The FDE Coach blog · new articles every day
Practical guides for engineers who ship AI.
Step-by-step builds with free AI tools, what just shipped in AI and why it matters, and Forward Deployed Engineer playbooks on interviews, pay and client work.
What just shipped in AI, and why it matters for engineers.
Page 9 of 11
AI News
Cactus Hybrid: Teaching Gemma 4 to Say “I Don’t Know”
Cactus Hybrid calibrates Gemma 4’s confidence so it knows when it’s wrong. Here’s how the method works, why it matters for engineers shipping reliable AI, and how to try it today.
7 min read
AI News
Echo Matches Fable at 1/3 the Cost: An Engineer's Blueprint for Open-Weight Distillation
Echo replicates Fable's narrative quality using open-weight models at one-third the cost. A technical deep-dive into distillation, cost benchmarks, and a practical blueprint for engineers to run it themselves.
10 min read
AI News
Kimi K3 vs Fable: Sparse Attention & RL Redefine Reasoning Models
Kimi K3 matches Fable using sparse attention and RL. Here's why that matters for building cheaper, faster reasoning agents and how to run it today.
14 min read
AI News
Claude Fable Generated a Counterexample to the Jacobian Conjecture: What It Means for Engineers
Claude Fable produced a counterexample to the Jacobian Conjecture—a 85-year-old math problem. We unpack the event, why it matters for FDEs building with LLMs, and how to harness reasoning models for hard technical work.
10 min read
AI News
OpenAI Presence: The Desktop Operator That FDEs Can Actually Use Today
OpenAI Presence isn't just a demo—it's an operator that takes over your desktop. Here's how it works, why it matters for Forward Deployed Engineers, and how to build your own today.
10 min read
AI News
The Reasoning Effort Knob: Trading LLM Cost for Accuracy
New research shows you can dial reasoning effort up or down in LLMs without retraining. Learn how this inference-time control works, why it matters for cost-sensitive pipelines, and how to implement it today.
10 min read
AI News
GigaToken: How a Custom Rust Tokenizer Achieves 1000x Speedup Over HF Tokenizers
A custom Rust tokenizer claims a 1000x throughput leap over Hugging Face tokenizers. We dig into the architecture, the memory tricks that make it fly, and what it means for your inference pipeline.
10 min read
AI News
Gemini 3.6 Flash Kills Sampling Knobs: What It Means for Your Pipelines
Google's Gemini 3.6 Flash ignores temperature, top_p, and top_k. Here's the engineering impact on determinism, caching, and pipeline design—and how to adapt today.
11 min read
AI News
Jane Street's Incremental: Engineering Self-Adjusting Computations for the Real World
Explore Jane Street's Incremental library, a practical OCaml tool for building self-adjusting systems. Learn how it transforms static code into dynamic, efficient pipelines—a powerful concept for engineers building data-heavy applications.
10 min read
AI News
Buzz by Block: Merging Team Chat, AI Agents, and Git into One Platform
Jack Dorsey's Block launches Buzz, an open-source platform fusing team chat, AI agents, and Git hosting. Here's why it matters for FDEs and engineers, plus how to spin it up today.
9 min read
AI News
Moonshine: A Rust-Based Server for the Moonlight Streaming Protocol
Moonshine reimplements NVIDIA's GameStream protocol in Rust, letting you stream PC games to Moonlight clients without GeForce Experience. A deep dive for engineers on why this matters for latency, safety, and self-hosting.
9 min read
AI News
Qwen-Image-3.0: What 'Deep Knowledge' Means for Text-to-Image Models
Qwen-Image-3.0 introduces 'Deep Knowledge' to text-to-image generation, moving beyond aesthetics to factual accuracy. We break down the architecture, why it matters for building reliable vision pipelines, and how to run it today.
11 min read
AI News
Gemini 3.5 Flash Cyber: Running Security-Focused LLMs at Scale
Google just shipped a security-hardened LLM. We dissect Gemini 3.5 Flash Cyber for engineers: how it differs from standard models, why it matters for FDEs, and how to run it today.
10 min read
AI News
Nativ: Run Frontier Open Models Locally on Your Mac Without the Ops Overhead
Nativ bundles a local API server, chat UI, and model downloader into a single native macOS app. Engineers and FDEs can now run Llama 3, Mistral, and other frontier open models locally with zero YAML wrangling.
11 min read
AI News
Measuring AI Writing on arXiv: Where Detection Breaks Down
A deep dive into measuring LLM-generated text on arXiv using statistical proxies like 'marker words.' We unpack the methodology, the engineering pitfalls of false positives, and why rigid detectors fail against evolving models.
8 min read
AI News
OpenAI Codex Context Window Shrinks 100k Tokens: What the PR Reveals
OpenAI quietly reduced the Codex model's context window from 372k to 272k tokens. We dissect the pull request, explain why this matters for engineering workloads, and how to adapt your agents.
8 min read
AI News
GPT-5.6 Found a $500k WordPress RCE: How AI is Reshaping Vulnerability Research
A security researcher used GPT-5.6 and $25 in API credits to find a WordPress RCE worth $500k on exploit markets. Here's what happened, why it changes the game for engineers, and how to start.
8 min read
AI News
Agent Swarms and the New Model Economics: Why Routing to Smaller Models Wins
The economics of AI are inverting. Agent swarms prove that a router directing tasks to specialized, smaller models often beats a single giant LLM on cost, latency, and accuracy.
8 min read
AI News
Fable 5 vs GPT-5.6 Sol on NP-Hard: Does /goal Actually Help?
We stress-tested Fable 5 and GPT-5.6 Sol on the NP-hard Traveling Salesman Problem. Here's the raw data, why the /goal command didn't magically solve intractability, and what it means for engineers building AI tooling.
9 min read
AI News
AI Advice Made Engineers 3x Less Accurate but 2x More Confident: What the Study Means
A new study shows AI advice suppresses critical thinking, making engineers 3x less accurate but 2x more confident. Here's the raw data, why it matters for FDEs, and how to build guardrails.
9 min read
AI News
Why Claude Code Moved to Bun-in-Rust (And What It Means for Your Toolchain)
Claude Code's move to a Rust-compiled Bun runtime isn't just a trivia footnote—it's a signal about where the agentic coding toolchain is heading. Here's the engineering breakdown.
9 min read
AI News
AI vs. ZK: How an LLM Found Real Bugs in OpenVM’s Verifier
An LLM auditing zero-knowledge proof code found genuine cryptographic flaws. Here’s exactly what happened, why it changes the game for engineers and FDEs, and how to start using AI for security review today.
11 min read
AI News
Inside VulnHunter: How Capital One Built an Agentic Code Security Scanner
Capital One open-sourced VulnHunter, an agentic AI tool that hunts code vulnerabilities. We break down the multi-agent architecture, why it matters for FDEs, and how to run it today.
9 min read
AI News
Headless AI Programming: Setting Up a Spare Mac for Claude Code to Control
Turn a spare Mac into a headless AI coding agent. A step-by-step engineering guide to remote SSH access, VNC fallback, and security hardening for Claude Code.
12 min read
259 articles and counting