Skip to content

ViralLoop

Your next post already worked.

← All creators

Posts
Likes
Typical
Best

Every post of theirs in the library

AllXLinkedInInstagramTikTok

FilterOrderGridTableSelect

Loading their posts

← All creators

Haider.

Writes about AI and Revenue. 100% of posts open with an observation. A typical post reaches 373 likes.

Remix their style

Posts: 10
Likes: 6.9K
Typical: 373
Best: 2.8K

What they write about

AI: 90%
Revenue: 20%

How they open

observation: 100%

Every post of theirs in the library

AllX LinkedIn Instagram TikTok

Haider.@haider1

Andrej Karpathy says today's agents aren't ready to work like real coworkers or interns They lack intelligence, can't use computers, aren't multimodal, lack continual learning, and forget what you tell them Fixing these gaps will take about a decade

Opens with an observation

2.8K likes188 repliesAIX10mo ago

Haider.@haider1

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

Opens with an observation

1.2K likes49 repliesAIX2mo ago

Haider.@haider1

finally, gemini 2.5 pro has been dethroned after a long time o3 beats gemini 2.5 on LiveBench, and it looks like both models ain't going anywhere at least in coding: - o3 score 73 - o4-mini score 74 - gemini 2.5 score 58 also, o4-mini is 2x cheaper than gemini 2.5 pro

Opens with an observation

749 likes59 repliesAIX1y ago

Haider.@haider1

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

Opens with an observation

487 likes8 repliesAIX8d ago

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

Opens with an observation

412 likes21 repliesHealthX1mo ago

Haider.@haider1

sam altman leaked gpt-5 tweet is probably fake but it feel like he has a plan to spoil the google i/o announcements openAI has done this every time, so it seems they’re holding back some big updates just for this reason

Opens with an observation

333 likes26 repliesAIX1y ago

Haider.@haider1

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

Opens with an observation

293 likes14 repliesScienceX20d ago

Haider.@haider1

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

Opens with an observation

256 likes11 repliesAIX9d ago

Haider.@haider1

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.

Opens with an observation

186 likes34 repliesMarketingX9mo ago

Haider.@haider1

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

Opens with an observation

185 likes17 repliesAIX1mo ago

Haider. (@haider1) — what they write about and why it works · ViralLoop