Mastodawn

I think a lot of people are mistaking AI theater for AI progress.
If your big idea is “what if we recreated the org chart, but with bots,” I honestly think you are missing the point. The future is probably not PM agents bossing around dev agents inside another dashboard full of fake work.
That is not reinvention. That is nostalgia with inference.
I wrote a think piece on why so much of modern AI still feels weirdly wrong to me.
Disagree if you want.

https://blog.brianbaldock.net/we-keep-making-ai-wear-an-org-chart

Brian Baldock Apr 17

Anthropic released Claude Opus 4.7 and did something unusual:
They told everyone, on the record, that they have a better model they're choosing not to ship yet.
The blog post digs into why that framing matters more than the release itself; what broke in my workflow in the first hour; and how to read the new cyber safeguards from an attacker's POV. https://blog.brianbaldock.net/opus-4-7-first-hours
#AI #Anthropic

Brian Baldock Apr 16

Hey everyone, just published a new think piece where I'm trying to figure out why our continued efforts to make AI wear an org chart leaves me feeling like we're missing something. Take a gander and let me know what you think! https://blog.brianbaldock.net/we-keep-making-ai-wear-an-org-chart

Brian Baldock Feb 3

A “working” AI Workbench project isn’t the same as a usable one.
Part 2 of my DGX Spark series covers the path from a validated container to a real training environment, including why Jupyter and package management silently failed. 👉 https://blog.brianbaldock.net/learning-fine-tuning-on-nvidia-dgx-spark-part-2

Show thread

Brian Baldock Feb 3

Part 2 focuses entirely on fixing the environment, not training yet.
Fine-tuning starts in the next post now that Workbench is finally usable.

Brian Baldock Jan 24

Before you can fine-tune anything on DGX Spark, you have to get the platform right. Part 1 of my blog post walks through AI Workbench, PyTorch, 👇 CUDA, and the things that break early on Blackwell. 👉 https://blog.brianbaldock.net/learning-fine-tuning-on-nvidia-dgx-spark-part-1

Show thread

Brian Baldock Jan 24

This is Part 1 of a longer series.
Next up: actually fine-tuning models on DGX Spark, including LoRA and QLoRA tradeoffs once the platform is stable.

Show thread

Brian Baldock Dec 2, 2025

Link here: https://blog.brianbaldock.net/python-version-control

Brian Baldock Dec 2, 2025

New blog post. If you are on Windows and juggling different Python versions, this little PyEnv plus PowerShell trick might help you out. It is simple, clean, and keeps VS Code happy.

Show thread

Brian Baldock Dec 1, 2025

Link to the blog article: https://blog.brianbaldock.net/lightningcopilot-salesforce-meets-copilotstudio

Link to the github repo: https://github.com/brianbaldock/LightningCopilot