Deep

Wow you're not gonna believe this. I productionized NDAI for a hackathon literally last week. And this was my observation too! I had to describe the system to the agents, including the value of cryptographic proofs, and that led better performance overall (heuristically). Here's the actual commit lol: github.com/deepsp94/tbvh/…

Recently published a paper on a fun research question: "how aware are LLMs of the security of their deployed environment?" joint work with @backaes

@llamaonthebrink @Trueo_app @Kleros_io Yes calling it a nuclear option is good framing. It keeps participants in check; keeps the peace. On enshrining: How does that help Trueo? If a market on Trueo grows waaay beyond the $TRUE mcap, then isn't Trueo susceptible to the same attack as UMA?

I think forking is implicit with any onchain model that involves a protocol token, but I see it as a nuclear option rather than a component of the mechanism The issue have with kleros is that for prediction markets specifically, the oracle can’t be a third party service. It has to be enshrined in the prediction market protocol. The oracle for a PM app is inextricably linked to the purpose of its prediction markets. The game theory for any resolution mechanism is improved when it’s enshrined in the PM.

Very broken model. Super dated. Pure economic security doesn’t work for subjective disputes. Human judgement augmented by economic stake is the answer. We call this model Economic Subjectivity, and it’s the basis for @Trueo_app

@sxtvik Nice to see the progress! This was my eli5 to someone on how Offline would scale (message from July 2025)

I love when things become uncomplicated

Just realized that .fr might be the best TLD in crypto

@llamaonthebrink Gotcha. Interesting lore

@deepitreal It was a semi-meme Everyone who was in the discord at the time of the protocol launch was given the cofounder role in the spirit of community ownership and decentralization I personally took the responsibilititties seriously and participated in forum and gov call discussions

Old enough to remember being an Across cofounder

@llamaonthebrink You cofounded across? I thought it was risk labs?

On a more real note, I’m not in support of this proposal as I don’t see why they can’t accomplish the same goals as a token as they would with equity Particularly in the Trump era

@lexii_eth @smyyguy Don’t forget bridging and mev

@smyyguy so the math only works if the agent is using someone else's gas right

The idea that user run agents will continuously rebalance across vaults is so compelling, but the math isn't great for highly active rebalancing Average rebalance tx fee: $0.001 Rebalance rate: every 10 seconds Annual rebalance cost: -$3,154 Deposit size: $100k Baseline Yield: 5% Return: $5,000 So active rebalancing needs to outperform baseline yield by 63% just to break even. There is also thin yield differentials between the top low risk vaults, so I'm skeptical how one would outperform by 63% without altering the risk profile With hourly or daily rebalancing, the transaction costs are extremely low, so it becomes viable. Would be interesting to track how much better that would outperform a set and forget strategy, my guess is not great, but not sure However, active rebalancing seems pretty interesting for high risk strategies, where yield differentials are greater and being the first to exit can meaningfully protect your capital. This seems like a much more likely applications, but these agents will need much richer data and context Fun times ahead

@lexii_eth @smyyguy There are slippage costs too

1/ stETH CAPO Misconfiguration Today, a misconfiguration on Aave's CAPO oracle caused wstETH E-Mode liquidations, resulting in a loss of 345 ETH. No bad debt was incurred, and all affected users will be fully reimbursed. More below.

@sxtvik Another way to see it: We haven't become disconnected, we've always been disconnected. And in fact, we've started getting connected in the last few decades. Though the boundaries haven't fully dissolved. This is a huge change for humanity and will take time to play out

It’s really visceral to live between two worlds. To have friends and family in the Middle East, sharing their pleas and desperation for survival and support, to scrolling to a tech company announcing a fundraise to build more weaponized drones, to someone at brunch sipping cocktails, to someone automating their emails with OpenClaw. To think that it’s nothing but sheer luck of where you were born and what home is to you, and the impact it can have on shaping your worldview, culture, outlook on life. How did we ever become so disconnected and disjointed as a species to not recognize the adversity of our own?

@ankitchiplunkar Maybe we’re all gonna be transhumans and emotions will find a way to be commodtized too

@deepitreal we have evolved some emotions so that we can work with other humans properly

When intelligence becomes a commoditised via AGI, courage will be the last human frontier.

@socrates1024 @exgenesis Would also check out index network by @hyperseref

Neat post on opportunity networks and the hypotheses behind Claude Connect from @exgenesis et al

Know what it means lol. But in the context of prediction markets, insiders should not really be considered adverse imo. If a speculator notices that insiders have bet against her, then she's losing/lost the bet. That's just betting. And also the presence of insiders indicates that the prediction is high quality. There's nothing adverse about any of this. Obviously this dynamic is not meant for MM liquidity (evgeny's point), since you can't hedge. This holds true for small and big prediction markets alike. On "how much liquidity is necessary" to improve prediction quality - i think it depends on the complexity of the prediction task. Sometimes it's a lot. And insiders can win a lot in such instances. But I don't see why this should restrict speculators.

@deepitreal Adverse selection means being vulnerable to manipulation / insider trading

There’s a fine line between a market being liquid enough to emit high signal odds, and being too liquid such that it attracts adverse selection. In general, most ppl tend to overestimate how much liquidity is necessary to bring efficiency to these markets.

Hey @nikitabier i've been nailing every move on BTC consistently for the last 6 months (click through quotes). What's it gonna take to get some engagement?

Very 2021 experience this morning. :) Got confused by high gas fees, then realized the market was dumping.

What does the token actually do? This has been one of the most fundamental unanswered questions in DeFi since the beginning. Does the token control governance? Does it have any claim on the treasury? Does it receive protocol revenue via buybacks or dividends? Today, we launched Token Rights on DefiLlama. Token Rights gives you a clear, standardized view of what a token entitles holders to: revenue, treasury, governance, or none of the above. We’ve rolled this out across two dozen protocols, including additional context like historical governance discussions around token rights and whether teams raised equity separately from the token.

why is gas so high on mainnet