There's a lot of excitement around vibecoding because it lowers the bar for shipping software.
I'm all about that.
But the cracks are showing in traditional software development lifecycle processes/systems.
Who is building the future of professional software engineering?
Would you attend a cohort course taught by o3/o4 ai chat models?
With peers who are interested in the same topics as you?
it's actually a needle point felted style throwback
gpt-4o image gen makes old avatars new again
new pfp, who dis
welcome @ziki.boo!
somewhere between "ok boomer" and "based" www.shruggingface.com/microblog/20...
why is og modest mouse so good?
objectively, they're kind of bad, but subjectively, they're so good
Do I know anyone in need of part-time product design help?
I have a very legit product designer friend looking for contract work, DMs open if you're interested!
GEN Z IS SO BACK
what is the actual relationship between building a great product, and success?
feels like there are many success stories for not-great products
does this thing work?
MVB: minimum viable buddhism
How does the watercolor fill work?
what's tiktok?
is teenage engineering premium mediocre?
what is that thing and where can I get one?
schelling points are the easy defaults that emerge from our unexpected ‘aha’ moments of bisociation
the worst thing about using cool new programming languages and frameworks is waiting for the LLM cutoff date to catch up with the present reality.
there's significant selection pressure to pick old tools because they're easy to use with llm code generation
Kubernetes is a charismatic trap.
how do i ban politics from my bluesky discover feed?
I've filtered out a bunch of keywords, but people just get past it by putting text in images...
other thing: in a world where test-time compute is the king of quality (seems increasingly likely with o3/qwq/qvq), hella fast inference is actually super important
one of them or both of them are going to get bought by frontier labs
personal take is that groq is more developer-friendly and has support for broader models, while their hardware is one generation behind cerebras
once the 2nd gen hardware is online, i think they might be the clear winner
but I guess we'll see
who will win between groq and cerebras?
cute interaction when creating a new @astro.build project 🥰
Why haven’t there been many new UX / interaction advancements for LLMs?
Feels like we’re stuck in the chat paradigm when there could be so much more.
In 2025, embrace non deterministic outputs, and leverage them to go where you never thought you could.
lollll my t2x tool somehow found its way onto HN a month after sharing news.ycombinator.com/item?id=4255...
www.shruggingface.com/microblog/20...
