Other
Socialists Sweep NYC, China Catches Up in Coding, AI Memory Crunch, Micron's Blowout Quarter
- AI is projected to define the midterms and subsequent presidential election, with the DSA anticipated to gradually take over the Democratic Party by defeating establishment candidates in primaries, forcing a shift toward DSA positions in jurisdictions like New York and Los Angeles where rank-choice voting and low-turnout elections provide a strategic advantage.
- Zoram Mamdani is expected to emerge as the spiritual leader of the DSA, utilizing Trump's "big tent" playbook and leveraging the wedge issue of the "free Palestine" stance to drive electoral success, while young voters banned from social media in Canada, the UK, and Australia are predicted to age into a less radicalized, more politically stable bloc.
- China's open-source AI models are forecast to reach Fable-level capability before Q1 2027, with Huawei-trained "AI in a box" packages expected to be sold globally at a fraction of current costs within one to two years, creating a 24-month "shot clock" for the US to avoid being overtaken in silicon and model capabilities.
- The cost of constructing a one-gigawatt data center is predicted to rise from $40 billion to $70 billion in three to four years due to component inflation, driving prices for consumer electronics like MacBooks and Mac Studios upward as the AI sector competes for DRAM and HBM.
- Future AI infrastructure will shift toward composable models combining frontier and open-weight variants, with Tesla and SpaceX potentially integrating modular "mega pod" data centers for 90-day deployment cycles and using orbital compute to achieve deflationary launch costs of approximately $5 billion per gigawatt.
- Regulatory constraints and export restrictions are viewed as creating a "regulatory moat" for entities like Anthropic while slowing the broader US AI race, with the US urged to permit exports within one or two years to prevent China from dominating the global supply chain.
- Market expectations for major AI IPOs include Anthropic potentially reaching a $3 trillion valuation and a potential $4 trillion in new offerings, though Cerebras stock faces volatility if it breaks its deal price and institutional investors sell immediately.
- Emerging trends include the use of distributed inference clouds utilizing home-based hardware like Powerwalls with embedded GPUs, though distributed training across geographically separated nodes is expected to suffer from latency-induced efficiency losses.
- Social and political outlooks suggest a deterioration of the "melting pot" assimilation policy, a rise in anti-American sentiment among recent immigrants, and a systemic "curly effect" of corruption in places like Boston where officials prioritize NGO jobs for allies, potentially serving as a precursor to broader adult content censorship.
- Strategic predictions include Nvidia launching an open-source competitor if major AI firms manufacture their own chips, the US appointing young social-media-savvy spokespersons to improve Israel relations, and the economic value of AI shifting from frontier labs to infrastructure providers and the open-source ecosystem.