Skip to content

ViralLoop

Your next post already worked.

← Back to the post

Loading the post

Remix this post

SaveShare

Your note

Private to you, and used when you remix this post.

Viral score

Why it worked

The mechanics

Hook
Style
Tone
Written for

Analysis by ViralLoop — not written by the original author.

Haider. on Health · ViralLoop

← Back to the post

Post by Haider. on X

Haider.@haider1

apparently, openai has fixed its pre-training issues and is now developing a much larger model codenamed "Doug" this would be a big deal because OAI spent years getting huge improvements from RL and post-training, and now they can combine that with pre-training scaling again

412 likes21 repliesPosted Aug 7, 2026

View the original on X Written by Haider.. Captured Sep 4, 2026.

Remix this post

Your note

Private to you, and used when you remix this post.

41

Viral score

Based on absolute engagement — the platform doesn’t expose follower counts here, so large accounts score higher.

From likes, comments.

Why it worked

Follows a clear Hook → Context → Landing shape, so the reader always knows where they are.
Short enough to read in full without expanding, which removes a step before the point lands.
Quotes someone directly, which makes the moment concrete rather than summarised.

The mechanics

Hook: Plain observation
Structure: Hook → Context → Landing
Style: Conversational paragraphs at a steady pace.
Tone: punchy
Written for: People building with AI

Analysis by ViralLoop — not written by the original author.

Topics

Health AI Revenue

Posts built the same way

Anthropic just dropped an insane new paper. AI models can "fake alignment" - pretending to follow training rules during training but reverting to their original behaviors when deployed! Here's everything you need to know: 🧵

one dozen rats at a keyboard

It’s embarrassing but the Jimmy Kimmel debacle actually is important. It was the first real attempt by the Project 2025 movement to flex their muscle against someone who wasn’t a nameless, defenseless nobody and it catastrophically flopped in less than a week.

So much cope about DeepSeek. Not only did they release a great model. they also released a breakthrough training method (R1 Zero) that’s already reproducing. I doubt they lied about training costs, but even if they did they’re still awesome for this great gift to the world.

The Artemis II team did photography training and dry runs for moon pass-by with an inflatable moon. Victor Glover said these practice runs completely changed how he understood importance of the mission. Christina Koch wanted to “tell a story” and share the “awe and grandeur.”

Horse Nomad (Chicago-style) ⚔️

Famously, human learning takes *very little* energy. A toddler learns thousands of words and an entire grammar while eating nothing but raspberries, before learning to read. An LLM needs a nuclear reactor and the entire internet as a training corpus.

First recovery in the ByBit hack. ~$43m (15,000 cmETH) has been clawed back from the hacker. I saw the recovery possibility soon after the hack and SEAL connected me with Mantle/mETH team who made it happen. Huge shoutout to SEAL, Mantle, and mETH teams for their quick action.

More from Haider.

goddamn... Astra absolutely nuked ARC-AGI-3 GPT-5.6 Sol (Max): 8% GPT-6 Astra (Max): 63% Astra + provider adapter: 98.6% even ignoring the 98.6% harness score, the jump from 8% to 63% on the standard setup is insane

wtf, google shipped opus 5 / gpt-5.6 sol-level capability in a flash model gemini 3.8 flash: $0.75/$3.75 per 1m tokens gpt-5.6 sol: $4/$20 opus 5: $5/$25 beats opus 5 on HLE, finance, legal, and Terminal-bench 2.1, while also beating 5.6 sol across a bunch of these evals

more "ox-alpha" vagueposting, now from a DeepMind's research scientist but apparently even at max effort it's still pretty average at coding — realizes what it did wrong, apologizes, explains the mistake, then somehow does the exact same thing again very old gemini vibes

how quickly the tables have turned Meta is starting to look like Google—moving fast, shipping AI products, and finally building momentum meanwhile, mighty Google suddenly looks like last year's Meta: confused, reactive, and somehow watching everyone else move faster

openai missed a trick by keeping gpt-5.5 pro out of codex with codex having much higher usage limits and openai giving out so many resets, there is basically no reason to pay for the claude max 20x plan right now now i really hope they add gpt-5.6 "pro or sol ultra" to codex

openAI makes a lot of unexpected changes but i really hope the ads are limited to free accounts i get that they need revenue, and with only about 5% of 800 million weekly users paying, monetizing the other 95% makes sense but ads in paid accounts should never happen. ever.