce quality results not found in frontier