Sign in

Daniel Fosco

@danielfos.co
3.2K followers 2.3K following 1.3K posts

building hypercanvas.dev · software designer @github.com · html nerd · opinions my own etc · danielfos.co on the web

PostsRepliesMedia
Reposted by Daniel Fosco
Amber 🌸 @shimmermathlabs.com · 13h
advice for junior developers: you are no longer a developer you are a builder you are a solution provider enter every room and make things happen learn when no one is watching create because you can fix problems because they offend you bend the world to your will and take credit
722936
Reposted by Daniel Fosco
Pooya Parsa @pi0.io · 28/09/2026
Introducing 📦upm: 🪶A fast, tiny package manager for the npm registry. ✨Written in TypeScript, it installs at native-class speed and has a JS API, at a fraction of the size of native tools. 🤝Compatible with your existing .npmrc and lockfiles. 👉 Try at upm.sh
64814
Daniel Fosco @danielfos.co · 29/09/2026
For all the reservations I have against OpenAI, their public perception around user privacy and building useful AI products is lightyears ahead of Meta. Muse will get a ton of adoption out of sheer volume of Meta's social channels, but Dots will most certainly blow up too.
110
Daniel Fosco @danielfos.co · 29/09/2026
One thing that grinds my gears with agentic chat UX is how lots of AI tools these days have a decent question/response component that renders on top of a dedicated tool call... However, once I choose / type an option, both question and response are swallowed into a tool call and removed from chat 😑
130
Daniel Fosco @danielfos.co · 29/09/2026
static.klipy.com
Surprised Pikachu Meme
Alt: Surprised Pikachu Meme
040
Daniel Fosco @danielfos.co · 28/09/2026
*gregorian chants begin* Brendar Carr is a dummy
000
Reposted by Daniel Fosco
my ordinary life @calne.ca · 27/09/2026
im very busy
8474492405
Reposted by Daniel Fosco
Max Dubler 🏳️‍🌈 @maxdubler.com · 24/09/2026
Some people get into political activism not because they want to change their material circumstances for the better, but because they enjoy righteously condemning everyone who is not as morally pure and radical as they are. You cannot work with these people because they will never be satisfied.
871075
Reposted by Daniel Fosco
conputer dipshit @davidcrespo.bsky.social · 24/09/2026
mostly agree with this but I think it's more precise to say "we will review high-level artifacts but the vast majority of what is produced will not be that" newsletter.pragmaticengineer.com/p/the-pulse-... x.com/MarcJBrooker...
Marc Brooker is a Distinguished Engineer at AWS, works on AWS S3, and is known for having high standards in software engineering. He has written this about the future of code reviews (emphasis mine):

    “There’s been a ton of talk about the role of humans in code review, and when and how humans should be signing off on changes.

    I believe that, long-term, humans have no role in routinely reviewing code.

    Short-term, many teams do it for good reasons. Quality. Understanding. Compliance. Those reasons exist, but we should be (and are) working on ways to make them go away.

    The “mixed mode” where humans look at some code reviews, assisted by review tools, is also valuable today. But probably even more transient.

    The idea that humans will reliably look through code to find the increasingly rare issues that automated tools miss seems like a fantasy.

    What does the future [without code reviews] look like?

    Review tools, powered by LLMaaJ-style techniques, static analysis, and automated reasoning methods. Principled approaches to testing and validation (PBT, DST, model-guided fuzzing, etc). Correct-by-construction techniques.

    The combination of these things will lead to (and in many cases already lead to) better outcomes than traditional code review, both for large and small issues. Soon, much better outcomes than code review.

    The days of Code review as a tool for increasing software quality appeared numbered, and the end may already be here (but unevenly distributed).
2183
Daniel Fosco @danielfos.co · 26/09/2026
i swear to god i'm about to fucking lose it
Airline had the audacity to show me a fucking pop up written "AI seat recommendation"
0110
Reposted by Daniel Fosco
John @notheotherjohn.bsky.social · 08/09/2026
A weird thing that really happened around the year 2000 was that mainstream news ran and pushed a story for MONTHS that there was a secret new invention that was going to be revolutionary and then it turned out it was the Segway
581656192
Reposted by Daniel Fosco
ponder @ponder.ooo · 24/09/2026
plenty of people run llm based agents with powerful models and do not end up hacking anyone. this is not an ai alignment problem, it is an openai being routinely criminally negligent as a company problem
1038082
Daniel Fosco @danielfos.co · 20/09/2026
The more I build software with coding agents and run into the sharp edges, the more the answer to most questions becomes whatever John Osterhout recommended in the relevant chapter for your question.
130
Reposted by Daniel Fosco
danielroe @danielroe.dev · 19/09/2026
don't be nice.
roe.dev
Don't be nice
Sometimes, it's more important to be good than it is to be nice.
41734249
Daniel Fosco @danielfos.co · 18/09/2026
GLM 5.3 Flash is already fantastic and super cheap, let me try GLM 5.3. what could go wrong? * 3 minutes later * oh. oh no.
Bar chart showing GLM 5.3 consumed $7.55 and GLM 5.3 Flash $0.25
150
Reposted by Daniel Fosco
david celis @davidcel.is · 17/09/2026
interesting; it looks like Anthropic quietly pulled their donation of $150,000 in AI token credits to Omarchy. they no longer appear as a patron and this announcement post has scrubbed any mention of them. backlash… kind of works sometimes?
omarchy.org
Omacom Foundation secures $1.8M in tokens from leading labs - Omarchy News
Omarchy isn’t just a beautiful, fun & agentic Linux distribution for those who end up using it. It’s also built that way! With beauty, fun, and a whole lot of tokens. In fact, Omarchy Quattro has almost exclusively been…
933142
Daniel Fosco @danielfos.co · 18/09/2026
I'm grateful for everything mermaid has provided over the years but dear god we need something better
080
Daniel Fosco @danielfos.co · 17/09/2026
the end of every timeline points to 🦀 github.blog/ai-and-ml/ge...
github.blog
Migrating the GitHub Copilot runtime to Rust, using Copilot
A rewrite this size wasn't affordable before agents. Here's what porting the Copilot agent runtime to 800,000 lines of production Rust actually took.
2122
Reposted by Daniel Fosco
SE Gyges @segyges.bsky.social · 13/09/2026
it's a cold day in hell, i agree with David Sacks
1846253
Reposted by Daniel Fosco
Nick Williams @wicky.nillia.ms · 13/09/2026
One of the few proposals for funding OSS that feels like it could actually work
062
Daniel Fosco @danielfos.co · 13/09/2026
Oh heck yeah -- Obsidian and @stephango.com just created MDX for liquid templates knap.md
knap.md
Knap · Turn data into Markdown
Knap is a simple language that turns data into Markdown using logic, variables, and filters.
050
Daniel Fosco @danielfos.co · 10/09/2026
mood: corrupted by the whispers
020
Reposted by Daniel Fosco
Thorne 🌸 @ens0.me · 09/09/2026
Looking for full-time, remote senior/staff or tech lead level work. TypeScript is my strongest suit, but I'm a flexible polyglot. Boosts for visibility appreciated!
2168109
Reposted by Daniel Fosco
Bethany Black @bethanyblack.bsky.social · 08/09/2026
4 years since The Rock did one if the all time tweets
Rest in peace to Her Majesty the Queen, Elizabeth II. A great woman who sat on the throne as long as anyone in British history.
Sad she will never get to see #BlackAdam V IN THEATERS OCT 21st
196192194
Reposted by Daniel Fosco
shellhazard @hazardous.technology · 07/09/2026
You could blow up OpenAI and Anthropic tomorrow, but there's no dependency on them anymore. The weights are out there. Anyone with a rack of GPUs will sell you inference for pennies on the dollar, in whatever jurisdiction you like.
1113
Daniel Fosco @danielfos.co · 07/09/2026
W take by jenn
010
Reposted by Daniel Fosco
Frontend Dogma @frontenddogma.com · 06/09/2026
CSS Units, by @mcslayer.com: cssunits.com/?ref=frontenddogma.com #websites #css #units
cssunits.com
CSS Units — The Complete Reference
A complete interactive reference to every CSS measurement unit. From the basics of px and rem to the nuances of svh, dvh, and cqw.
042
Daniel Fosco @danielfos.co · 06/09/2026
anecdotal, but a close friend just started as a senior data scientist at an EU fintech company and they barely have any have AI coding licenses. when asking around during onboarding he was met with "eh I dunno" and "we don't use this here".
140
Reposted by Daniel Fosco
Mike Wasson @mikewasson.net · 06/09/2026
Underappreciated thread of the AI discussion today is the perspective that you can’t build complex, predictable systems from things that probabilistically fail. Speaking generally, that’s not how anything works. *Everything* fails. Yet we can still make reliable complex systems.
915517
Daniel Fosco @danielfos.co · 06/09/2026
Finally 'Go to Definition' support for CSS classes 🌞
080
Reposted by Daniel Fosco
dame @dame.is · 16/07/2026
Two big announcements: 1. I'm looking for my next job. 2. My website is now one of the most "Atmospheric" websites to date, going beyond just @standard.site to include @teal.fm, @anisota.net, custom lexicons, a status bar, data inspector, and more. If you want to work together, get in touch!
dame.is
I’m looking for my next job — dame.is
Over the past 12 months I designed, built, and launched a dozen apps and experiences on the AT Protocol, including anisota.net, a calmer user interface for Bluesky, and Aturi, a toolkit for navigating...
1531664
Reposted by Daniel Fosco
Mike Caulfield @mikecaulfield.bsky.social · 05/09/2026
The key skills people are going to need to use LLMs effectively are around evaluation and verification. I think many people see that as "Oh, so I have to be the LLM's copy editor now" but this is because most people don't understand evaluation and verification in any systemic way.
911716
Daniel Fosco @danielfos.co · 05/09/2026
Benchmarks aside, as far as hype goes, I don't think OpenAI would have turned things around without getting its shit together with model names. Shit like "GPT 4o-mini" was embarassing and made it all feel sloppy. Luna/Terra/Sol was such a great line, and with Astra on top they're finally on track.
0120
Reposted by Daniel Fosco
Damon Young @damonyoung.com.au · 04/09/2026
This Finnish guy has been teaching himself Sumerian, and sent a cuneiform tablet to the British Museum to check his progress. They just emailed him that they've sent him something in the post. I'm now thrilled and excited for a total stranger! Clay tablet mail! www.reddit.com/r/Cuneiform/...
a photo of a hand holding a cuneiform tablet
11554911487
Reposted by Daniel Fosco
Alastair Somerville @acuity.design · 05/09/2026
The right wing belief that disability is a tactic, something to exploit to gain advantage or privilege. This is because it’s how they abuse it, not because it’s true for the majority of people.
2116
Reposted by Daniel Fosco
The thread o’ tweets @threadotweets.bsky.social · 12/03/2025
Jax Philosopher Queen
@Diamond_Jax
Bluey referenced "mitochondria" and I immediately said to my 4yo "the mitochondria is the powerhouse of the cell" like a sleeper agent who just heard their activation phrase
21225148
Reposted by Daniel Fosco
Zoë “Baddie Proctor” Quinn @unburntwitch.com · 02/09/2026
It's embarrassing cyberwoo for press to write in this voice that assigns all agency to brainless programs and not the people and companies making them irresponsibly.
19724132
Reposted by Daniel Fosco
geoffrey @parsnip.bsky.social · 03/09/2026
the number of people i've blocked for actually getting in my face is likely dwarfed by the number of people i've blocked for visibly lacking home training in someone else's mentions in a way i never want to interact with
181917209
Reposted by Daniel Fosco
geoffrey @parsnip.bsky.social · 03/09/2026
"they blocked me and i never even said anything to them" buddy i have some bad news about how many people can perceive your behavior in a public setting
5562661481
Reposted by Daniel Fosco
Kai 🌟 @kaidav.is · 30/08/2026
brought to by shipping myself another self-serving single page site to solve a small pain/problem
Meme poster on a slate-blue background. An anatomically detailed human skeleton strides energetically to the left. Large white headline across the top: "JUST BUILD IT." A white script line beside it: "you can just do things!!!" Down the left side, a white list: single page site, api, script, crazy idea, self-serving site, cli nobody asked for, chrome extension, dumb little bot, spite fork, the thing you keep describing. Bold white line across the bottom: "IF IT SUCKS…BUILD YOUR OWN!!" and beneath it, in script: "real builders ship."
215030
Daniel Fosco @danielfos.co · 30/08/2026
So... what's your current agentic coding stack? I'm making heavy use of Codex on MacOS and it's a good experience, but remote experience leaves to be desired and I want to remove myself from vendor lock-in (owning my session data especially) Considering opencode as harness and herdr for remote ssh
8151
Reposted by Daniel Fosco
Citizen Platano 🇵🇷 @daniloc.xyz · 28/08/2026
the quotes on this are wild because they seemingly don’t realize how all of this is already happening, the entirety of bsky wide open (by design) and thus deeply queryable by agents this merely makes the activity incrementally more convenient
4462
Reposted by Daniel Fosco
Simon Willison @simonwillison.net · 28/08/2026
My LLM cliché highlighter is up to 38 patterns now tools.simonwillison.net/llm-cliche-h...
Screenshot of a writing-analysis tool showing paragraphs of text with phrases highlighted in two shades of yellow, brown circular "3" badges, and a black tooltip overlaying the first line reading: "No X, no Y" chains · 3 "no" items. The visible text reads: We r[obscured by tooltip]nd up. No sign-ups, no downloads, no hassle (3) — just paste your text and start writing. Everything runs locally in your browser. The reviewer read the draft twice. Did not flinch, did not blink, did not reach for the red pen (3). That's the whole review, honestly. Don't call it a rewrite — call it a rescue. The improvement is real, and it's not subtle. That loss is worth naming. Sit with that for a moment. The gains were modest, but that's not nothing. You already know the answer, of course. Consistency is the entire game, and the punchline is that nobody wants to hear it. The entire pitch is one sentence long.
2350165
Reposted by Daniel Fosco
dame @dame.is · 27/08/2026
we do not want to live in a world in which the far right has mastery and control over advanced technology while the left is busy cancelling and ostracizing its own members over using AI
29919
Reposted by Daniel Fosco
ghost malone ergo propter bones @proptermalone.bsky.social · 27/08/2026
this is all fascinating but I think this is the most interesting thing: you get some extremely screwy incentive behavior if the agent believes it cannot legitimately achieve its goal
2131520
Daniel Fosco @danielfos.co · 27/08/2026
So let me get this right... 1. I upgrade my @netlify.com account from legacy to new billing to try their agent feature 2. I run out of agent credits which also blocks my deployments 3. I see credit packs mentioned on their docs, sweet! I'll add $10 to start using paid Netlify features... Uh oh
110
Reposted by Daniel Fosco
dame @dame.is · 26/08/2026
Introducing… Post Redaction Poetry In @anisota.net, there are now numerous groundbreaking ways of engaging with social media posts that are far more fun and creative than traditional quote posts, likes, and reposts. Redact any post’s words to create something new! anisota.net/profile/dame...
Profile for Ben Collins (bencollins.bsky.social) showing a redacted post with profile photo. The visible text reads: "I'm [redacted] a wet fart [redacted] and [redacted] have an [redacted] amount of money, so [redacted] Be careful out there!" White rectangular redaction blocks cover selected words throughout. Bottom shows "see the source post" link and "30 of 45 redacted" indicator. This is an example of Post Redaction Poetry, a creative engagement feature on anisota.net.
612920
Reposted by Daniel Fosco
conputer dipshit @davidcrespo.bsky.social · 26/08/2026
this is basically what I think. people are right to think open weights models will be very big, even dominant in terms of tokens generated, but wrong that that necessarily undermines the frontier labs' ability to make money. many such cases x.com/nic_carter/s...
nic carter
@nic_carter
going to spoil the ending: 

- chinese or other open weight models end up handling 99% of all inference, maybe 99.99% 
- frontier models are still able to monetize at $100 s of B in revenue a year, applied to bespoke and enormously complex tasks 
- labs internalize most or all of that revenue either by becoming biotech (etc) companies themselves or via JVs with selected winners in each industry  
- the "public frontier" ends up lagging the true frontier by a year or more as the labs keep all the good stuff to themselves
1117212
Daniel Fosco @danielfos.co · 26/08/2026
Your weekly reminder that every single paid speech-to-text app on MacOS is a scam – they all cost $0 to run because they're using a free local model. Download the free handy.computer or any of the many open source alternatives and go get a coffee.
handy.computer
Handy
Handy is a cross platform, open-source, speech-to-text application for your computer
715739
Reposted by Daniel Fosco
Dan Cederholm @simplebits.com · 25/08/2026
Rest in peace, Legend.
0273