Maciej Biłas

TIL: > ping example.com.localhost PING localhost (127.0.0.1): 56 data bytes 64 bytes from 127.0.0.1: icmp_seq=0 ttl=64 time=0.061 ms anything *.localhost resolves to localhost on macOS

@krzyzanowskim @tmuxvim […] ridge regression, random forest, XGBoost, and an MLP […] doesn't seem to relevant, does it? The claim is plausible, but I'd expect you to quote the article with a caveat about its age and limits of its applicability to LLMs.

@tmuxvim research do identify quality degradation over time, that is also playing a role nature.com/articles/s4159…

"Anthropic, OpenAl and Google release their new models with high quality from day one then slowly nerf them until the next model, so when the next model hits, its perceived as a bigger jump than it actually is" sounds right what's happening

@odd_joel @steipete Do you have plans to add Codex support as well? It’d make Moshi 10x more useful for me.

@maciejb @steipete nice discovery - running SSH from iPhone is a game changer once you get into it. I built Moshi for exactly this kind of workflow. it uses the mosh protocol so sessions survive wifi switches and phone sleep, plus Face ID for SSH keys which removes a lot of friction on mobile

TIL you can run SSH scripts natively on an iPhone. This is how I use @steipete’s great summarize.sh now.

@odd_joel @steipete Let me give it a try. I’ve been using Blink for a while now.

(I should, of course, quote and perhaps sanitize the input)

@thsottiaux Are those daily patterns or weekly? If daily, just expand to regions where your off-peak is middle of the day!

With Codex the there is quite the gulf in load between peak and off-peak times, and we would like to achieve more of a smoother traffic pattern as that would be a more optimal use of our compute. We have ideas, but curious what you all think we should do? Would more usage during off-peak and surge multiplier during peak times make sense?

Same; switching from Claude to GPT made me change my habits.

Here’s what I’d do if I was in charge of GitHub, in order: 1. Establish a North Star plan around being critical infrastructure for agentic code lifecycles and determine a set of ways to measure that. 2. Fire everyone who works on or advocates for copilot and shut it down. It’s not about the people, Im sure theres many talented people, youre just working at the wrong company. 3. Buy Pierre and launch agentic repo hosting as the first agentic product. Repos would be separate from the legacy web product to start since they’re likely burdened with legacy cross product interactions. 4. Re-evaluate all product lines and initiatives against the new North Star. I suspect 50% get cut (to make room for different ones). The big idea is all agentic interactions should critically rely on GitHub APIs. Code review should be agentic but the labs should be building that into GH (not bolted in through GHA like today, real first class platform primitives). GH should absolutely launch an agent chat primitive, agent mailboxes are obviously good. Etc. GH should be a platform and not an agent itself. This is going to be very obviously lacking since I only have external ideas to work off of and have no idea how GitHub internals are working, what their KPIs are or what North Star they define, etc. But, with imperfect information, this is what I’d do.

I, for one, am not upset that AI is taking my job. I’m grateful to have been one of the people who paved the road of computing in the barbaric times before Claude. Our heroic work was necessary to keep the computers running until the machines were ready to take over.

Does this mean Elon will stop telling Europeans how to vote?

I don't like tech CEOs trying to cause mass panic to the job market. Jensen and Elon too. Excavator + crane didn't mean that construction (shovel) jobs got cut down by 10x. We started building massive bridges and skyscrapers. Early numbers show just that (much more commits).

Local barber suggests everyone come in for a trim once a week

❗️Ground-based robotic systems this year could replace up to a third of 🇺🇦Ukrainian infantry on the line of combat contact — commander of the 3rd Army Corps Andriy Biletskyi “If we move in the direction of technological innovation, then this year, I am convinced, it will be possible to remove up to 30% of infantrymen from the line of combat contact, and in the near future — up to 80%.”

@levelsio Driving in cities is not fun for the most part. It’s fun when you go out of urban areas, or at least can be fun.

I think my next OpenClaw side project, once my home Internet is back up, is automating complains to @Orange_Polska about their fiber-to-home outages.

Warsaw has reduced its level of fine particulate matter (PM2.5), a particularly harmful type of air pollution, by 46% since 2010. That was the second biggest decrease among 19 global cities included in a new international report, behind only Beijing (48%) notesfrompoland.com/2026/03/15/war…

@levelsio You won’t force humans to use self driving cars. Most folks like driving too much. If any politician tried to– it would’ve been bad expenditure of political capital.

Now that I got my driver's license Seeing people drive like complete goons every day now I think our #1 priority should be to get every car brand to self driving ASAP on a worldwide gov mandate Humans are way too retarded to drive

@michuk @HarryStebbings Fun fact: a coworker who has a stellar education record complained to me that job candidates he interviews are scared of him. An unexpected perk of attending top universities, I guess.

@HarryStebbings Gaining skills has never been the main reason to study. People study to meet like-minded people, make friends, party. Top unis also get you status. Whatever you do in your life you'll always be Stanford/Caltech/Oxford alum. Don't waste the best years of your life working ;)

University has never been more worthless. Curriculums are unable to keep up with the speed of AI. Do not waste three of your most productive years of your life, gaining debt to learn skills that will be obsolete by the time you leave.

@SebAaltonen @TukiFromKL Having AI do a first pass on a PR is actually a better use of everyone’s time. As a reviewer I can focus on higher level aspects on the change and the direction the codebase is taking.

@TukiFromKL A tech lead does so much more than just 2 code reviews a day. That's a minor part of the job.

🚨 Are you paying attention right now? Claude just put a price tag on your senior developer. $15–25 per code review. AI agents that start Instantly The second you open a PR. > Your tech lead makes $200K/year That's ~$400 per code review assuming 2 per day > Claude charges $15 That's a 96% pay cut > And the AI doesn't call in sick on Monday The people protecting their $200K salaries by gatekeeping code reviews are about to have a very bad quarter.