Nick DiRienzo

We hit an all time high in commits to main this week. Every PR is squashed so a commit represents a real change. Opus 4.6 with @conductor_build is a gamechanger.

anyone have a good emoji reaction for vibe coding?

@rosslazer worth

@nickdirienzo's poor sleep score was worth it

An engineer joined the team today and submitted a PR within ~3 hours. You don't need to be a big company to have good platform engineering.

@rosslazer trying to harden the network. accidentally taking down staging. everything is fine.

We will get through SOC2 in a week. Inshallah.

@rosslazer @equacc How many do we need? Hard to imagine ordering just a single pie.

@equacc Instead of a benchmark dataset, we should have @nickdirienzo's Italian cousin in NJ try to order a grandma's pie.

How can we take claims of AGI seriously if it can't even order a pizza? Read more in my short post rosslazer.com/posts/personal…

We’re hiring our 5th team member at @miragesecurity! Check the link below for all the details—and, just for fun, see if you can find the easter egg we tucked in there. 👀

@NotTuxedoSam @rosslazer hetzner, linux, sqlite, php

@rosslazer drop list

The most understated perk of doing a startup is being small enough that you can pick your best-of-breed tools.

@psjeffrey @getshayapp congrats on the launch! lets gooo 🚀

is there a way to not have this sidebar show up every single time i open a @googlesheets? @googledrive

Founder fellowship @southpkcommons applications close Friday! Time is running out. Take the leap. Embrace ambition. Get $1M in funding and the best team, community and program in -1 to 0. Become a worldbuilder. I'll help.

Meta just dropped Llama 3.1, and they quietly released a paper on new safety benchmarks for large language models (LLMs). Here's what you need to know about these safety tests and why their social engineering test has a glaring flaw. A 🧵

founder fellowship is a game changer. go apply 👇

@rosslazer What if we used irc

I miss IRC 😔

wooo they did the thing 🎉

@chuga @copilotmoney sad i missed this and couldn’t swing by! maybe next time!

Any @copilotmoney users near SOMA in San Francisco? I have a few stickers and will be around at a coffee shop for the next hour or so. DM for details!

@whelan_boyd the AI overlords will remember this kind gesture

Welp I just waved at a waymo for letting me cross the street 🫠

@lyndsysimon vscode & copilot. took me a bit to change how i code to make it more useful. have heard good things of cursor if you wanna try another ide.

Anyone using an AI coding assistant in their editor/IDE? If so, I’d love to hear what tool(s) you’re using, how it helps you, and any positive or negative experiences you’ve had getting to this point. I’m working on my Neovim setup. Gitlab Duo was a disappointment.

@rosslazer glhf registry.terraform.io/modules/twstew…

@nickdirienzo But can you run JS on the edge with this? How will you update the database with sub second latency?

Vercel is a tax on people who don't want to spend 3 hours learning how computers work.

𝗚𝗼𝗼𝗴𝗹𝗲 𝗜𝘀 𝗪𝗿𝗼𝗻𝗴. 𝗣𝗵𝗶𝘀𝗵𝗶𝗻𝗴 𝗦𝗶𝗺𝘂𝗹𝗮𝘁𝗶𝗼𝗻𝘀 𝗗𝗼 𝗪𝗼𝗿𝗸 𝗪𝗵𝗲𝗻 𝗧𝗵𝗲𝘆 𝗗𝗼𝗻’𝘁 𝗦𝘂𝗰𝗸. 𝗛𝗲𝗿𝗲’𝘀 𝗪𝗵𝘆: A recent debate ignited by Matt Linton’s Google blog post challenges the traditional tactics of cybersecurity training, specifically the efficacy of phishing simulations. It’s a provocative claim that questions the foundation of our current security practices. 🌵 In response, I’ve researched the topic deeply in a new blog post. Mira, our AI avatar, summarizes our critical findings: 🎬 𝗞𝗲𝘆 𝗧𝗮𝗸𝗲𝗮𝘄𝗮𝘆𝘀: • 𝗘𝘃𝗶𝗱𝗲𝗻𝗰𝗲 𝗦𝗽𝗹𝗶𝘁: Some research backs the effectiveness of phishing simulations, yet other studies suggest they may actually enhance susceptibility to real threats. • 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝗙𝗹𝗮𝘄𝘀: Not all studies are on equal footing. Mira highlights key methodological shortcomings that could skew results. 📊 𝗙𝘂𝗹𝗹 𝗔𝗻𝗮𝗹𝘆𝘀𝗶𝘀: Don’t just take our word for it—watch the video, read the blog, and look at the data. 👇 𝗪𝗮𝗻𝘁 𝗺𝗼𝗿𝗲 𝗶𝗻𝘀𝗶𝗴𝗵𝘁𝘀? Dive into the full blog post—link in the replies. @miragesecurity #Cybersecurity #Phishing #SecurityAwareness