A new paper out of KRICT and KAIST just put a number on something Qubic's architecture has been built around since 2022. Multiple AI agents working in coordination reduce extrapolation error by up to six orders of magnitude versus isolated deep neural networks. Up to a million model parameters compress into 5 to 40 interpretable ones. Tested across deterministic and previously uncharacterized dynamics. The collective recovered the underlying governing equations across all of them. @c___f___b  described the design premise of Qubic's coordinated mining layer this way: "The work of one miner benefits from the work of another miner, and their combined work is greater than the sum of their works measured separately." That is not a metaphor. The MCI paper is empirical confirmation of the mechanism. The researchers tested on a small lab cluster and noted the obvious next question: what happens when this runs at network scale? Qubic's 676 Computors and broader mining base sit directly on the other side of that question.  Coordinated, network-scale machine intelligence is what Aigarth is being built for. The frontier is catching up to the foundation.

🚀 The Agentic Economy is coming. But where does Qubic fit in? Join @AEON_Community × @_Qubic_ for an exclusive X Space! We’ll dive into AEON’s vision for the #AgenticEconomy#, what’s next on the #AEON# roadmap, how #Qubic’s# decentralized intelligence could fit into this emerging narrative, and where #AI#, decentralized networks, and payments are heading next. 🎁 $200 Reward Pool $100 for 20 winners who follow @AEON_Community & @_Qubic_ + $100 for 5 selected questions about #Qubic# 📅 Sept 17 | 13:00 UTC 🔗 Join the X Space: Drop your questions, the proof of following and qubic-network:native address below! 👇

Today we debut our 256 physical qubit system: IonQ Superion 256. Our company has evolved through many stages to get to this day: 🔸 From Dr. Chris Monroe and Dr. Jungsang Kim’s foundational trapped ion breakthroughs 🔸 To integrating trapped ions on semiconductor chips 🔸 To adding networking, sensing and security technologies 🔸 To releasing our Walking Cat architecture 🔸 And to adding U.S.-based quantum manufacturing capabilities that underpin the entire quantum ecosystem It’s the first generation in our new line of quantum computers. Read the full announcement → #IonQ# #QuantumIsNow# #QuantumComputing# #WalkingCat# #FTQC#

Today we debut IonQ Superion 256, our 256 physical qubit system, and the IonQ Superion product line. Developed in partnership with SkyWater Technology, it operationalizes our Walking Cat architecture for full fault-tolerant quantum computing, the blueprint for scaling beyond 10,000 qubits by 2028. Watch Chris Ballance share the full view in our livestream today, September 8, at 12:30 PM ET → Read the full announcement → #IonQ# #QuantumIsNow# #QuantumComputing# #WalkingCat# #FTQC#

$IONQ EC (2) IonQ recognized $4.3M(Q2) and $11.3M(1H) from Chicago Uni. The 256-qubit system has now moved beyond the design stage and reached the fully integrated QPU prototype stage. The 10,000-qubit chip is already being developed in parallel, with tape-out work underway. SkyWater will remain a merchant foundry serving other quantum companies, rather than becoming a fab dedicated exclusively to IonQ. An upgrade pipeline is beginning to emerge in the field, with Tempo customers potentially progressing to 256-qubit and eventually 10,000-qubit systems.

U.S. GOVERNMENT BACKS $RGTI & $QBTS WITH UP TO $200M FOR QUANTUM COMPUTING R&D Rigetti has signed a definitive $100M agreement with the U.S. Department of Commerce to tackle key bottlenecks in scaling superconducting quantum systems, including cryogenics, readout electronics and high-connectivity chip fabrication. D-Wave separately finalized access to up to $100M to advance its annealing and gate-model platforms, including development of a 100,000-qubit annealing system and a 10,000-qubit gate-model system targeting 100 logical qubits. As a condition of the funding, the U.S. Department of Commerce will receive minority, non-controlling equity stakes in both companies.

A TON OF THINGS HAPPENED IN THE STOCK MARKET TODAY. Here's a full recap: 1. Palantir $PLTR named Nebius $NBIS its preferred sovereign AI infrastructure partner, announcing a strategic partnership to integrate Nebius compute and inference endpoints directly inside Palantir’s enterprise perimeter. The integration will allow eligible Palantir commercial customers to run open AI models on Nebius infrastructure, continuously adapt them using proprietary company data, and maintain control over their compute, models, and data. The companies also plan to bring new AI compute capacity online faster, including modular data-center deployments in locations where power is already available. Palantir CEO Alex Karp said, “Nebius’ compute infrastructure powers your ability to run your own AI models under conditions you control. Our ontology and their infrastructure will undergird the sovereignty our partners are demanding.” Nebius CEO Arkady Volozh added that the partnership will help commercial clients run optimized open models on trusted infrastructure while maintaining control over their data and models. 2. The EU and Canada are reportedly set to announce a major “all-encompassing” partnership spanning trade and security after President Trump’s latest tariffs, according to Bloomberg. The goal is to build a new alliance that helps offset global power politics dominated by the U.S. and China. The partnership is expected to be announced on September 16, with Canada and the EU reportedly aiming to get “as close as legally possible” short of EU membership. Canada has also become the first country outside the EU to join the bloc’s $175B military procurement fund, underscoring how quickly trade tensions are pushing allies toward deeper economic and defense cooperation. 3. BlackRock’s Bitcoin ETF $IBIT has attracted $3.7B of inflows so far this quarter, putting it on pace for its largest quarterly intake since Q3 2025. The fund has already added $459.8M in September inflows, following $3.0B in August, its biggest monthly inflow since October 2025. As a result, $IBIT AUM has climbed to $62.6B, near its highest level since mid-May. Since the start of July, the fund’s AUM has increased by $19.6B, or 46%, showing renewed institutional and retail demand for Bitcoin exposure through ETFs. 4. OpenAI CFO said at the Goldman Sachs conference that the consumer business is accelerating, with July revenue up 20% MoM. The enterprise business grew even faster, with July revenue up 32% MoM. OpenAI’s revenue mix is now roughly 50/50 between enterprise and consumer, showing that growth is broadening beyond ChatGPT subscriptions and into business adoption. 5. Schwab retail activity in August shows investors selling the software rip and buying the semiconductor dip. The top retail buys were $SPCX, $MU, $NVDA, $INTC, and $GOOGL, while the top sells were $NOW, $ADBE, $MSFT, $PLTR, and $ORCL. The shift points to more active portfolio management from retail investors, with money rotating out of crowded software winners and into AI, chips, and compute-related names after pullbacks. 6. Copper prices on the LME have surged to a record $14,533/ton, now up 47% over the last 12 months. The move is being driven by tight supply and rising demand from data centers, renewable energy, and power grids. Expectations that President Trump could expand tariffs to refined copper imports have also pushed U.S. Comex prices above global prices, creating an incentive for traders to ship hundreds of thousands of tons of copper into the U.S. to capture the price spread. As a result, global stockpiles are becoming increasingly concentrated in the U.S., draining LME inventories to critically low levels and leaving less readily available copper for buyers elsewhere. 7. Meta $META launched Muse, a personal AI agent that can take actions across apps and the web, including sending emails, booking travel, filling out forms, and making purchases with user approval. Muse runs on Meta’s new Muse Secure VM, giving each user a dedicated cloud environment with a separate Sentinel agent that reviews actions before they reach the internet. The agent can continue working after the app is closed, remember user preferences, and proactively move longer-term tasks forward. Meta is also integrating Stripe Link for checkout, with Shop Pay and 1Password support coming later. 8. IREN $IREN said its 2 GW Sweetwater Hub has been conditionally included as base load in ERCOT Batch Zero. Sweetwater 1 accounts for 1.4 GW and Sweetwater 2 for 600 MW, both part of IREN’s broader 5 GW+ global data-center development pipeline. At Sweetwater 1, the high-voltage substation is already energized, with 300 MW of gross data-center capacity under construction and targeted for delivery in Q4 2027. IREN also said additional large-scale projects across its development pipeline have been included in Batch Zero, though ERCOT classifications remain conditional and subject to further approvals. 9. The top 10 most active options today by contracts traded were $TSLA with 2.4M contracts, $NVDA with 2.2M contracts, $INTC with 1.3M contracts, $SPCX with 847K contracts, $AAPL with 689K contracts, $MU with 673K contracts, $AMD with 615K contracts, $META with 568K contracts, $AMZN with 524K contracts, and $CRWV with 477K contracts. 10. The U.S. government is backing Rigetti $RGTI and D-Wave $QBTS with up to $200M for quantum computing R&D. Rigetti signed a definitive $100M agreement with the U.S. Department of Commerce to address key scaling bottlenecks in superconducting quantum systems, including cryogenics, readout electronics, and high-connectivity chip fabrication. D-Wave separately finalized access to up to $100M to advance its annealing and gate-model platforms, including work on a 100,000-qubit annealing system and a 10,000-qubit gate-model system targeting 100 logical qubits. As part of the funding, the Commerce Department will receive minority, non-controlling equity stakes in both companies. 11. Uber $UBER President and COO Andrew Macdonald purchased 70,000 shares for roughly $5.3M. Macdonald is one of Uber’s longest-tenured executives, making the insider buy notable given his deep knowledge of the business, operations, and long-term strategy. 12. U.S. data center construction spending surged 57% YoY in July to a record annualized rate of $75B, following a 46% YoY increase in June and marking the fastest growth since mid-2025. Since the start of 2021, data center construction spending has soared by $66B, or 717%. Since the end of 2023 alone, spending has risen by $51B. Over that same period, all other private construction spending — including housing, shopping centers, and offices — has declined by $120B, showing how dramatically AI infrastructure is diverging from the rest of the construction market. WALL STREET IS THE GREATEST SHOW ON EARTH.

TL;DR: libp2p TCP transport with dual trainer/syncer roles = chef's kiss For model training, don't bottleneck and single-point-of -failure yourself with a blob store (S3/R2). Also, don't bash your forehead against a wall using listening sockets on nodes that may be behind firewalls/NAT/port mapped containers/etc. Just set up a separate backbone (sync only, non-GPU/training nodes) across the world with super fast WAN and use push only from training/GPU nodes to this layer. Easy peasy, works like a charm. And bonus, libp2p's TCP transports are reliably better across providers/networks/countries compared to default quic/udp. Skynet?

📢 𝗝𝗨𝗦𝗧 𝗜𝗡: $IONQ IonQ and Synopsys Quantum Breakthrough Cuts Complex Simulation Times by Up to 14.6% 👉 𝗞𝗲𝘆 𝗛𝗶𝗴𝗵𝗹𝗶𝗴𝗵𝘁𝘀: ➤ 𝗜𝗼𝗻𝗤 and 𝗦𝘆𝗻𝗼𝗽𝘀𝘆𝘀 achieved simulation acceleration of up to 𝟭𝟰.𝟲%. ➤ Hybrid quantum approach targets major 𝗰𝗹𝗮𝘀𝘀𝗶𝗰𝗮𝗹 𝘀𝘂𝗽𝗲𝗿𝗰𝗼𝗺𝗽𝘂𝘁𝗶𝗻𝗴 bottlenecks. ➤ Quantum algorithm was integrated directly into 𝗔𝗻𝘀𝘆𝘀 𝗟𝗦-𝗗𝗬𝗡𝗔 simulation software. ➤ Industrial models included automobiles, jet engines, impellers, and drill components. ➤ Tests involved meshes containing up to 𝟯𝟱 𝗺𝗶𝗹𝗹𝗶𝗼𝗻 data points. ➤ Numerical simulations scaled to 𝟭𝟱𝟬 𝗾𝘂𝗯𝗶𝘁𝘀. ➤ Physical validation used IonQ's 𝟯𝟲-𝗾𝘂𝗯𝗶𝘁 𝗙𝗼𝗿𝘁𝗲 quantum computer. ➤ Runtime improvements ranged from at least 𝟱.𝟵% to 𝟭𝟰.𝟲%. ➤ A seven-day simulation could save approximately 𝗼𝗻𝗲 𝗳𝘂𝗹𝗹 𝗱𝗮𝘆. ➤ Research won 𝟭𝘀𝘁 𝗣𝗹𝗮𝗰𝗲 𝗕𝗲𝘀𝘁 𝗣𝗮𝗽𝗲𝗿 at IEEE Quantum Week 2026. 👉 𝗪𝗵𝘆 𝗧𝗵𝗶𝘀 𝗠𝗮𝘁𝘁𝗲𝗿𝘀: ➤ Demonstrates a practical 𝗵𝘆𝗯𝗿𝗶𝗱 𝗾𝘂𝗮𝗻𝘁𝘂𝗺 use case in engineering software. ➤ Faster simulations could reduce 𝗰𝗼𝗺𝗽𝘂𝘁𝗲 𝗰𝗼𝘀𝘁𝘀 and energy consumption. ➤ Shorter runtimes could accelerate automotive, aerospace, and industrial product development. ➤ Shows quantum systems can complement existing supercomputers before fault-tolerant quantum computing. 👉 𝗘𝘅𝗽𝗲𝗿𝘁 𝗦𝘁𝗮𝘁𝗲𝗺𝗲𝗻𝘁𝘀: 𝗗𝗿. 𝗠𝗮𝗿𝘁𝗶𝗻 𝗥𝗼𝗲𝘁𝘁𝗲𝗹𝗲𝗿, IonQ’s Vice President of Quantum Solutions and Technical Fellow and co-author of the paper: "Engineering simulations for jet engines or automotive crash tests require extremely large classical compute clusters running for days at a time. By embedding an advanced quantum algorithm into the core workflow, we reduced total simulation times by up to 14.6 percent," said Dr. Martin Roetteler, IonQ’s Vice President of Quantum Solutions and Technical Fellow and co-author of the paper. "Because this complex organizing step only needs to happen once at the start of a simulation, the time saved pays off repeatedly. That means faster solutions, reduced compute costs, and lower energy consumption for industrial users." 𝗣𝗿𝗶𝘁𝗵 𝗕𝗮𝗻𝗲𝗿𝗷𝗲𝗲, Senior Vice President of Innovation at Synopsys: "Together with IonQ, we are advancing the development of quantum computers and hybrid simulation workloads to unlock new possibilities for engineering design and optimization,” said Prith Banerjee, Senior Vice President of Innovation at Synopsys. “Quantum computers have the potential to drastically accelerate large simulation workloads while improving predictive accuracy. By focusing on achievable breakthroughs in the NISQ era, we can unlock some of that potential in the near-term for our customers while preparing for the full promise of fault-tolerant quantum computing."