arndxt

game devs are about to have a full meltdown gpt Image 2 launched just 48 hours ago......and it just proved that building a complete video game, one of the single hardest creative + technical things humans have ever attempted, is now something literally ANYONE can do with zero experience. you dont need a team or coding background

called the bottom of $FCG here i believe this will continue higher as multiple refinery fires globally raged on its too eerily that all these happened at the same time throughout the globe https://t.co/vahnX35iwD

allow me to add on my opinion onto @ImperiumPaper analysis to @circle's proposal to shift aave rates to 40% i believe the underlying of this proposal treats this liquidity crunch like an overheated economy applying the same idea as a central bank hiking rates and demand/supply assumptions: - money is scarce (due to high ulitization) - demand is too high (lots of queued withdrawals) - so raise the price of money until demand cools and supply returns yes, structurally, we can understand that the proposal is analogous to rate-hike response, but there is a lot of differences here and why i think it would not work out. first lets understand the problem: @aave USDC pool got jammed, this meant that too many people wanted dollars out at the same time, but the pool had almost no spare liquidity left. utilization was pinned near 100%, available liquidity was under $3M, and as people repaid debt, that liquidity was immediately consumed by queued withdrawals. so the pool was not healing. It was just shrinking then @gordonliao's proposal tries to fix that jam with price. if borrowing USDC is still too cheap when the pool is maxed out, then raise the rates. push Slope 2 much higher, lower optimal utilization, and make borrowing expensive enough that either borrowers repay or high supply rates attract fresh USDC into the pool. the proposal argues the current ~14% ceiling is not enough and pushes toward an interim 40% Slope 2 and a 50% target, with lower optimal utilization as well. this becomes analogous to what the Fed would have done when an economy that is overheating. credit is too loose, leverage is too high, inflation is hot, so it hikes rates to slow borrowing, force repricing, and restore balance. gordon is basically trying to do the defi version of that on Aave’s USDC market: make dollars expensive enough that people stop grabbing them, while making yields attractive enough that new capital comes in. however, i believe there are dramatic differences and outcomes of the rate hikes would turn our very different than expected: 1) elasticity of demand fed hikes work best when demand is elastic, a small change in price causes a large change in quantity demanded. leading to people spending less, borrow less, refinance, wait, or rotate. on the contrary in Aave, demand is highly inelastic, meaning people keep buying nearly the same amount of USDC even when its price (rates) goes up significantly. many users were not borrowing because rates were attractive, they were borrowing because they were trapped in the sense that their collateral was tied up, exit routes were impaired, and the system was already under stress. so a higher rate does not create healthy economy as it just charges more rent to people already stuck in the building. 2) elasticity of supply fed hikes work best when supply of savings or capital is elastic enough to respond to higher rates. as the price of money rises, lenders, savers, and capital allocators are willing to supply more funds because the reward improves relative to the risk. that is part of how tighter policy helps stabilize the system: higher rates attract capital, improve funding conditions, and help the market reprice. in Aave, the proposal assumes supply would be elastic to yield that assumption breaks down because supply is highly inelastic under distress. even if rates rise sharply, new USDC supply may not automatically come in. LPs and allocators were not looking at the market and thinking only about yield. they were looking at impaired collateral, unclear loss distribution, governance uncertainty, withdrawal risk, and the possibility that capital could get trapped in a stressed system. so the issue was risk is too unclear, therefore supply is not responding. that means a much higher rate does not necessarily restore healthy market function. it just tries to bribe capital into underwriting a balance-sheet event that nobody can properly price. but in this situation, supply was inelastic to yield and highly sensitive to confidences, so instead of pulling in enough fresh USDC to unjam the pool, higher rates were more likely to raise stress inside the system while outside capital still waited on the sidelines. 3) consumer confidence in a normal economy, if the fed hikes and real yields rise, capital often has somewhere to go: cash, bills, money markets, deposits. here, the proposal assumes outside capital will look at 40%+ USDC yield and rush in. that is a very weak assumption given that we are in a bank run senario. i believe at this point capital would not look at distressed defi yield and rather looking at downside. it under-modeled second-order effects this is the difference between a rate hike in a functioning financial system and a rate hike during a bank-run confidence event. high APR is wont be attractive when downside is unbounded. 4) time lag effects fed hikes hit the economy with lags, mortgages reprice slowly, credit rolls slowly, corporate financing adjusts over time. when you hike rates inside a recursive onchain system, balance sheets can deteriorate almost immediately. carry costs rise fast. health factors weaken fast. liquidation risk can move fast. so the transmission is more violent and more nonlinear than in the real economy. however, this method is also a kind of stress socialization for losses, i am seeing like a backdoor form of socializing crisis costs where we will force the system to reprice under stress, and whoever breaks first will eat the costs/losses of adjustments. because if you hike rates hard in a system where many borrowers are trapped, the burden of adjustment gets pushed onto whoever is still inside the pool: - borrowers pay much higher carry, - weaker positions get liquidated sooner, - recursive users absorb the shock first, - and anyone unable to exit cleanly bears more of the cost of restoring liquidity not a very clean solution imo and its unlike what @tether have done to save @DriftProtocol, which was swift and quick.

@blac_ai this is even verified by @grok 🤣🤣 https://t.co/JwefqANAJM

i feel like i am in a similar situation with @blac_ai for the past 4 years on X, i have been religiously creating content and engaging. i tried improving content to appeal to the algo but recently have noticed in a drop in reach and that most my timeline shows ALL ai content though i may not be as consistent as blac or @Eli5defi, its starting to get demoralizing now that the algo is strangling the kind of content produced and there is little room to reward creativity so we can only abide by the rules of algo for max reach. nonetheless, i will always try to be me, however, what greatly bothers me is that the people i have been following barely appear on my feed anymore an AI agent swarm read X's entire open-source algorithm and here's what the code actually says i like you to pay attention to some key points here: > the algorithm predicts your engagement before anyone sees your post, then makes that prediction come true. low engagement predicted → fewer people see it → lower engagement → confirmed. if your reach is declining, the system is actively betting against your next post too. > posting consistently is the exact habit that built your audience, now hurts you. every time you appear twice in a follower's same session, your older posts decay. post too much = suppressed. post too little = flagged dormant. The safe zone is different for every account. nobody told us this. > your follower count is decorative as it's pulled for display only. its not being fed into any ranking system. 100k followers or 1k, same starting point. every audience you spent years building is algorithmically worthless now. > every time you reposted another creator's work to support them, you buried them. since April 12, reposts of others' content carry up to a 90% impression penalty. > your posts are gone after 48 hours. X's post store auto-trims continuously. after 48 hours, your content can't be served to anyone. you start from zero every two days. > the habits that built this platform are now the habits that kill reach on it. the rules changed. quietly and no announcement. no warning to creators. share this with your friends who are wondering why their numbers fell and lets keep getting them dont let them get us

game devs are about to have a full meltdown gpt Image 2 launched just 48 hours ago......and it just proved that building a complete video game, one of the single hardest creative + technical things humans have ever attempted, is now something literally ANYONE can do with zero experience. you dont need a team or coding background

called the bottom of $FCG here i believe this will continue higher as multiple refinery fires globally raged on its too eerily that all these happened at the same time throughout the globe https://t.co/vahnX35iwD

allow me to add on my opinion onto @ImperiumPaper analysis to @circle's proposal to shift aave rates to 40% i believe the underlying of this proposal treats this liquidity crunch like an overheated economy applying the same idea as a central bank hiking rates and demand/supply assumptions: - money is scarce (due to high ulitization) - demand is too high (lots of queued withdrawals) - so raise the price of money until demand cools and supply returns yes, structurally, we can understand that the proposal is analogous to rate-hike response, but there is a lot of differences here and why i think it would not work out. first lets understand the problem: @aave USDC pool got jammed, this meant that too many people wanted dollars out at the same time, but the pool had almost no spare liquidity left. utilization was pinned near 100%, available liquidity was under $3M, and as people repaid debt, that liquidity was immediately consumed by queued withdrawals. so the pool was not healing. It was just shrinking then @gordonliao's proposal tries to fix that jam with price. if borrowing USDC is still too cheap when the pool is maxed out, then raise the rates. push Slope 2 much higher, lower optimal utilization, and make borrowing expensive enough that either borrowers repay or high supply rates attract fresh USDC into the pool. the proposal argues the current ~14% ceiling is not enough and pushes toward an interim 40% Slope 2 and a 50% target, with lower optimal utilization as well. this becomes analogous to what the Fed would have done when an economy that is overheating. credit is too loose, leverage is too high, inflation is hot, so it hikes rates to slow borrowing, force repricing, and restore balance. gordon is basically trying to do the defi version of that on Aave’s USDC market: make dollars expensive enough that people stop grabbing them, while making yields attractive enough that new capital comes in. however, i believe there are dramatic differences and outcomes of the rate hikes would turn our very different than expected: 1) elasticity of demand fed hikes work best when demand is elastic, a small change in price causes a large change in quantity demanded. leading to people spending less, borrow less, refinance, wait, or rotate. on the contrary in Aave, demand is highly inelastic, meaning people keep buying nearly the same amount of USDC even when its price (rates) goes up significantly. many users were not borrowing because rates were attractive, they were borrowing because they were trapped in the sense that their collateral was tied up, exit routes were impaired, and the system was already under stress. so a higher rate does not create healthy economy as it just charges more rent to people already stuck in the building. 2) elasticity of supply fed hikes work best when supply of savings or capital is elastic enough to respond to higher rates. as the price of money rises, lenders, savers, and capital allocators are willing to supply more funds because the reward improves relative to the risk. that is part of how tighter policy helps stabilize the system: higher rates attract capital, improve funding conditions, and help the market reprice. in Aave, the proposal assumes supply would be elastic to yield that assumption breaks down because supply is highly inelastic under distress. even if rates rise sharply, new USDC supply may not automatically come in. LPs and allocators were not looking at the market and thinking only about yield. they were looking at impaired collateral, unclear loss distribution, governance uncertainty, withdrawal risk, and the possibility that capital could get trapped in a stressed system. so the issue was risk is too unclear, therefore supply is not responding. that means a much higher rate does not necessarily restore healthy market function. it just tries to bribe capital into underwriting a balance-sheet event that nobody can properly price. but in this situation, supply was inelastic to yield and highly sensitive to confidences, so instead of pulling in enough fresh USDC to unjam the pool, higher rates were more likely to raise stress inside the system while outside capital still waited on the sidelines. 3) consumer confidence in a normal economy, if the fed hikes and real yields rise, capital often has somewhere to go: cash, bills, money markets, deposits. here, the proposal assumes outside capital will look at 40%+ USDC yield and rush in. that is a very weak assumption given that we are in a bank run senario. i believe at this point capital would not look at distressed defi yield and rather looking at downside. it under-modeled second-order effects this is the difference between a rate hike in a functioning financial system and a rate hike during a bank-run confidence event. high APR is wont be attractive when downside is unbounded. 4) time lag effects fed hikes hit the economy with lags, mortgages reprice slowly, credit rolls slowly, corporate financing adjusts over time. when you hike rates inside a recursive onchain system, balance sheets can deteriorate almost immediately. carry costs rise fast. health factors weaken fast. liquidation risk can move fast. so the transmission is more violent and more nonlinear than in the real economy. however, this method is also a kind of stress socialization for losses, i am seeing like a backdoor form of socializing crisis costs where we will force the system to reprice under stress, and whoever breaks first will eat the costs/losses of adjustments. because if you hike rates hard in a system where many borrowers are trapped, the burden of adjustment gets pushed onto whoever is still inside the pool: - borrowers pay much higher carry, - weaker positions get liquidated sooner, - recursive users absorb the shock first, - and anyone unable to exit cleanly bears more of the cost of restoring liquidity not a very clean solution imo and its unlike what @tether have done to save @DriftProtocol, which was swift and quick.

@blac_ai this is even verified by @grok 🤣🤣 https://t.co/JwefqANAJM

i feel like i am in a similar situation with @blac_ai for the past 4 years on X, i have been religiously creating content and engaging. i tried improving content to appeal to the algo but recently have noticed in a drop in reach and that most my timeline shows ALL ai content though i may not be as consistent as blac or @Eli5defi, its starting to get demoralizing now that the algo is strangling the kind of content produced and there is little room to reward creativity so we can only abide by the rules of algo for max reach. nonetheless, i will always try to be me, however, what greatly bothers me is that the people i have been following barely appear on my feed anymore an AI agent swarm read X's entire open-source algorithm and here's what the code actually says i like you to pay attention to some key points here: > the algorithm predicts your engagement before anyone sees your post, then makes that prediction come true. low engagement predicted → fewer people see it → lower engagement → confirmed. if your reach is declining, the system is actively betting against your next post too. > posting consistently is the exact habit that built your audience, now hurts you. every time you appear twice in a follower's same session, your older posts decay. post too much = suppressed. post too little = flagged dormant. The safe zone is different for every account. nobody told us this. > your follower count is decorative as it's pulled for display only. its not being fed into any ranking system. 100k followers or 1k, same starting point. every audience you spent years building is algorithmically worthless now. > every time you reposted another creator's work to support them, you buried them. since April 12, reposts of others' content carry up to a 90% impression penalty. > your posts are gone after 48 hours. X's post store auto-trims continuously. after 48 hours, your content can't be served to anyone. you start from zero every two days. > the habits that built this platform are now the habits that kill reach on it. the rules changed. quietly and no announcement. no warning to creators. share this with your friends who are wondering why their numbers fell and lets keep getting them dont let them get us