---
title: Writing
url: https://www.consultingai.dev/blog/
original: https://www.consultingai.dev/blog/
---


# Writing

Notes on significant moves in AI, and how to build agents, RAG and evaluations that hold up. Every post opens with a short answer, then shows the evidence.

Writing is seed pods drifting down a thread of light · Stockcake · Stockcake (public domain).

1. [**Between two Samsung lines, the eyes post: the shortage is now a payout story**](/blog/jukan-eyes/)AI news · 29 Oct 2026Jukan's wordless 👀👀 post sat between Samsung's 2027 shortage line and its buyback remarks: what repriced that morning was the KRW 55–75 trillion return plan.
1. [**Samsung’s HBM4 sold at “target price” is the memory bill for Rubin**](/blog/jukan-hbm4-pricing/)AI news · 29 Oct 2026Mid-to-high $4 per gigabit is about $36–40 per GB, or $9,000–10,500 of HBM4 in one 288 GB Rubin GPU. Sold out at target price makes the shortage contractual.
1. [**Powertech’s $400M panel fab is where Broadcom’s ASIC volume goes when CoWoS is full**](/blog/jukan-powertech-plp/)AI news · 29 Oct 2026Taiwan cleared Powertech's US$400M Singapore panel-level packaging fab on 28 October: the second queue for AI packages when TSMC CoWoS is full, with Broadcom ASIC volume.
1. [**Samsung’s 60–70% LTA target is a mortgage on 2030 capacity**](/blog/jukan-samsung-ltas/)AI news · 29 Oct 2026Samsung wants 60–70% of memory on long-term agreements through 2030. The contracts lock volume, not price: a demand schedule with deposits that now justifies the capex.
1. [**Samsung’s 2028 line is a lead-time statement, not a demand forecast**](/blog/jukan-tight-supply-2028/)AI news · 29 Oct 2026Samsung guides KRW 84.4 trillion of 2026 capex and still calls 2028 tight: money buys cleanrooms and tools on 12–18 month lead times before it buys output.
1. [**What is DataVendor by HUD, and who pays for your data?**](/blog/datavendor-hud/)AI news · 28 Oct 2026DataVendor is HUD's new marketplace where teams license codebases, RL tasksets and raw data to AI labs. Here is how a sale works, what it costs, and what is still unverified.
1. [**Can you steer sycophancy out of a small model without retraining it?**](/blog/sycophancy-steering-small-models/)Interpretability · 2 Dec 2025LLM sycophancy, removed with activation steering: it shows up as a linear direction in a small model's hidden states. Subtracting that direction while the model generates raised accuracy on sycophantic prompts from 12.5% to 53.8% on Qwen3-0.6B, with no fine-tuning.