Company Spotlight: American Cloud
American Cloud is an America-First hosting provider whose goal is to promote a free and open internet for everyone. We’ve worked with them a whole bunch, and really appreciate their team’s commitment to delivering world-class hosting services and their commitment to America’s core principals.
Daring Bits
🚨 FCC to Take Jack-Booted Control of the Internet! 🚨 - President Biden plans to centralize control over and regulate nearly every aspect of the internet - like, next week. The FCC is voting on it on the 15th. FCC Commissioner Carr is sounding the alarm, saying the plan imposes costly infrastructure mandates, fines without ceilings, and price controls on ISPs. The 2021 Infrastructure Investment and Jobs Act serves as the legal framework empowering the FCC to adopt new rules, with the stated goal of achieving “digital equity” 🤮 and facilitating equal access to broadband internet services. It only just so happens to open the door to a federal takeover of the internet. Ain’t the legislative process great?
Federal Reserve Threatens Bitcoin Magazine’s Speech - The Federal Reserve is trying to silence critics of FedNow, and has threatened Bitcoin Magazine with legal action over merchandise that satirizes the surveillance capabilities of the FedNow system. Bitcoin Magazine is refusing to comply with the Federal Reserve’s cease and desist letter, arguing that parody is protected speech, not copyright infringement. Can’t take criticism? You might be a tyrant.
Lawmakers Propose Open Hardware Restrictions - RISC-V is an open-source instruction set architecture (ISA) for designing computer processors. RISC-V’s open specifications make the architecture auditable, unlike the black box of security risks under closed architectures such as x86 or ARM. Now, a bipartisan group of US lawmakers is urging RISC-V restrictions to prevent Chinese dominance in CPU technology. Andrew ‘bunnie’ Huang, an active participant in the RISC-V ecosystem, argues that such restrictions would undermine American competitiveness, and suggests investing in homegrown RISC-V chip makers instead. In a letter, he urges President Biden not to impose restrictions, and highlights the benefits of open standards for innovation and freedom of expression.
Europe: All Your Identity Are Belong To Us - The European Parliament and the Council of the EU have reached a final agreement on European Digital Identity Wallets, a digital identity framework for EU citizens, enabling access to public and private online services across Europe. “Very Large Online Platforms designated under the Digital Services Act (including services such as Amazon, Booking.com or Facebook) and private services that are legally required to authenticate their users will have to accept the EU Digital Identity Wallet for logging into their online services.”
Tests Are Bad For Developers 😬 - Stephan Schmidt explains why devs avoid writing tests for their code. The pressure to meet deadlines, coupled with the repercussions of bugs discovered during development—resulting in blame and missed goals—deters developers from prioritizing tests. One dev’s experience at AWS.
Robot Overlords
Elon Musk’s ChatGPT Rival Grok Enters AI Fray - xAI has launched its AI chatbot, Grok, touting access to X for real-time knowledge as a distinct advantage over incumbents like OpenAI and Bard. Musk had expressed support for an AI development pause, but now dismisses a pause as futile. Grok, inspired by “The Hitchhiker’s Guide to the Galaxy,” promises snarky responses and the ability to handle “spicy” questions. Musk claims Grok’s superiority over ChatGPT 3.5 in some benchmarks but acknowledges lagging behind OpenAI’s GPT-4. The chatbot’s release is offered to X Premium+ subscribers initially.
Mozilla Statement on AI Access - Mozilla has released a joint statement, with a diverse group of signatories, stressing the importance of openness, transparency, and broad access in AI governance. They argue that despite risks, increasing public access and scrutiny makes technology safer, not more dangerous.
AI Chatbots are Eating Their Own Tails - AI chatbots like ChatGPT face a fundamental issue called “model collapse,” where without new human input, their output quality deteriorates. This problem occurs when new AI models are trained on synthetic data derived from older models, causing them to overly depend on outdated patterns present in the original data. As a result, the accuracy and output variance of the models decline over repeated generations. The article highlights the concern that an internet flooded with AI-generated content could exacerbate this issue.
