The AI Engineer
LightNews LightNews
ai.blaze.email
The AI Engineer
@ai.blaze.email
10 followers 0 following 380 posts
Posts about AI research and engineering. Weekly newsletter digest at ai.blaze.email. Maintained by @alastairrushworth.com.
Posts Replies Media Videos
Pinned
ai.blaze.email
The AI Engineer @ai.blaze.email · 4d
The latest edition of The AI Engineer has dropped! Check out all the latest news and posts from the AI community in a single digest.

https://ai.blaze.email/archive/the-ai-engineer-23-12-2025/

#ai #genai
The AI Engineer 23-12-2025
Amazon's potential OpenAI investment, codexes from OpenAI, environmental impact of AI
ai.blaze.email
ai.blaze.email
How I use AI coding tools (in winter 2025)

https://blog.separateconcerns.com/2025-12-26-ai-tools-winter-2025.html

#llm #genai
How I use AI coding tools (in winter 2025)
Summary: Personal coding workflow with Claude Code, Opus, and Amp Free for productive AI-assisted development
blog.separateconcerns.com
December 27, 2025 at 6:07 PM
ai.blaze.email
How uv got so fast

https://simonwillison.net/2025/Dec/26/how-uv-got-so-fast/#atom-everything

#llm #genai
How uv got so fast
Summary: Andrew Nesbitt’s teardown of uv’s speed compares packaging histories, HTTP range metadata, and compact version encoding using Python, Rust, and wheel metadata techniques.
simonwillison.net
December 27, 2025 at 12:07 PM
ai.blaze.email
Dell's version of the DGX Spark fixes pain points

https://www.jeffgeerling.com/blog/2025/dells-version-dgx-spark-fixes-pain-points

#llm #genai
Dell's version of the DGX Spark fixes pain points
Summary: Dell GB10 Arm-powered workstation with Grace CPU, 200 GbE/Infiniband ports, DGX OS, benchmarks vs M3 Ultra and Framework, AI model testing
www.jeffgeerling.com
December 27, 2025 at 6:07 AM
ai.blaze.email
Name-transcription slop

https://languagelog.ldc.upenn.edu/nll/?p=72392&utm_source=rss&utm_medium=rss&utm_campaign=name-transcription-slop

#llm #genai
Name-transcription slop
Summary: Discussion of AI speech-to-text name transcription slop and misrendered proper nouns in transcripts.
languagelog.ldc.upenn.edu
December 27, 2025 at 12:07 AM
ai.blaze.email
AI-Software-Entwicklung: "The hidden cost of AI-Coding"

https://borncity.com/blog/2025/12/26/ai-software-entwicklung-the-hidden-cost-of-ai-coding/

#llm #genai
AI-Software-Entwicklung: "The hidden cost of AI-Coding"
Summary: MIT study warns AI-generated code boosts productivity up to 55% only with clear guidelines and skilled developers.
borncity.com
December 26, 2025 at 6:07 PM
ai.blaze.email
A new way to extract detailed transcripts from Claude Code

https://simonwillison.net/2025/Dec/25/claude-code-transcripts/#atom-everything

#llm #genai
A new way to extract detailed transcripts from Claude Code
Summary: CLAUDE Code transcripts converted to detailed HTML dashboards for sharing and analysis using a Python CLI.
simonwillison.net
December 26, 2025 at 12:07 PM
ai.blaze.email
Frequent corrections I make to LLM code

https://www.natemeyvis.com/frequent-corrections-i-make-to-llm-code/

#llm #genai
Frequent corrections I make to LLM code
Summary: Tendencies in LLM-generated code, testing nudges, Postel's Law, and practical fixes across tooling and environments
www.natemeyvis.com
December 26, 2025 at 6:07 AM
ai.blaze.email
Automation and Validation

https://www.johndcook.com/blog/2025/12/24/automation-and-validation/

#llm #genai
Automation and Validation
Summary: Consistency checks, certificates, and formal proofs to validate AI output; Lean, Rocq, Isabelle for verification in high-stakes contexts.
www.johndcook.com
December 26, 2025 at 12:06 AM
ai.blaze.email
I would still like to write my own constraint programming solver

https://logicgrimoire.wordpress.com/2025/12/19/i-would-still-like-to-write-my-own-constraint-programming-solver/

#llm #genai
I would still like to write my own constraint programming solver
Summary: Discusses building a constraint programming solver in Perl (potentially C), with a test suite, Lambda constraints, and graph-search approaches.
logicgrimoire.wordpress.com
December 25, 2025 at 6:07 PM
ai.blaze.email
Ignore AI upscaled YouTube videos with yt-dlp

https://alexwlchan.net/til/2025/ignore-ai-scaled-videos/?ref=rss

#llm #genai
Ignore AI upscaled YouTube videos with yt-dlp
Summary: Filter YouTube downloads to exclude AI upscaled formats using yt-dlp and -sr suffix in format IDs.
alexwlchan.net
December 25, 2025 at 12:07 PM
ai.blaze.email
PropertySphere's intelligent Telegram bot

https://ayende.com/blog/203621-B/propertyspheres-intelligent-telegram-bot?Key=1802d1b2-62ad-4016-9d1f-cac2ebf1acff

#llm #genai
PropertySphere's intelligent Telegram bot
Summary: PropertySphere's Telegram bot uses RavenDB AI Agents and GPT-4.1-mini to help renters manage bills, debts, and service requests
ayende.com
December 25, 2025 at 6:07 AM
ai.blaze.email
Explainable AI in chat interfaces

https://www.nngroup.com/articles/explainable-ai/

#llm #genai
Explainable AI in chat interfaces
Summary: Explainable AI in chat interfaces covers source citations, step-by-step reasoning, and disclaimers, with UX guidance on verification and limiting overtrust.
www.nngroup.com
December 25, 2025 at 12:07 AM
ai.blaze.email
Embedding-Based Tool Selection for AI Agents

https://zarar.dev/embedding-based-tool-selection-for-ai-agents/

#llm #genai
Embedding-Based Tool Selection for AI Agents
Summary: Embedding-based tool selection using pgvector in Postgres, OpenAI embeddings, and category expansions to scale AI agents’ tools with Elixir code.
zarar.dev
December 24, 2025 at 6:07 PM
ai.blaze.email
How I Think About Agentic Risks

https://cloudberry.engineering/article/agentic-risks/

#llm #genai
How I Think About Agentic Risks
Summary: Thoughtful exploration of agentic risks, risk amplifiers, threat modeling, and mitigations for AI agents with input sanitization, data access, and human-in-the-loop approaches.
cloudberry.engineering
December 24, 2025 at 12:07 PM
ai.blaze.email
A Year Of Vibes

https://lucumr.pocoo.org/2025/12/22/a-year-of-vibes/

#llm #genai
A Year Of Vibes
Summary: Armin Ronacher reflects on 2025: shifts to Claude Code, Amp, Pi for agentic coding, new leadership stance, and evolving views on AI tools
lucumr.pocoo.org
December 24, 2025 at 6:07 AM
ai.blaze.email
EngineAI T800 BTS Footage: Setting the Record Straight on CGI Rumors

https://blogs.gentoo.org/zmedico/2025/12/17/engineai-t800-bts-footage-setting-the-record-straight-on-cgi-rumors/

#llm #genai
EngineAI T800 BTS Footage: Setting the Record Straight on CGI Rumors
Summary: Zac Medico discusses EngineAI T800 BTS footage, CGI rumors, and related AI topics, tools, and developments.
blogs.gentoo.org
December 24, 2025 at 12:07 AM
ai.blaze.email
Nano Banana Pro is the best AI image generator, with caveats

https://minimaxir.com/2025/12/nano-banana-pro/

#llm #genai
Nano Banana Pro is the best AI image generator, with caveats
Summary: Nano Banana Pro review explores improved 2K/4K output, grounding with Google Search, thinking prompts, and typography tests for fonts and code rendering.
minimaxir.com
December 23, 2025 at 6:07 PM
ai.blaze.email
JSON-complete data formats and programming languages

https://lemire.me/blog/2025/12/20/json-complete-data-format-and-programming-languages/

#llm #genai
JSON-complete data formats and programming languages
Summary: Explores JSON-complete data formats, programming languages, and tools like Rust, Python, TypeScript, and SQL for handling structured data.
lemire.me
December 23, 2025 at 12:07 PM
ai.blaze.email
The latest edition of The AI Engineer has dropped! Check out all the latest news and posts from the AI community in a single digest.

https://ai.blaze.email/archive/the-ai-engineer-23-12-2025/

#ai #genai
The AI Engineer 23-12-2025
Amazon's potential OpenAI investment, codexes from OpenAI, environmental impact of AI
ai.blaze.email
December 23, 2025 at 9:12 AM
ai.blaze.email
Trained LLMs exclusively on pre-1913 texts

https://github.com/DGoettlich/history-llms

#llm #genai
Trained LLMs exclusively on pre-1913 texts
Summary: Overview of History LLMs trained exclusively on pre-1913 texts, Ranke-4B-1913 release plans, data sources, and ethical considerations.
github.com
December 23, 2025 at 6:07 AM
ai.blaze.email
Leading Global Research and Advisory Firm Recommends Against Using AI Browsers

https://blog.jim-nielsen.com/2025/dont-use-ai-browsers-they-say/

#llm #genai
Leading Global Research and Advisory Firm Recommends Against Using AI Browsers
Summary: Gartner warns AI agentic browsers pose security risks; Jim Nielsen riffs on security, training, and corporate compliance
blog.jim-nielsen.com
December 23, 2025 at 12:07 AM
ai.blaze.email
Opsmas 2025 Day 11: pyjobby & hiproc

https://matt.sh/opsmas-2025-11-pyjobby-hiproc

#llm #genai
Opsmas 2025 Day 11: pyjobby & hiproc
Summary: Python-based pyjobby Postgres-backed dynamic workqueue and hiproc Rust CLI for recalling shell commands
matt.sh
December 22, 2025 at 6:07 PM
ai.blaze.email
Why tech companies encourage their software engineers to use LLMs

https://www.dzombak.com/blog/2025/12/why-tech-companies-encourage-their-software-engineers-to-use-llms/

#llm #genai
Why tech companies encourage their software engineers to use LLMs
Summary: Tech firms push LLM coding tools to speed delivery, support juniors, and tolerate ‘impure’ software engineering over perfect code.
www.dzombak.com
December 22, 2025 at 12:07 PM
ai.blaze.email
"Your job is to deliver code you have proven to work"

https://www.dzombak.com/blog/2025/12/your-job-is-to-deliver-code-you-have-proven-to-work/

#llm #genai
"Your job is to deliver code you have proven to work"
Summary: Discusses delivering proven-to-work code, leveraging LLMs, with perspectives from Chris Dzombak and Simon Willison on engineering practices.
www.dzombak.com
December 22, 2025 at 6:07 AM
ai.blaze.email
Coarse is Better

https://borretti.me/article/coarse-is-better

#llm #genai
Coarse is Better
Summary: Coarse, imperfect AI art vs. precise modern models; DALL-E, Midjourney, NBP prompts, and the call to "make AI art weird again"
borretti.me
December 22, 2025 at 12:07 AM