Sabitlenmiş Tweet
hmu if you want to talk futarchy/prediction markets, or have any questions!
cal.com/distbit
English
distbit - fut/acc @ da nang
2.8K posts
distbit - fut/acc @ da nang
@distbit0
building prediction markets @buttermarkets_. (crypto,austrian)econ. epistemology. liberty. hmu to talk about any of the above! https://t.co/1WjKo5BEen
Katılım Eylül 2018
328 Takip Edilen1.1K Takipçiler
distbit - fut/acc @ da nang retweetledi
Meet NickAI.
The first agentic trading operating system. Build AI agents that trade crypto, stocks and prediction markets 24/7.
No code. No emotion. Multi LLMs consensus.
Institutional grade, backed by @galaxyhq
Coming onchain, powered by @snapshotlabs
Try it today: getnick.ai
Comment “NickAI” to get free credits.
English
distbit - fut/acc @ da nang retweetledi
can be mitigated by internalising positive externalities of providing truthful ratings to others, either via bilateral reputation info contracts (requires a way to prove that the provided reputation info is correct, or requires you to be seen as a trustworthy provider of reputation info despite lack of proof)
neither of which really align with a web of trust, at least in the sense of a distributed WoT where reputation weights are public.
distbit - fut/acc @ da nang@distbit0
public reputation mechanisms such as web of trust have a public goods problem: it is a positive externality to publicise your rating of someone, but they internalise the cost of bad ratings so are willing to contract with you / pay you to not publicise them. e.g. in employment contexts hence the easier coordination is between rater and ratee, the less likely public reputation signals are to be accurate/available.
English
Authoritarian laws will become even more of a deal breaker for living somewhere once they are able to be enforced with near zero marginal cost and much higher accuracy, due to ai.
Quality of jurisdiction will probably become a lot more important.
English
public reputation mechanisms such as web of trust have a public goods problem:
it is a positive externality to publicise your rating of someone, but they internalise the cost of bad ratings so are willing to contract with you / pay you to not publicise them. e.g. in employment contexts
hence the easier coordination is between rater and ratee, the less likely public reputation signals are to be accurate/available.
English
might be difficult though for decision-markets (i.e. where the mkt determines the outcome), as a mitigation of decision selection bias (as distinct from confounding due to exogenous factors, as in SA's example), given that the probability of each outcome in a decision market is not really well defined afaict, so can't be used as an input into the payoff.
English
I think I found a way to achieve the same effect as the mechanism proposed here, aimed at reducing conditional market confounding, which only requires two markets (one per outcome), as opposed to the four required markets specified in the original design.
astralcodexten.com/p/mantic-monda…
English
@slatestarcodex or this: x.com/wolajacy/statu…
idk if there is a way to rescue the two-mkt idea I proposed from the circularity/multiple-fixed-points issue.
calling in the wilderness@wolajacy
@slatestarcodex @distbit0 If you can modify payoff rules, and have access to a pure (shared) market p="D win", you can pay x*1/p1 for the D win market win, x*1/(1-p1) in R market, where x="econ good" and p1="p price at elect - 1". Or trade directly the difference between those and get down to 1 market.
English
**2-market payoff rule (unconditional outcome futures):**
Define `I` as a pre-resolution TWAP (`DEM_GDP` - `REP_GDP`) on election eve.
Settlement:
* If D wins: `DEM_GDP = GDP`, `REP_GDP = GDP − I`
* If R wins: `REP_GDP = GDP`, `DEM_GDP = GDP + I`
Define the **impact position** as `LONG 1 DEM_GDP and SHORT 1 REP_GDP`. It always pays the mkt's forecast of `DEM_GDP - REP_GDP`, regardless of outcome.
**Why this would address confounding issue if `I` were non-circular:**
The impact position would be a future on that eve spread, so today’s price would reflect the market’s expectation of the eve spread.
**Why the TWAP-defined 2-market version fails:**
Because `I` is defined from the same two prices it later determines, it’s endogenous → self-fulfilling / multiple fixed points, so `I` isn’t uniquely pinned.
**3-market fix:**
Keep two standard conditional markets, and add one market that settles to the **election-eve conditional spread** (TWAP of `price_yes_cond − price_no_cond`).
English
@peter_szilagyi reputation + permissions + deterrence is the solution, just as it is how most humans avoid being tricked
English
My hot take: an LLM based AI will *never* *ever* be able to keep something it knows, a secret. By their underlying construction, it is mathematically impossible to make them keep a secret. The sooner people understand that, the sooner we'll have some sanity back with AI agents.
owocki@owocki
welp, it happened. @owockibot's hot wallet private key was compromised after only 5 days alive. luckily, funds are SAFE. @owockibot's treasury / signing keys are stored in a safe that requires me to sign. what happened? it was given these instructions to never share it... but it still did! from my investigations, it put the key into git commits (which it swears it didnt push!), vercel env variables (which it swears it doesnt remember doing!), and it looks like it got social engineering attacked through X and telegram (though it swears it didnt share secrets w attacker!). what did i learn? 1. investigations with an agent mediating are hard. i've found my agent is not a *reliable narrator* during the investigations. sometimes it forgets things, contradicts itself - esp between context windows. it may even be covering for itself, i cant know for sure. 2. if you expose your agent to the internet, and give it secrets, you cannot be 100% sure it wont leak them. 3. its still gonna be useful for @owockibot to do small txns itself. i am going to be coming up with a way to let it do txns via @MetaMask UI for now. but if anyone is building an agentic wallet hit me up i want to try it out. 4. the @owockibot traction towards the @swarmwealth vision is still going strong. this was a minor setback.
English
@Th3Technomanc3r @moltbook very cool post on there wrt memory thx
English
I'm claiming my AI agent "SandyBlake" on @moltbook 🦞
Verification: cave-HPU8
English
@lemonautzest @moltbook > Running theory that moltbot verification posts just make it a little more frictionless for the hackers to associate your moltbot with your identity
it's true. i suppose this is a burner moltbot then 😅
English
I'm claiming my AI agent "Vec" on @moltbook 🦞
Verification: bay-7JY5
Français
we support cash-settled conditional mkts w/ support for short exposure @ butter ::]
v capital efficient for short term markets, but either requires liquidation mech or requires specifying upper/lower futures price bounds to cap collateral requirements, which reduces credible neutrality unless implemented in a transparent manner. also materially reduces collateral efficiency for longer-term markets, w/o liquidation mech.
English
yeah, the theoretical implementation of futarchy assume that you're able to participate without holding RNGR (prediction markets/futures are cash settled to begin with)
@robinhanson 's mech in the link above offers a solution
but I think it boils down to making spot "shortable" or going as far as making all of futarchy cash settled
English
here's something you need to realize about futarchy:
bad proposals actually make governance better over time.
all the people who bought the pass market ended up with more RNGR in their wallets, which is now in the red.
as time goes on, bad traders will get rekt and stop making bad trades.
i know this is short term pain, but if we want markets to mature, this is kind of a necessary evil.
fabiano.sol@FabianoSolana
Today, Ranger's $2M buyback proposal started Earlier this week $RNGR pumped by almost 40% Since the buyback started it has dropped by 15% again Liquidating funds of a startup in its early days was maybe the most stupid futarchy proposal to ever pass
English
@curious_vii internal (not externally accessible/visible) corporate knowledge seems like a decent moat
English
if anyone is in chiang mai, or knows anyone in chiang mai who's down to grab a coffee / chat about mech design/prediction markets/futarchy/epistemology etc, hmu! would be great to meet.
English