The Seattle "Eastside" is not really Seattle, it's Redmond, Bellevue, Kirkland, Issaquah and Sammamish - overpriced suburbs. I live in one of them and work at MSFT.
I mostly agree with your assessment of the culture. I have been here for 6.5 years and have never seen source tree, I work in Azure and everything is in Azure Devops and we commit and deploy chaotically.
I hope they would. I'm just comparing MSFT compensation for a SWE in the USA which might range from 150k-400k per year unless you get to quite high levels, with companies like Meta, nvidia, openAI, Anthropic where someone might make 1m a year with 2 years of experience due to crazy stock appreciation and very high offers for specific domain expertise
I don't really know the legality of it, but I can tell you that last year around April there was quite a big wave of sudden performance terminations within Microsoft and these people did not get severance. I know because I was a manager of one of these people and had to deliver the news. I know that many of these people had had a below 100 rating on their previous performance review, but we were unaware at the time that this meant a death sentence for these employees several months later when we got the email a few days in advance of the termination. It was hundreds, if not thousands of affected people.
Very few people are level 70. It means VP or Distinguished Engineer. Probably far less than <1% of the company. These people are making over $1m a year most likely, if not more. For Microsoft, that is big money. They don't pay like Meta or the AI labs. We are more comparable in pay to Amazon and other 2nd/3rd tier big tech companies.
It seems to depend on your level (above or below L65 which is first Principal rung). 12 months of health insurance is nice, but the pay seems to be between 3-9 months worth of pay depending on various factors.
It's something like
- 3 months base salary
- plus either 2 weeks or 1 month of pay per year of service (the former if you are at principal level or above)
- 12 months health insurance
The worst case scenario is not taking the offer then getting let go for performance, which means zero severance, health insurance terminated same day. This has been happening at Microsoft for the last few years, sometimes in large waves. Many of these people were not egregious non-performers, just people who had been marked as not quite meeting expectations, and thought they had another chance.
I just saw Julian Lage a few nights ago here in Seattle - it was so great. I went alone and soaked it in. I'm not even really a jazz fan, but as a lifelong guitarist he is undeniable. He had John Medeski on organ and piano and it was absolutely sublime.
I've been recording myself at home for many years (I write songs on guitar and sing, and some other instruments), and this really resonates with me.
When I sing and play guitar at the same time (which creates recording challenges with mic bleed and phase cancellation), and especially without a click track, there is an undeniable difference and there is magic in that difference. But the perfectionist in me wants to present a polished recording, so I end up "comping" (recording many times and picking the "best" bits), using a metronome, and a bit of gridifying, although usually just with drums if I played them. I draw the line at vocal tuning, I just can't.
This blog post is nothing new to me, I've had these thoughts a thousand times. But it's a great reminder. Especially in the age of AI. Our imperfection and vulnerability is our gift.
I wonder how good it sounds. I could see my kid using this for Spotify, I current have to keep reminding her to AirPlay her iPad to our old homepods so she isn't always listening to tiny, shrill audio.
It seems totally inappropriate to play an AI video in court, but it seems like a waste of resources to waste more court time on this. But also, how can you murder someone and get only 10 years in prison?
My team has been struggling to understand whether or not we should do "spec-driven" development or not. It makes a lot of sense to me to have one developer iterate with the model/harness to generate a markddown document that is a high-level of what will be implemented, and then have the team PR review it before and agent attempts to do the actual implementation work. Do you think this is a good practice?
This is so true. After 20 years in the tech business I have so rarely seen perfect execution. It's mostly scrambling and chaos and a miracle anything works in the first place.
Sad! I really enjoyed going to the Kusama exhibit with my (now) wife years ago, and my daughter had a Kusama phase when she was about 5 years old and dressed up like her with a hot pink wig and polka dot outbit.
It would be very cool to have an Obsidian viewer or something.
On many occasions I have compiled notes, sometimes manually created sometimes LLM curated, and put them into a .epub file to send to my kindle so I can read as I am out walking.
It would be cool to skip a few steps there, and have "live" documents at the ready, particularly in markdown if possible, without having to sync/publish etc
I'm absoltutely not defending the environmental decision, but these datacenter projects seem to be funded mostly by free cash flow (not borrowed money), from the hyperscaler profit margins.
I definitely wish I could use something like this. I am not a great note-taker, and am often distracted during meetings with actual work items, or side pings.
Many of us can't just arbitrarily record things though. I work in a highly regulated part of the industry with major data compliance standards.
Totally, Airwindows is a very cool project and his videos are fun to watch. I wanted to bring a nicer UI along with it, using JUCE, and more focus on live usability.
Is it lame that (especially for super long discussions), I often open a gemini tab and summarize the comments? I like reading HN comment threads, but I often only have a few minutes and just want to get the gist of the reactions from the community for some topic.
The gap between China's chip manufacturing capabilities and the USA's is only going to shrink, right? ASML obeys some export controls for their most sophisticated machines, but those machines are in China's backyard (Taiwan).
Taiwan manufactures the world's most advanced chips. CCP wants "re-unification" with Taiwan. AI may be THE key to world dominance. These are scary times.
Something that I think we will see soon (if it doesn't exist already), are programming languages that are specifically designed to be used with LLMs doing the coding, with natural language as the primary input. Maybe LLMs will get so good that it's just not necessary, but it's hard to imagine a world in which this new paradigm of most users using coding harness (or event totally autonomous agents) doesn't end up having serious influence on language design itself.
One outcome I would love to see is average programmers (or even vibecoders) being able to create performant applications. Imagine if the issue of making an attractive cross-platform application didn't mean using a chromium-based Electron type framework.
Or, perhaps things we used to code in Java or C# can be more easily brought into existence in a lower-level language akin to Rust, but far more easy to create and maintain due to LLM assistance.
I'm using copilot cli because I work at microsoft and it's free for me. It has all the models. I have not loyalty to any of them. Even the copilot harness can be swapped out for claude, codex or whatever.