Until recently my M1 Max 32GB 2021 was great until memory became the bottleneck.
My M1 Pro 8GB lasted me less than a year bcs with Docker dev stuff made it sluggish.
32GB was great for years until AI assisted productivity got so high with Codex.
Running 3 projects, 3 different worktrees, each either Docker compose or running/building Electron apps. So in this high AI era productivity my Mac is using insane amount of RAM and SSD space (bcs of all the worktrees)
Agree, with you but I'm still using 3.6 Flash because of tok/s/ latency/ uptime with high context. Tried Grok 4.6 and it was scoring lower on some internal benchmarks or slower.
Thanks for the post. For https://Document.bot (Kinda Cursor IDE for knowledge workers), im already trying to improve the harness (besides spotting bugs) using hillclimb experiments. More and more i'm using a AI harness engineering skill in Codex to further improve the app. This blog post helped me to improve the skill a bit.
im not sure if grok is on here, i think that's an X thing
but if you were to ask me, we're more fully featured than cloudflare, and anecdotally a ton more reliable in terms of error rates. back when it was brand.dev, i actually tried to use cloudflare's apis and it was quite unreliable, so we built our own stack instead
I'm not disagreeing with the article, but I think you are right that if you use AI in the right setting it can do much more than the average ChatGP that outputs Linkedin like crap:
interdisciplinary teams/person + SOTA AI + Right context (customer/ internal pain points, access to the code, focussed Hackathon) = amazing cool things.
The biggest benefit for me has been getting side projects done. Little bugs or todos that would take more time than I have to spare, but don’t detract from the project.
I’ve had Claude do stuff like interrogate IoT devices, look into proprietary file formats and even assisted debugging electronics from a text description and datasheets. It has stupidly broad knowledge as long as you watch out for things like getting part numbers/boards correct. Have you ever tried to ask for help on an electronics forum full of miserable greybeards? $20/mo for getting ballpark answers immediately is excellent value.
Do assess if you want the learning experience or not, and whether you care if the problem is solved for you.
For writing I get the appeal to bloggers, but I can’t stand reading anything that has the hallmarks of generative AI.
Same here. Little side projects and convenience tooling for $day_job that otherwise wouldn’t have gotten built for lack of time. Doesn’t need to be perfect, beautiful code for an audience of one. It just needs to work.
From X:
Today, we are releasing Rampart: a 14.7MB machine learning model designed to protect citizens’ privacy by redacting personal information directly in your browser before it gets sent to any server.
Same. Solving it by moving complex and sensitive data to an offline desktop app https://document.bot that support offline (self hosted) AI models (and optionally EU/ US AI providers).
However, it doesn't integrate yet with shared (org) drives.
