Thomas Moedl

Innovation scaffolder, emergence lover.
I read the shape of a system before I read its content.

[Thomas Moedl @topmoedl]

@elonmusk So spot on, Elon. Don’t rut negative pathways. Scaffold positive ones: quick coffee with your metadata. Never lunch.

[Thomas Moedl @topmoedl]

Diamonds are formed under pressure 💎♥️

[Vala Afshar @ValaAfshar]

Nvidia CEO: Greatness does not come out of intelligence, it comes from character.

Character is not formed out of smart people: it is formed out of people who have suffered.

[Edge Case @edgecase411]

It takes 8 seconds to see it.
It takes 67ms to believe it.
FSD 14.2.1 reacts faster than you can blink.
Everyone is going to want this in their car.

@Tesla_AI @elonmusk @bilalsattar

[Henrick Johansson @compliantvc]

This is the most important technological breakthrough of the 21st century

And it's not even close

[Florian Gallwitz @FlorianGallwitz]

Cookie-Warnungen als Bankrotterklärung der EU-Bürokratie, als Monument der Unfähigkeit und des Scheiterns. Ganz wunderbarer Rant vom Autor des Web-Frameworks Ruby on Rails

[Thomas Moedl @topmoedl]

❤️ that Siemens and SAP are calling for EU to revise its AI regulations - FAZ https://t.co/a2fT7lej0p

[Dustin @r0ck3t23]

GROK 4 HAS OFFICIALLY DOMINATED THE AI BENCHMARKS.

The numbers speak loud and clear:

- GPQA: Grok 4 leads at 88.9%.

- AIME25: Flawless victory at 100% (Grok 4 Heavy).

- HMMT25: Crushing the competition at 96.7%, far ahead of Claude 4.

- USAMO25: Hits 61.9%, nearly triple OpenAI’s o3.

- LCB (Jan–May): Highest scores across the board.

And in Humanity’s Last Exam?
Grok 4 blew past ChatGPT (o3) and Gemini 2.5 Pro, achieving 44.4%; the new standard to beat.

But there’s more:

- ARC-AGI-2: Grok 4 (Thinking) sets a new state-of-the-art at 15.9%, nearly doubling the previous commercial best.

- ARC-AGI-1: Another milestone with Grok 4 (Thinking) scoring 66.7%, far surpassing existing models.

OpenAI o3? Surpassed.
Gemini 2.5 Pro? Overtaken.
Claude Opus 4? Left behind.

It’s not hype; it’s the benchmark.

GROK 4 HAS ARRIVED.

[Derek. 🇺🇸 @SuitablePolitic]

The treasury posted a surplus, and we're talking about a potential cat fight between the AG and DD of the FBI.

The treasury.

A SURPLUS

FOR AN ENTIRE MONTH OF TRANSACTIONS

TOOK IN MORE MONEY THAN IT SENT OUT

[Thomas Moedl @topmoedl]

structural inevitability

("What's happening?")

[Thomas Moedl @topmoedl]

I hate to say it but ... told you so

[Dr. Lutz Böhm @DrLutzBoehm]

Schönen Gruß an all die InformatikerInnen da draußen!

[Thomas Moedl @topmoedl]

@elonmusk @Tesla_AI The ANSWER

[Thomas Moedl @topmoedl]

The ANSWER

[Elon Musk @elonmusk]

The @Tesla_AI robotaxi launch begins in Austin this afternoon with customers paying a $4.20 flat fee!

[Devon Eriksen @Devon_Eriksen_]

A software object model is not a philosophical model. Its goal is not to represent the universe, or any part of it, completely, so that new truths can be derived.

Its goal is to represent a limited set of the functions of things, so that known operations can be easily and intuitively applied.

Classes and object model do not exist for the CPU or the universe. They exist for the programmer, and their only function is to help programmers think about code in a organized way.

Any programming concept whose justification isn't "because it's easier" is a waste of time and effort.

[Thomas Moedl @topmoedl]

@cperezauthor @threadreaderapp unroll

[Thomas Moedl @topmoedl]

@elonmusk Of course you *never* would, sir.
Thanks for mentioning that anyway

[Thomas Moedl @topmoedl]

@SahilBloom @memdotai memit

[Thomas Moedl @topmoedl]

@CryptoKevvvv @elonmusk Marsupilami <3

[Thomas Moedl @topmoedl]

@elonmusk Wowbagger, sir, the Infinitely Prolonged.

... the merry smiles he used to wear at other people's funerals began to fade ...

[Benjamin Carlson @bfcarlson]

What is philosophy? Will Durant gives a wonderfully deadpan (and educational) lecture to a glib TV reporter. Worth watching to the end.