SUNDAY, JUL05
1. Node data goes public, 2. Nuclear thaws for AI, 3. Whales buy the dip, 4. The local-AI bar
From Proto and Bitkey - part of the Bitcoin ecosystem at Block, Inc.
1. moody
The developer Clark Moody has added two modules to his widely used bitcoin dashboard: a node-distribution breakdown tracking the share of nodes running Bitcoin Core, Knots, and other implementations, and a service-bits breakdown that includes utreexo and BIP-110 signaling, according to his post on X. The additions turn abstract governance debates into measurable data, letting anyone watch how node-software diversity and signaling for contested proposals actually shift over time. For a protocol-minded audience, the dashboard work matters because the Core-versus-Knots split and the BIP-110 data-carrier fight have generated far more heat than hard numbers, and public instrumentation grounds those arguments in what the network is really doing rather than what any faction claims. Making implementation share and signaling visible is a quiet public good, the same transparency ethic that lets anyone audit the chain itself. As consensus debates intensify, tools that measure rather than assert give the community a shared factual basis to reason from.2. nuclear
A Washington Post opinion piece argues that public and political opposition to nuclear power is slowly but steadily melting away, as surging electricity demand from data centers and artificial intelligence forces a reappraisal of the carbon-free baseload that nuclear provides, the paper reported. The shift reflects a broader realignment in which the computing economy’s power hunger is reviving interest in generation sources long stalled by cost, regulation, and public resistance. For an audience tracking the energy-and-compute collision, the piece captures the demand side of the story that also runs through bitcoin mining: the same appetite for steady, dispatchable power that drives miners toward stranded energy and flexible contracts is now reshaping the politics of large-scale generation. Bitcoin miners, long criticized for their energy use, have argued they can help fund and stabilize power infrastructure. A warming public attitude toward nuclear, driven by AI’s needs, reframes the broader question of who builds and pays for the generation that energy-hungry industries require.3. whales
Large bitcoin holders accumulated roughly 270,000 BTC, worth about 16.7 billion dollars, over a two-week stretch even as US spot bitcoin exchange-traded funds recorded their worst month of outflows ever in June, bleeding a record four billion dollars, according to CoinDesk citing on-chain data. The divergence, with whales absorbing supply while institutional ETF investors sold, is described as a pattern that has appeared near past cycle bottoms. For a bitcoin audience, the data is a study in holder behavior rather than a price forecast: it shows conviction concentrated among large, long-horizon holders at the same moment that the newer, ETF-wrapped institutional demand pulled back sharply. The split illustrates the different time horizons and temperaments of bitcoin’s holder base, from patient accumulators buying weakness to fund investors reacting to short-term momentum. On-chain flow data like this illuminates who is buying and who is selling beneath the headline price, without predicting where the price goes next.4. local
HashiCorp founder Mitchell Hashimoto argued that local AI models have improved dramatically over the past year, moving from what he bluntly called dogshit to good actually in roughly twelve months, but still fall short of what autonomous workflows need, in a post on X. He named frontier-class quality running locally as the threshold that would make the world spill over toward private, on-device intelligence, citing Anthropic’s Opus 4.5 as the bar to clear. For a freedom-tech audience watching the open-versus-closed AI fight, the post sharpens the stakes: incremental open-weight progress is real, yet the gap to frontier-class private compute remains the bottleneck for sovereignty-minded builders. Hashimoto has been shipping local AI tooling rather than theorizing about it, which makes the observation a builder’s benchmark rather than a venture narrative. When models good enough for real work run on hardware you control, users regain not just compute but the judgment and privacy that cloud models quietly absorb by default.Consider subscribing and sharing OP_Daily with your community.