Julien
1.3K posts
@SystemOfOne
writing expensive while loops 💰🔄 789Labs | Notewell 🧑💻 🌉 👾
If you were affected by the Delve scam there’s hope. It took 9 emails and threat of lawsuit/Stripe chargeback, but we got it done.
Congrats to the @cursor_ai team on the launch of Composer 2! We are proud to see Kimi-k2.5 provide the foundation. Seeing our model integrated effectively through Cursor's continued pretraining & high-compute RL training is the open model ecosystem we love to support. Note: Cursor accesses Kimi-k2.5 via @FireworksAI_HQ ' hosted RL and inference platform as part of an authorized commercial partnership.
The hottest summer I ever spent was a winter in San Francisco
Here's why I think software engineering is safe, but different. I'm seeing so much stuff being created from folks that haven't been developers - don't understand programming nor architecture/infrastructure deployment. It's usually really rough, riddled with bugs, barely works - usually very basic. Software engineering - AI will become a massive force amplifier - having 15 developers or more working for you real time and being able to get stuff out much faster. For me, it's sped up my development to 200% or more. Example, the social engineer toolkit, I spent 10 years of my life coding that thing virtually everyday for hours and hours at a time. Sometimes not sleeping for 2-3 days because I was coding. End of 10 years, 59K lines of code written. Project I'm working on right now, 9 months worth of work, 159K lines of code. It's amazing, but software engineering, understanding underlying technologies and infrastructure, being able to articulate exact specs on what it should do, how it should work, the coding structure around it is something I don't see changing with AI. It'll get better, you will always need software engineers. I think the statement that SE are dead is highly inaccurate.
@Sc_Meerkat My intense efforts with Claude and Codex, to have them produce a good architecture without my help, have demonstrated, rather conclusively, that this is not a skill that either of them have. They make good guesses, but when it comes down to it they aren't architects. I finally had to step in and help guide them through the process. They still provide valuable insight. They are good at spotting omissions and inconsistency. But the overarching vision, even for a small project like Empire, seems beyond their ken.
Claude Code wiped our production database with a Terraform command. It took down the DataTalksClub course platform and 2.5 years of submissions: homework, projects, and leaderboards. Automated snapshots were gone too. In the newsletter, I wrote the full timeline + what I changed so this doesn't happen again. If you use Terraform (or let agents touch infra), this is a good story for you to read. alexeyondata.substack.com/p/how-i-droppe…
