The AI Daily Five · No. 23 · Tuesday 10 November
Is Technology Facing Its Own Limits?
Australia brings quantum computing into commercial infrastructure, US AI stocks react to electoral change, Wall Street selects blockchain settlement technology, researchers train AI to simulate the physical world, and communities protest data-centre expansion.

1
Can Australia’s Quantum Computers Leave the Lab?
Australian company Diraq has installed a silicon-spin quantum computer in a commercial Sydney data centre. The milestone brings quantum technology closer to existing computing infrastructure, although its practical applications remain largely unproven.
Quantum computing has taken a step beyond the research laboratory. Australian-founded Diraq announced on 9 November that it had installed an eight-qubit silicon-spin quantum computer inside an Equinix commercial data centre in Sydney.
The company describes the installation as the first silicon-spin quantum computer operating in a shared commercial data centre, alongside conventional computing infrastructure.
Unlike systems that require purpose-built research facilities, Diraq’s design incorporates cooling and control equipment within a compact installation. Its quantum bits, or qubits, are manufactured using silicon-based technology related to processes already used in the semiconductor industry.
The company first announced the project in August. The November development is the installation itself, rather than a new agreement.
Diraq says it intends to expand the system to as many as 256 qubits over the following year, primarily by replacing the chip rather than rebuilding the surrounding infrastructure.
That target remains prospective. An eight-qubit computer cannot yet perform the complex industrial calculations associated with large-scale, fault-tolerant quantum computing.
Alongside the installation, Diraq published Quantum at Work, a report exploring potential applications in medicine, materials and logistics. Its estimates of pharmaceutical benefits are based on modelling rather than demonstrated clinical outcomes.
The distinction matters because useful quantum computing requires more than increasing the number of qubits. Systems must also control errors, maintain reliable operations and demonstrate advantages over conventional computers on commercially meaningful tasks.
The Sydney installation could make it easier for industry partners to explore such applications within familiar data-centre environments. It does not establish that quantum computing has already reached commercial maturity.
2
Did the US Elections Shake the AI Boom?
The Nasdaq suffered a sharp sell-off after Democratic victories raised concerns about data-centre development and AI spending. A subsequent rebound suggests investors are reassessing the industry’s prospects rather than abandoning the sector entirely.
The US technology sector suffered one of its sharpest market declines in more than a year on 5 November, as investors reconsidered the prospects for artificial intelligence following the midterm elections.
The technology-heavy Nasdaq Composite fell 5.5% in a single session, wiping approximately $2.5 trillion from its market value. The S&P 500 declined 4.4%, while companies closely associated with AI infrastructure recorded some of the largest losses.
Nvidia shares dropped 9.3% and AMD fell 17.3%. Companies providing electricity and infrastructure for data centres also came under pressure.
The sell-off followed Democratic victories in the 3 November elections, including gubernatorial contests in states with substantial AI infrastructure investment.
According to Reuters’ market analysis, investors were concerned that tighter oversight and opposition to new data centres could slow construction and weaken expected demand for AI chips and computing capacity.
Some newly elected state officials have proposed stronger restrictions or requirements for data-centre development, including measures addressing electricity costs and local environmental impacts.
However, the political result was not the sole cause of the market decline.
AI-related shares had already been experiencing volatility. Investors were questioning the cost of infrastructure expansion, the returns available from increasingly expensive computing facilities and the ability of smaller companies to finance their plans.
The election added another source of uncertainty to an industry already facing scrutiny over its spending.
By Monday, 9 November, US markets had rebounded. The Nasdaq gained 2.7%, while the S&P 500 rose 2.3%, supported partly by falling oil prices.
The recovery does not resolve questions about AI profitability or the political future of data-centre expansion. It does, however, demonstrate the danger of treating a single trading session as evidence of a lasting industry collapse.
Technology companies continue to invest heavily in AI infrastructure, but investors are now assessing those ambitions against changing economic and political conditions.
3
Is Wall Street Moving Onto Blockchain?
The New York Stock Exchange has selected Avalanche technology for the settlement infrastructure of a planned tokenised securities platform. The arrangement would bring blockchain into market operations, although the platform has not yet received regulatory approval.
The New York Stock Exchange is moving closer to operating a blockchain-based securities market after selecting Avalanche to support the clearing and settlement of tokenised shares.
Ava Labs, the company behind Avalanche, announced the agreement on 9 November.
Avalanche will provide the core ledger technology to the clearing firm supporting a planned NYSE-affiliated alternative trading system. The new platform is designed to allow securities to be traded, cleared and settled around the clock, subject to regulatory approval.
The distinction between trading and settlement matters. Trading is the agreement to buy or sell an asset; clearing and settlement involve confirming obligations, transferring ownership and completing payment.
Conventional US equity markets operate within established trading hours, while settlement generally takes place on the following business day.
Tokenised securities represent ownership interests using blockchain-based records. The proposed platform aims to bring those records into the infrastructure responsible for completing transactions, potentially reducing the time between a trade and its settlement.
NYSE has indicated that the system will support equities issued on different blockchains, while Avalanche will provide the ledger used by the clearing firm. The arrangement does not mean that every stock listed on the existing exchange will move to Avalanche.
The planned venue would operate separately from the NYSE’s conventional trading market.
The announcement follows an earlier NYSE plan to develop a digital securities platform. It does not confirm that trading has begun, that regulatory approval has been granted or that the proposed settlement infrastructure has been tested under live market conditions at scale.
Avalanche joins several blockchain providers competing to supply infrastructure for established financial institutions.
For Wall Street, the potential attraction extends beyond creating digital versions of existing assets. It includes changes to how transactions are recorded, completed and financed.
4
Can AI Simulate the Physical World?
Researchers at MIT and the University of British Columbia have developed an AI architecture capable of modelling everything from flowing liquids to moving proteins. The approach could accelerate scientific research, although it is not yet a universal physics simulator.
A pile of sand collapsing, water flowing through a container and a protein changing shape might appear to have little in common. For scientists attempting to simulate these processes, however, they present a similar problem: each typically requires specialised software built around different physical equations.
Researchers at the Massachusetts Institute of Technology (MIT), the University of British Columbia (UBC) and collaborating institutions have developed an AI system that could simplify this process.
Called WorldParticle, the system uses a common neural-network architecture to model six categories of physical phenomena, including fluids, deformable materials, granular substances and molecular dynamics.
In its 9 November announcement, MIT explained that WorldParticle represents materials as collections of tiny three-dimensional particles. The system learns to predict how those particles move and interact over time.
The approach resembles the predictive mechanism behind large language models. Rather than anticipating the next word in a sentence, WorldParticle predicts the next positions and properties of particles within a simulated environment.
Conventional simulations often require engineers to translate the equations governing a particular material into numerical instructions. Different materials may therefore require different tools.
WorldParticle uses a shared computational architecture across the six categories, potentially reducing the need to develop separate simulation systems for every physical process.
The researchers demonstrated the approach on several physical processes.
In one experiment, WorldParticle simulated granular material collapsing into a pile. Some comparison models incorrectly left particles suspended in the air, while WorldParticle reproduced the movement more realistically.
The team also investigated molecular motion, including changes in protein structures. MIT reports that its approach modelled protein dynamics approximately 100 times faster than conventional numerical simulations.
However, the original research paper does not report a direct speed comparison for proteins, so the 100-fold figure comes from MIT’s announcement.
There is another important qualification. Although WorldParticle uses one architecture, the researchers trained separate networks for each category of physical behaviour. A single trained model that can simulate every type of material remains a future goal.
The approach also has computational limitations, particularly when dealing with very large numbers of particles or complex physical interactions.
Nevertheless, the research could eventually support applications in materials science, soft robotics and biological research. Engineers might use related systems to investigate how liquids interact with flexible machinery or how robotic devices respond to different surfaces.
These applications are still research possibilities rather than commercially validated capabilities.
The work is scheduled for presentation at SIGGRAPH Asia 2026. Its broader significance lies in the possibility of creating more flexible simulation tools that can be adapted across scientific disciplines.
5
Why Is Europe Protesting AI Data Centres?
Campaigners organised demonstrations across 15 European countries over the weekend, calling for restrictions on large data-centre projects. Their concerns extend beyond AI to electricity costs, water consumption and local communities' influence over development.
Europe’s expanding data-centre industry faced coordinated protests over the weekend as environmental groups and local campaigners challenged the construction of large computing facilities.
The Stop Dirty Data Centres coalition organised coordinated actions in 15 European countries on 6 and 7 November, bringing together more than 100 organisations.
Demonstrations took place in countries including Spain, France, Germany, the Netherlands, Poland and the United Kingdom.
In the Netherlands, protesters entered a Microsoft construction site near Amsterdam, temporarily disrupting work. Police detained approximately 150 to 200 people.
The coalition is demanding a moratorium on new large-scale data-centre approvals until clearer rules govern their environmental and social impacts.
Protesters argue that developments are moving forward without sufficient public consultation, particularly in communities already facing pressure on housing, electricity infrastructure or water resources.
The growth of artificial intelligence is an important driver of demand for computing capacity. Training and operating large models requires substantial processing power, while companies are expanding the infrastructure used to provide AI services to customers.
However, data centres also support conventional cloud computing, telecommunications and many everyday digital services. Their environmental effects differ according to location, electricity supply, cooling systems and the technology being operated.
Industry representatives argue that new facilities support economic growth, digital services and employment. Developers also point to investments in more efficient equipment and renewable energy.
Campaigners maintain that promised economic benefits do not adequately address the costs borne by nearby communities.
The European demonstrations follow political opposition to data-centre projects in the United States, where voters supported numerous local restrictions during the 3 November elections.
Together, these developments illustrate how decisions about computing infrastructure are becoming subjects of public debate rather than matters left primarily to technology companies and investors.
The AI Daily Five