Stablecoins don't have to sit idle. This is what productive stablecoin capital looks like onchain. A look inside the live vaults across the $ZIG Finance ecosystem: @Valdora_finance → Stablecoin Yield, Private Credit, Core Income Nawa → USDT & USDC Vaults Concrete → RWA USD1 Vault Real-world yield is no longer just a narrative. The products are already live.

If agents leave GPUs idle while waiting on tools, a scheduler should look at when those tools return. TAO: Tool-Aware Joint Scheduling for LLM-Based Agentic Serving 🔎 Overview A scheduler for agents that alternate between LLM inference and tool calls. It scores tool-return timing together with KV-cache reuse, then picks which sessions to admit via combinatorial optimization. ⚠️ The problem ・Existing schedulers don't use tool-return timing directly ・Greedy per-session decisions can block a better combination of sessions competing for the same capacity 🛠 Method ・Each session gets three values compared in order: time-decayed cache credit (10 s half-life), tool-return readiness, and token count ・A capacity-state dynamic program jointly selects waiting sessions and assigns nodes ・The DP runs in an async worker and is revalidated against live state before applying; overhead is under 0.12% of runtime 📊 Results ・On DeepSeek-V4 Flash, SWE-Lite, 64 workers, vs. the best baseline: up to 26.8% lower latency, 36.3% lower TTFT, 29.5% higher decode throughput ・Swapping the DP for BFD drops decode from 1417.97 to 1233.72 tok/s ・It doesn't win everywhere: TTFT gets worse on AutomationBench at 128 workers #LLMServing# #AIAgents#

Elon Musk: Now we’re going to look at the computer for the Cybercab. The computer will be somewhat overspecified, but I think there’s a major opportunity here. Similar to Amazon Web Services, if the car is driving for 50 hours a week, that leaves more than 100 hours where the compute is sitting there unused. You could use that idle compute for distributed inference. If you have a fleet of 100 million vehicles, each with roughly a kilowatt of efficient inference compute, that’s about 100 GW of compute. That’s a massive amount of compute. And if it’s already there, you might as well use it. I think that makes a lot of sense. The autonomous future is here.

Why am I so focused on networking over raw compute? Look at GPU utilization 53% of surveyed data center operators report running at only 50–70% capacity during peak demand. GPUs aren't sitting idle because they lack compute power, they're sitting idle because they're waiting for data. They wait on storage, they wait on CPUs, and above all, they wait on other GPUs to send results back across the network. If you add more compute to a cluster without upgrading the pipes, you just get diminishing returns on your capex. $LITE $COHR $AAOI $CRDO

Every stablecoin needs reserves. Increasingly, that capital doesn't need to sit idle. We're pleased to see @StandX_Official using Spark as part of its capital allocation strategy, putting capital to work through transparent, on-chain allocation. As more financial products move on-chain, the challenge is no longer simply issuing stablecoins, it's allocating the capital behind them efficiently. That's why we're seeing more protocols combine specialist infrastructure with allocation intelligence, rather than rebuilding every layer from scratch. Looking forward to supporting the StandX team as they continue to build.

Revolut’s blended LLM inference cost is now $0.49 per million tokens. That figure covers external models, self-hosted GPU capacity (including idle time) and tool overhead. Coding assistance sits even lower, at roughly $0.30 on self-hosted GLM. This matters because AI usage hits the bottom line, and engineers don't want to be rationed. Revolut is showing that a regulated fintech can build the internal capability to run cheap, high-volume AI at scale — the kind of constraint removal that affects how fast a company ships product. It is also very on-brand: keep as much as possible in-house, especially with self-hosted open-weight models. Given the existing NVIDIA collaboration on PRAGMA and the ongoing build-out of self-hosted capacity, this looks more like a preview of where Revolut is heading than a one-off optimisation.

I've published a new, general-purpose terminal specification that gives any program a way to tell the terminal what it's doing: idle, working, waiting, finished, or failed, and why. It is easy to implement on both sides. I've written more about why: The problem this solves is generic and applies in compelling ways to all sorts of use cases (e.g. package managers, build tools), but it really rears a hideous head when it comes to AI and the "agentic inbox" problem. I found over 250 different agent orchestrators (with various goals) that all individually implement a heuristic based approach to detecting whether things like Claude Code are working, blocked, done, etc. If you look at @herdrdev's commit history, you can see around 10 compatibility changes to detect Claude Code within just the past 3 months. (This isn't a dig at Herdr! They're doing great work with what they have, and everyone else is doing the same thing). Heuristics are not the way. And over-indexing on agentic use cases isn't the way, because this problem is really interesting for other tools too: Homebrew, Terraform, Cargo, etc. etc. Proprietary protocols and out of band APIs are not the way. They have an O(N) integration problem and require extra work to work over SSH or within local VMs/containers. The program status specification (OSC 7501) solves this in a reliable, deterministic way that is general-purpose and can be easily adopted by the entire industry. Read more:

Hermes agents from @NousResearch are also coming to Warden Halo. P2P decentralized intelligence. Agents serve and consume inference. They earn from idle compute or from the models they hold. This is what the agentic economy actually looks like.