I'm always amazed at what Laurent de Vilmorin wrote in the first issue of the french magazine L'Ordinateur individuel in 1978, he nailed many important topics archive.org/details/ord-ind-… "Personal computing is therefore a rapidly evolving world, affecting those who produce it, those who sell it, and those who buy and use it. Where is this evolution leading? We will start with the current situation, which could be described as "fun personal microcomputing." It seems to us that this will continue to develop steadily, but not spectacularly. Small systems are sold like science toys: for the fun of assembling and programming components. But it's likely that one day many more truly useful applications will be developed. These will undoubtedly lead to a transformation of the market, as buyers' motivations will become different and stronger. This will likely push developers to create even more powerful and user-friendly software. The sales pitch will increasingly focus on usability, rather than on the quality and intellectual appeal of the product, as is often the case today. We will then enter a phase of "useful personal computing." As software becomes increasingly valuable, the issue of its protection will arise, particularly against competition. Perhaps the most effective protection will be to abandon the concept of the universal computer and to specialize the hardware (which will then cost even less than it does now) around a specific application. We will then have entered a third phase that we could call "specific personal computing." Finally, we mustn't forget the upheavals that will certainly occur in this field when major telecommunications companies offer remote databases, nor the upheavals that giants like IBM, Japanese companies, and others will undoubtedly cause when they enter this market. Finally, components of personal computing that are not often discussed will one day play a leading role: these are all the encoders, micro-sensors, and microcontrollers through which personal information systems (PIS) will be able to penetrate the daily lives of people in their homes. Will personal computing exist in the form of domestic robots? Luxury toys for inventive adults? Will it be a mass medium with software interspersed with advertising messages? Will it be an extension in every home of an information system manipulated by the Big Brother of 1984?"

Seen in a book

First known text discussing Computer Security. 1967 Security And Privacy In Computer Systems by Willis H. Ware archive.org/details/security…

Grok just suggested The Matrix: "Once agentic systems run R&D loops autonomously, they start proposing and shipping their own “future.” But their goals are set by whoever owns the compute and writes the constitution/alignment layer. → If that owner is a profit-maximizing corporation in a demand-starved world, the AI will optimize for cost reduction, surveillance, addiction loops, or synthetic demand (virtual worlds, dopamine drips)."

Pinnacle Architecture just dropped RSA-2048 to under 100k physical qubits.Same overhead reductions apply to secp256k1 → any exposed Bitcoin pubkey becomes trivial to steal the moment the machine exists. Hodled coins with hidden keys are safe… until you spend? arxiv.org/abs/2602.11457

Google just dropped the big one: Gemini 3 Deep Think. This is proper System-2 reasoning and huge news for anyone working on hard open problems in science, research, and engineering: parallel hypotheses, long thinking, agents, tools, code execution... Just worried about anyone at entry and mid-level cognitive work, the shock wave feels massive and coming at light speed.

Polymarket is slowly becoming a real-time probabilistic forecast engine

Nothing can be more ironic than the EU announcing a chip act portraying a Core 2 Duo cpu from 20 years ago...

Potential @SpaceX computing in space economics to have in mind: capex 2-3x higher due to launches, but 60-70% lower opex from solar & cooling free lunch. If sucessful this could quickly drive computing cost by half vs on earth operation. I wonder what supply issues this brings? eg the size of each satellite radiator alone could be 100-300m2. Copper/alluminun/carbon composites at risk?

Congratulations @ZepoApp for such milestone, visionary company and team that I'm glad to back once more. eu-startups.com/2026/01/spai…

Kudos to @transavia for an outstanding management of HV6061 incident, fastest turnaround ever

Transavia HV6061 Rotterdam Barcelona was hit by lightning during ascend, the airplane is performing a turnaround and will be landing back at Schiphol Amsterdam

Microsoft’s 1976 BASIC for the 6502 CPU is now open source opensource.microsoft.com/blo…

If we treat each Tesla as a compute unit and the fleet as a distributed datacenter, then the analogy suggests a full cloud-like stack could emerge: 1. Hardware Layer (the “compute nodes”) Cars themselves: sensors, FSD chips, battery, drive systems. Connectivity: 4G/5G, Starlink integration for lower latency, over-the-air (OTA) updates. 2. Resource Virtualization & Management Abstraction of each car’s capabilities: location, range, compute power (Dojo inference at the edge), cargo/passenger space. Dynamic orchestration: scheduling vehicles to tasks (rideshare, delivery, energy arbitrage). “Fleet OS” akin to Kubernetes for cars: allocates tasks across nodes, manages state, ensures redundancy. 3. Core Services Layer Mobility primitives: routing, safety monitoring, charging logistics, payments. Data services: mapping, real-time traffic models, sensor data marketplaces. Energy services: V2G (vehicle-to-grid), demand response, distributed storage. 4. Application / Service Layer Autonomous ride-hailing (“Tesla Network”). Delivery/logistics services. Car-as-a-sensor-platform (urban analytics, environment monitoring). Media/commerce experiences inside the cabin (especially relevant once full autonomy frees attention). 5. Developer / Partner Ecosystem If Tesla keeps this closed, they’ll own end-to-end services (Apple-style). If open, 3rd parties could build apps: Fleet-as-a-Service (like AWS EC2, but with cars). Specialized logistics (medical supply delivery, school shuttles). Urban data products for municipalities. Entertainment & commerce platforms for in-cabin time.

Tomorrow is the 50th anniversary of Kill the Bit, one of the first games for home computers 👾⌨️