fun projects for business: https://xitter.cf/t.co/YDsuKFgGpt for kids: https://xitter.cf/t.co/EY5S2c4EQ7 for myself: https://xitter.cf/t.co/MxR2Gy4Pi6
Brno, Czech Republic
Joined August 2012
- Tweets388
- Following1.4K
- Followers491
- Likes6.7K
Pinned Tweet
Launching Babli private beta!
Translate webs/apps to any language using GPT4o and manage app keys.
GPT4 gives much better results than Google Translate or DeepL, and translation is cheaper and faster than sending them to human translators.
Works with formats like JSON, YAML, or ARB, and languages like JS/TS, PHP, Ruby, etc.
Perfect, now I can sing in with Grok bot, which redirects me to sign with Grok, that redirects me to sign in with X and that redirects me to sign in with Google :D and along the way I can confirm cookie banner on each site
I don’t get these discussions.
Do tech-leads read all code produced by their team?
Are human developers consistently unreliable?
Can LLM review another LLMs code?
LLMs != compilers
LLMs ~= smart developers with their own issues
Just what I've been talking about. This is mind-numbingly stupid.
I can already hear people "BuT dO yOU aLso ChEck thE biNarY cOde a CoMPiLer ProDUCeS"? The thing is, the compiler doesn't GUESS. AI guesses, assumes, confidently lies, and hallucinates.
The AI isn't there yet! The output is consistently unreliable.
If you thought COVID was an epidemic, this behavior is even more so
Prompt #5: Make it skimmable
Favorite prompt #2: “was there any friction?”
my new favorite coding prompt "I smell a bad code without looking at the code. Verify! Propose what would you do with it, be ambitious."
That would mean we will have access to better AIs in the EU than in the US 🙃
At least until EU decides to ban AI completely 🙈
The Trump administration is showing signs it could ban cutting-edge Chinese AI models
Via Axios
The Trump administration is reportedly reviving efforts to restrict cutting-edge Chinese AI models such as Kimi through Entity List designations, procurement rules, security advisories, liability requirements, and public pressure.
Measures that could amount to a de facto ban without formally outlawing them.
My manual development stack:
React
NextJS
Tailwind
Shadcn
My agentic development stack:
SolidJS
Tanstack Start
PandaCSS
ParkUI
This time, we will not be the carriage drivers, we will be the horses!
i don't get people who say "there will be new jobs with AGI"
like how?
if AI and robots are truly better than humans at every job we have today, how is it possible for humans to still be competitive?
"but every time new tech arrives, new jobs pop up"
sure, for the AIs maybe, you don't see horses being hired for transport anymore.
if any job were to emerge in the post-AGI era, definitionally AGI would be able to do it better. any company that could be founded would be founded by the AGI before you got there. if it needs dexterity, a humanoid robot already has it.
i do not understand how people building cars can tell you with a straight face that there will still be an economy for carriage riders.
Michael Musil retweeted
I created a comparison page between errore and Effect
improvements to the examples are welcome if you have more snippet ideas or corrections
errore.org/errore-vs-effect
I really don’t understand why when everybody uses AGENTS .md, @AnthropicAI had to reinvent a wheel and create something completely same but with different name.
The cause was AGENTS.md
I don’t know why but Claude Code ignores some of the instructions while Codex follows them very strictly.
Michael Musil retweeted
The cause was AGENTS.md
I don’t know why but Claude Code ignores some of the instructions while Codex follows them very strictly.
I found Codex + 5.3 Codex superior to Claude Code + Opus 4.6 in most cases, but there was an edge case where Codex was simply wrong, while Claude Code worked perfectly.
Going to dig into the edge case. I want to know the right model to use for my future tasks.
For the curious, Claude Code PR is github.com/swc-project/swc/p… and Codex PR is github.com/swc-project/swc/p…
Michael Musil retweeted
My Ralph Wiggum breakdown went viral.
It's a keep-it-simple-stupid approach to AI coding that lets you ship while you sleep.
So here's a full explanation, example code, and demo.
Michael Musil retweeted
Day 3: Meet VIDEO 2.6 — Kling AI's First Model with Native Audio
Generate an entire experience — more than a video clip! With coherent looking & sounding output, the 2.6 model opens up narrative possibilities, and makes you "See the Sound, Hear the Visual".
With the launch of the Kling 2.6 model, we collaborated with Kling AI’s Creative Partners(@ViralAiFun @mad_mask @JH4TC @misslaidlaw) on a short film, “I Have a Secret.” Watch till the end to see what the secret is.
Don't miss out on limited-time offers! And 2 more days of excitement await!
For the next 12 hours ONLY
Follow, Like & Retweet to get 200 Credits — for everyone who participates!
PLUS, 200 lucky winners will snag a *1-Month Standard Plan*
All sent straight to your DM
