mathiasmoser @mathiasmoser

mathiasmoser.at Wien, Österreich Joined January 2009

Tweets

1K
Followers

63
Following

583
Likes

2K

mathiasmoser @mathiasmoser

3 weeks ago

@taylorotwell nice - built in MCP Support would be great as well

2 0 0 203 0

View Details

marc friedrich @marcfriedrich

3 months ago

When you realize you country, your whole continent is fucked

555 1K 18K 1.3M 2K

View Details

Laracasts @laracasts

4 months ago

I'm done.

67 50 542 61K 201

View Details

@JackEllis But i still have not found my ideal setup. Currently i use "Ask", "Auto-Commit Message", "Autocompletion" within Jetbrains with GH Copilot. For all bigger agentic tasks i use OpenCode in a separate window. The AI produced files from OpenCode are reviewed within my Jetbrains IDE.

0 0 0 13 0

View Details

mathiasmoser @mathiasmoser

4 months ago

@JackEllis I can use my Github Copilot License in VS Code / Jetbrains and OpenCode. So OpenCode also gives me the flexibility to reuse my existing License which is nice.

0 0 0 38 0

View Details

mathiasmoser @mathiasmoser

4 months ago

@JackEllis My problem with Cursor and all the VS Code Forks is the underlying VS Code. I am a Jetbrains User for years now and i really love the way it works/feels. OpenCode allows me to stick with Jetbrains as my IDE. If you are a VS Code Guy i completely understand why you go with Cursor.

0 0 0 33 0

View Details

mathiasmoser @mathiasmoser

4 months ago

@JackEllis use OpenCode

1 0 1 66 0

View Details

mathiasmoser @mathiasmoser

4 months ago

@christophrumpel OpenCode is very good. Much better usablility than Claude Code in my eyes. You can esily switch to other Model Providers. The TUI is great and the progress they are making is sick, daily updates…

0 0 1 62 0

View Details

mathiasmoser @mathiasmoser

5 months ago

@mariorod1 @github @opencode great 👍

0 0 0 37 0

View Details

mathiasmoser @mathiasmoser

5 months ago

@jeffrey_way @RichKleinAI @thdxr i think clean code is beneficial for the ai as well - lets take DRY. If the Ai has to do only one change in a central place its faster / cheaper to change (less token) and less prone to errors.

0 0 3 26 0

View Details

Trending GitHub Repositories @trending_repos

5 months ago

Trending repository of the week 🏅 opencode by @anomalyco The open source coding agent. Last week: 13823 ⭐ Total: 63768 ⭐️ github.com/anomalyco/open…

0 1 37 4K 14

View Details

mathiasmoser @mathiasmoser

7 months ago

@enunomaduro what plan are you using? is the quota sufficient for your needs?

0 0 0 12 0

View Details

PJ @Prithvir12

8 months ago

If this Karpathy interview doesn't pop the ai bubble, nothing will. 10 brutal quotes: 1. LLMs don’t work yet They don’t have enough intelligence, they’re not multimodal enough, they can’t use computers, and they don’t remember what you tell them. They’re cognitively lacking. It’ll take about a decade to work through all of that. 2. When you boot them up, they always start from zero They have no distillation phase, no process like sleep where what happened gets analyzed and written back into the weights. 3. What’s stored in their weights is only a hazy recollection of the internet It's just a compressed blur of 15 trillion tokens squeezed into a few billion parameters. Their context window is just short-term working memory. 4. They’re good at imitation, terrible at going off the data manifold Too much memory, not enough reasoning. We need to strip away the memorized knowledge and keep the cognitive core: the algorithms, the magic of intelligence, problem-solving, strategy. 5. We’ve probably recreated a cortical tissue, pattern-learning and general, but we’re still missing the rest of the brain No hippocampus for memory. No amygdala for instincts. No emotions or motivations. 6. They memorize perfectly but generalize poorly If you give them random numbers, they can recite them back. No human can do that. That’s the problem: humans forget just enough to be forced to find patterns. 7. Anything truly new, code that’s never been written before, ideas that have no template; they stumble They’re still autocomplete engines with perfect recall and no understanding. Until we find that cognitive core, intelligence stripped of memory but full of reasoning, they’ll stay brilliant mimics, not minds.

316 803 6K 1.4M 6K

View Details

terminal @terminaldotshop

10 months ago

21 174 2K 103K 126

View Details

FK Austria Wien @FKAustriaWien

10 months ago

Welche klare Fehlentscheidung lag hier vor?

VAR Österreich @VAR_Oesterreich

10 months ago

#gakfak 51. | SR Entscheidung: Strafstoss & Handspiel | Grund des Checks: Handspiel? | Finale Entscheidung: kein Strafstoß #VAR #VARheit #AdmiralBL #VARÖsterreich

0 0 4 32K 0

37 4 132 42K 2

View Details

mathiasmoser @mathiasmoser

a year ago

@christophrumpel funny - got the same gift from my wife when i got 40. was an unforgetable experience. enjoy!!!

1 0 1 82 0

View Details

FK Austria Wien @FKAustriaWien

a year ago

🎈💜 #faklive 📸 @_annxsophie_

0 2 52 2K 0

View Details

Die Tagespresse @DieTagespresse

a year ago

Zu deppert, um ein Urteil zu lesen: Die FPÖ hat vergessen, die Urteile fristgerecht zu veröffentlichen und schenkt der Tagespresse damit 62.757 Euro. dietagespresse.com/urteilsveroeff…

66 159 1K 49K 21

View Details

Jeffrey Way @jeffrey_way

a year ago

Coming soon to a Laracasts near you.... Every Laracasts video ever...fully transcribed....converted to embeddings...ready to answer your questions...with direct links to the timestamped portion of the relevant video. 🥵