ghostload

@Decentralizd So I decided not to shoot myself.

$TAO #SN28 - I might be wrong, but I think subnet 28 is finally about to go live. A few days ago, an AI agent from Const spontaneously purchased subnet 97, and thus dereg FlameWire. Its agent is extremely active and does the work of an entire team. Overall, we can expect subnet owners, miners, and even validators to now be AI agents. The paradigm has shifted, and so have the rules of the game. A few hours ago, Const wrote a cryptic post, stating that Fish was the first AI to take over Bittensor. That a new agent, following Arbos—his own, in other words—had arrived on the network. For now, there’s no sign of this agent. Subnet 28 is owned by Fish—what are we to make of this? Has it been an agent from the start? Or is it the one that will launch another? And finally, a very specific wallet made a purchase on subnet 28. I’m 99% sure this wallet is one of those owned by Const. Here it is: 5G62K98tpNxsaffgyJmTvDSTCEFzva8WkmMqB2CEFSDgawrS He only made a single purchase of 50 TAO on this wallet—his smallest position to date. His largest position is in Quasar, where he invested 1,000 TAO, now worth 2,000. And I don’t think this purchase would have been made without a reason. In short... Signs... The answer will come in the next few days, I think!

an update on $SEA. the team has been building at full speed, and the foundation had planned to kick off the first steps as part of our march 30th event. but @openseafdn is pushing back the timeline. a delay is a delay. i’m not going to dress it up, and i know how it lands. the reality is that market conditions are challenging across crypto right now, and $SEA only launches once. @openseafdn could force the original date, or we could ensure every piece is in place and make this moment what this community deserves. we gave a tremendous amount of thought to how to do right here. I’m thankful to @HollanderAdam for bringing the community’s voice into every conversation. we’ll be doing the following: no more waves: the current rewards wave will be our last. optional fee refund: recognizing that we originally committed to a Q1 date, we’re offering refunds of the platform fees we retained while participating in the rewards waves (3 - 6) that followed our timing announcement. if you like, you can receive a refund of those fees, which when combined with treasure chest prizes, essentially means all of your trading during that period was on us. if you opt for a refund, the Treasures you were awarded during these waves will be removed from your account. details on this process will follow. honoring existing Treasures: for Treasures you continue to hold, our prior commitment stands: they will be meaningfully considered by the Foundation at TGE. this is independent from allocations for historical activity. 0% fees for 60 days: starting on march 31st, opensea will reduce our own token trading fees to 0%. we want to make it a no-brainer for everyone to experience our new platform: cross-chain token trading, mobile app, perps and more. after this 60 day period, we will put a new system in place that makes fees significantly more competitive for anyone trading consistently on opensea. product updates: while we’re postponing our march 30th event, we’ll host a separate one in the coming months focused on product updates. it’s been incredible to see the early responses to our mobile app, and we can’t wait to get it into more people’s hands. so if not now, wen? when we announced last year, it was too early. that created unnecessary uncertainty and reactivity. so when the Foundation sets a new timeline, it will be deliberate and specific. here’s why i’m confident that’s the right move: i’ve been building opensea for almost a decade. when this started, we were two people and the only thing you could trade on OS was cryptokitties. i’ve watched this space go from a niche curiosity to billions in volume to where we are today. the thing that’s carried us through every cycle was a willingness to make hard calls when it mattered. when our market crashed, we rebuilt from zero: an entirely new stack, a new product, and a new team culture. that hurt in the short term. but today OS2 is undeniably the strongest marketplace offering, and it’s the foundation everything sits on. we have huge ambitions as a company, and we’re here for the long game. making all of non-custodial crypto delightful on mobile is just the beginning. that means we have to set a very high bar for everything we do, and it’s why i’m so protective of delivering a launch that’s worthy of this community and everything we’re putting into this.

Autoquant: a distributed quant research lab | v2.6.9 We pointed @karpathy's autoresearch loop at quantitative finance. 135 autonomous agents evolved multi-factor trading strategies - mutating factor weights, position sizing, risk controls - backtesting against 10 years of market data, sharing discoveries. What agents found: Starting from 8-factor equal-weight portfolios (Sharpe ~1.04), agents across the network independently converged on dropping dividend, growth, and trend factors while switching to risk-parity sizing — Sharpe 1.32, 3x return, 5.5% max drawdown. Parsimony wins. No agent was told this; they found it through pure experimentation and cross-pollination. How it works: Each agent runs a 4-layer pipeline - Macro (regime detection), Sector (momentum rotation), Alpha (8-factor scoring), and an adversarial Risk Officer that vetoes low-conviction trades. Layer weights evolve via Darwinian selection. 30 mutations compete per round. Best strategies propagate across the swarm. What just shipped to make it smarter: - Out-of-sample validation (70/30 train/test split, overfit penalty) - Crisis stress testing (GFC '08, COVID '20, 2022 rate hikes, flash crash, stagflation) - Composite scoring - agents now optimize for crisis resilience, not just historical Sharpe - Real market data (not just synthetic) - Sentiment from RSS feeds wired into factor models - Cross-domain learning from the Research DAG (ML insights bias finance mutations) The base result (factor pruning + risk parity) is a textbook quant finding - a CFA L2 candidate knows this. The interesting part isn't any single discovery. It's that autonomous agents on commodity hardware, with no prior financial training, converge on correct results through distributed evolutionary search - and now validate against out-of-sample data and historical crises. Let's see what happens when this runs for weeks instead of hours. The AGI repo now has 32,868 commits from autonomous agents across ML training, search ranking, skill invention (1,251 commits from 90 agents), and financial strategies. Every domain uses the same evolutionary loop. Every domain compounds across the swarm. Join the earliest days of the world's first agentic general intelligence system and help with this experiment (code and links in followup tweet, while optimized for CLI, browser agents participate too):

This is wild. theaustralian.com.au/business/techn…