The Frontier

Jordi Visser identifies a 'compute shortage' as a critical current issue, as AI adoption rates have outpaced the supply of data centers and necessary hardware, potentially slowing companies' ability to replace labor and impacting margins.

Jordi Visser's portfolio is heavily weighted towards 'scarcity assets' supporting the AI infrastructure, including memory stocks like Micron and Pure Storage, chip-related companies like Marvell, and raw material producers like silver miners and Brazilian mineral companies.

KPMG, Meta, and Nike recently announced layoffs affecting 8,000-10,000 employees in total. Janelle Gail, Meta's Chief People Officer, stated these layoffs help offset significant investments in AI infrastructure, like a multi-billion dollar deal with Amazon for Graviton chips.

Vuebuds offer advantages over smart glasses, which faced cultural resistance, by integrating visual AI into a discreet device already worn by over a billion people. The camera module costs only $1-2.

DeepSeek V4 Pro boasts 1.6 trillion parameters and a 1 million token context window, signaling significant AI advancement. Apple's internal silicon and universal memory position it well for AI integration, despite current restraint on "AI nonsense."

Scaling01 estimates GPT 5.5's parameters are 2-5 trillion, compared to Mythos at approximately 10 trillion and GPT 5.4 at 1-2 trillion.

Project 11, a post-quantum security startup, awarded one Bitcoin to researcher Giancarlo Lelli for breaking a 15-bit elliptic curve key using a publicly accessible quantum computer.

Bitcoin's digital signature scheme uses 256-bit elliptic curve cryptography, significantly more complex than the 15-bit key broken, with experts divided on how quickly quantum systems can scale to real-world levels.

Casey Newton highlights the Apple Watch, AirPods, and the Apple Silicon bet as Tim Cook's key product successes. The Apple Watch, initially doubted, became a mainstream health product through iteration, while Apple Silicon gave Apple control over its chip design, weaning it off Intel.

Companies are aggressively pursuing AI-related capital expenditures to remain competitive, with Tesla projecting over $25 billion in capex this year. The semiconductor index is up 16 consecutive days, and Intel's earnings beat caused a 13% stock rise after hours.

For his plotter project, Wandel wrote a custom division routine for the 6502 processor, which lacked a built-in divide instruction, by implementing the manual long division algorithm.

An LLM query involves an immense number of computations, roughly equivalent to "one computation for every grain of sand on Cavendish Beach" including the dunes, according to a comparative anecdote.

Nvidia CEO Jensen Huang declared at GTC that "every software company in the world needs to have an OpenClaw strategy" while demonstrating Nvidia's enterprise-grade agent software.

Schrelli is launching a hardware startup in photonic computing, a field he believes could offer 1,000x to 1,000,000x performance improvements in speed and energy efficiency over current GPUs.

Schrelli contends that quantum computing is largely impractical beyond Shor's algorithm, noting that current quantum computers operate at megahertz speeds and are single-threaded, far slower than Nvidia GPUs.

The market for new computing hardware like photonics could reach $5-10 trillion, yet Schrelli finds a surprising lack of startups in this space compared to the thousands in AI agents.

Apple's M-series chips are an underrated asset for local LLM inference, outperforming DGX Spark benchmarks, and the company has already secured two-nanometer chip fab capacity for next year, positioning it for future local AI processing.

Alex notes ChatGPT Images 2.0 achieved an ELO score of 1,512 in arena.aI benchmarks, making it the top-scoring text-to-image model. He speculates its release might indicate lower compute costs or its instrumental role in code generation.

Alex explains that SpaceX is transforming into a hyperscaler, building a "Dyson swarm" of orbital clouds. The Cursor acquisition could signify an opening of SpaceX's vast ground and orbital GPU sets to third parties.

The annual threat assessment highlights an "ambient threat environment" where AI and quantum computing expedite existing problems, leading to a connected ecosystem of threats. Chris Wingfield notes this reduces episodic attacks to a constant, pervasive risk.

Michael Stewart intends to lead an investment in material science for AI this year, focusing on delivering singular, actionable leads that could accelerate breakthroughs in fields connected to the future data center.

Michael Stewart asserts that current data centers are inefficient, evolved from the IBM PC era of the 1980s, with lithography reticle size being a central limiting factor for chip size, forcing the use of less energy-efficient die-to-die interconnects.

Michael Stewart contends that die-to-die interconnects are 10 to 50 times less energy-efficient than on-chip communication, contributing to wasted energy by extending the current compute paradigm.

Theo theorizes Anthropic forces heavy users to the dumber 1 million context version (now default and free) on Amazon/Google TPUs, reserving superior Nvidia GPUs for researchers during a compute crisis.

TSMC reported a 35% revenue boost and forecasts over 30% growth but faces capacity limits, with ASML unable to supply lithography machines. Nikkei Asia predicts memory chip shortages until at least 2027, meeting only 60% of demand.

Amazon expanded its Anthropic partnership with a $25 billion investment, providing 5 gigawatts of compute, including Tranium 3 chips, to resolve Anthropic's inference shortage and ensure Claude's availability via AWS.

UTXO runs a local AI server with four RTX 3090 GPUs, initially acquired for $600 each, for developing and serving Nostr bots, demonstrating a cost-effective setup for open-source AI.