Large Language Models

Today's Stories

1. 1

   Apple sues OpenAI, alleging theft of trade secrets for hardware project

   Apple filed a lawsuit in California federal court accusing OpenAI of encouraging former Apple employees to share confidential information and stealing trade secrets to support OpenAI's nascent hardware business. Two former Apple employees named as defendants are Tang Tan, now OpenAI's chief hardware officer, and Chang Liu, a former electrical engineer. Apple alleges both accessed confidential hardware-related files after leaving the company. Apple and OpenAI partnered in 2024 to integrate ChatGPT into iPhones, but the relationship has shifted toward rivalry as OpenAI pursues a physical AI device alongside recruiting Apple talent. The lawsuit suggests OpenAI may be building its hardware effort partly on knowledge obtained improperly, which undercuts the legitimacy of that business line if the claims hold. For Apple, it signals the company is willing to pursue aggressive legal action to protect product development secrets as competition in AI hardware intensifies.

   OpenAI CFO Sarah Friar stated in April that the company plans to release "consumer hardware that will come towards the end of this year." Apple reached out to OpenAI in February about its concerns, but said OpenAI did not respond. The lawsuit also names io Products, the company OpenAI acquired for nearly $6.5 billion（約1兆円） to work with designer Jony Ive on the hardware project, as a defendant.

2. 2

   Researchers propose philosophical approach to AI alignment training

   Researchers have developed a metaethical argument — combining perspectival moral realism with evolutionary debunking as an epistemological approach — and are considering submitting it as feedback to Anthropic or publishing it for broader engagement with AI alignment researchers. Anthropic has stated its constitutional approach to AI training is meant to be revised and improved over time, and substantive philosophical contributions are rarer than bug reports. The argument presented takes a position distinct from common approaches in AI ethics literature, which tend toward either naive moral realism or preference-satisfaction consequentialism — making it potentially more likely to gain traction precisely because it addresses moral uncertainty in a less common way.

   Although the probability that any single submission changes training decisions is low, the expected value may be higher than it seems, given Anthropic's openness to revising its approach and the relative scarcity of rigorous philosophical input to AI training methodology.

3. 3

   DeepSeek cuts prices 75%, but agent AI consumes tokens faster than savings

   DeepSeek drastically cut pricing on its V4-Pro model by 75%. However, the cost reduction is being offset by a fundamental shift in how AI systems operate—agent systems (AI that chains together planning, retrieval, tool use, verification, and follow-up decisions) consume tokens far faster than traditional chatbots, which turn one user question into one model call. For two decades, software economics followed a pattern where infrastructure became cheaper each year while applications became more capable, and many assumed AI would follow the same path. Instead, the shift toward agents is reversing that dynamic—even as per-token costs plummet, total token consumption per user task is rising sharply, making it harder for vendors to maintain healthy margins despite price cuts.

   The gap between token price declines and agent token consumption rates. The body describes this as an exponential problem that "has begun crumbling," signaling that the business model assumption underlying AI margin improvement may no longer hold.

4. 4

   How to Route AI Calls Across Models Without Losing Cache Savings

   An article explains why simply routing easy prompts to cheaper AI models often fails to save money—because switching models mid-task destroys the prompt cache (which costs 90% less on repeat access), forcing the context to be re-billed at full rates. Production routers solve this by pinning a task to a single model once chosen, keeping the cache warm across all calls within that task, and switching models only when a new task begins. For teams running AI agents that make multiple sequential calls (planning, tool use, result analysis), naive routing can leave costs unchanged despite switching to cheaper models. Understanding how to preserve cache affinity means the cost savings actually materialize—the article reports that applying this exact four-stage pipeline (guardrail filter, router model, cost policy, model affinity) reduced usage by 2x on a real agent without changing a line of code.

   The pipeline is implemented in Plano, an open-source proxy that runs between your agent and model providers. The routing model uses Arch-Router, a 1.5B model trained on human preference data. Plano includes an observability console that shows per-request cost and which model answered each request, letting teams see routing decisions in real time.

5. 5

   Reddit user seeks budget Chinese AI models after Anthropic blocks access

   A Reddit user posted that Anthropic has blocked their IP address after they created over 100 accounts using disposable email services, and they are now looking to switch to Chinese AI models including DeepSeek-V4-Pro, MiMo-V2.5-Pro, and GLM-5.2. The post illustrates user frustration with access restrictions and pricing on mainstream AI services, and shows interest in exploring alternatives from Chinese providers as a cost-effective option.

   The user's comparison highlights three Chinese models they are considering—DeepSeek-V4-Pro from Deepseek, MiMo-V2.5-Pro from Xiaomi, and GLM-5.2 from z.ai—as cheaper alternatives to paid services.