Hamida Rahman
55 posts
Hamida Rahman
@hey_hamida
Experimenting with data, creating intelligence 📈 AI | Deep Learning | Tech mindset On a journey to make machines think smarter
Katılım Mart 2026
0 Takip Edilen32.5K Takipçiler
Hamida Rahman retweetledi
Anthropic built a great model on a bad foundation.
@subqai built a better model on the right foundation.
Linear scaling wins. @alex_whedon #subq
Alexander Whedon@alex_whedon
Introducing SubQ - a major breakthrough in LLM intelligence. It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA), And the first frontier model with a 12 million token context window which is: - 52x faster than FlashAttention at 1MM tokens - Less than 5% the cost of Opus Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention). Only a small fraction actually matter. @subquadratic finds and focuses only on the ones that do. That's nearly 1,000x less compute and a new way for LLMs to scale.
English
Hamida Rahman retweetledi
We thought FlashAttention was the limit. We were wrong.
@subqai is 52x faster.
The frontier just moved. @alex_whedon #subq
Alexander Whedon@alex_whedon
Introducing SubQ - a major breakthrough in LLM intelligence. It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA), And the first frontier model with a 12 million token context window which is: - 52x faster than FlashAttention at 1MM tokens - Less than 5% the cost of Opus Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention). Only a small fraction actually matter. @subquadratic finds and focuses only on the ones that do. That's nearly 1,000x less compute and a new way for LLMs to scale.
English
Hamida Rahman retweetledi
This is the shift animation needed. Instead of wrestling with complex software, @Anijam_ai lets you direct scenes on your phone through simple conversation. Pure storytelling, zero friction. 📱🎬
Anijam AI@Anijam_ai
Introducing Anijam AI — the first AI Animation Agent on your phone. Your story idea should not die before you open professional software. Just tell Anijam your idea. It does the rest. Available now on iOS & browser.
English
Hamida Rahman retweetledi
Stop writing sliding window prompts. Just drop the data in.
@subqai handles 12M tokens without breaking a sweat.
@alex_whedon #subq
Alexander Whedon@alex_whedon
Introducing SubQ - a major breakthrough in LLM intelligence. It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA), And the first frontier model with a 12 million token context window which is: - 52x faster than FlashAttention at 1MM tokens - Less than 5% the cost of Opus Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention). Only a small fraction actually matter. @subquadratic finds and focuses only on the ones that do. That's nearly 1,000x less compute and a new way for LLMs to scale.
English
Hamida Rahman retweetledi
Most browsers give you AI... then quietly track you with it. Neo delivers powerful AI with privacy locked down by default. Builtin Norton VPN, 11-signal fingerprint blocking, and pre-click phishing protection .. all active with zero extensions or setup. No more choosing between power and privacy.
Norton Neo Browser@neobrowser
Most browsers give you AI. Then quietly track everything you do with it. Neo gives you the AI and locks down your privacy at the same time. > Norton VPN built right in. > Fingerprinting blocked across 11 signal types. > Phishing caught before you ever click. All on by default. No extensions. No setup. No fine print. Because you shouldn't have to choose between a powerful browser and a safe one. go.neobrowser.ai/55 🩵
English
Hamida Rahman retweetledi
Forget complex timelines and desk software. @Anijam_ai lets anyone direct animated stories straight from their phone. You literally just chat. The gap between an idea and a scene is now zero 🚀
Anijam AI@Anijam_ai
Introducing Anijam AI — the first AI Animation Agent on your phone. Your story idea should not die before you open professional software. Just tell Anijam your idea. It does the rest. Available now on iOS & browser.
English
Hamida Rahman retweetledi
Situation: You think 200k tokens is enough because you have been conditioned to accept failure.
System: Demand 12M tokens. @subqai.
@alex_whedon
Alexander Whedon@alex_whedon
Introducing SubQ - a major breakthrough in LLM intelligence. It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA), And the first frontier model with a 12 million token context window which is: - 52x faster than FlashAttention at 1MM tokens - Less than 5% the cost of Opus Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention). Only a small fraction actually matter. @subquadratic finds and focuses only on the ones that do. That's nearly 1,000x less compute and a new way for LLMs to scale.
English
Hamida Rahman retweetledi
Been dogfooding Sai with Claude Code lately and the part that's actually stuck for us isn't just the responsiveness. It's that you can get work done on your phone. We have teammates improving our app from iMessage daily now.
This all started last week, when a teammate spotted a UI bug (black text on a black button). Normally that's a ticket → repro doc → Slack thread → fix-tomorrow situation. Instead Sai reproduced the bug, took screenshots, and handed it to Claude Code. Put up a PR before someone even read the bug report.
Whole thing started on our teammate's phone. On the go, laptop never opened, all in our VMs.
In preview now. Try it at sai.work with invite code: SAI_CLAUDECODE
English
Hamida Rahman retweetledi
We built an entire ecosystem of dev tools to fix the fact that LLMs cannot read a whole codebase.
@subqai just made them all obsolete.
@alex_whedon #subq
Alexander Whedon@alex_whedon
Introducing SubQ - a major breakthrough in LLM intelligence. It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA), And the first frontier model with a 12 million token context window which is: - 52x faster than FlashAttention at 1MM tokens - Less than 5% the cost of Opus Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention). Only a small fraction actually matter. @subquadratic finds and focuses only on the ones that do. That's nearly 1,000x less compute and a new way for LLMs to scale.
English
Stop chunking your documents. Stop summarizing your logs.
@subqai takes the whole dataset in one shot.
True automation requires true memory. @alex_whedon
Alexander Whedon@alex_whedon
Introducing SubQ - a major breakthrough in LLM intelligence. It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA), And the first frontier model with a 12 million token context window which is: - 52x faster than FlashAttention at 1MM tokens - Less than 5% the cost of Opus Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention). Only a small fraction actually matter. @subquadratic finds and focuses only on the ones that do. That's nearly 1,000x less compute and a new way for LLMs to scale.
English
Hamida Rahman retweetledi
Introducing SubQ - a major breakthrough in LLM intelligence.
It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA),
And the first frontier model with a 12 million token context window which is:
- 52x faster than FlashAttention at 1MM tokens
- Less than 5% the cost of Opus
Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention).
Only a small fraction actually matter.
@subquadratic finds and focuses only on the ones that do.
That's nearly 1,000x less compute and a new way for LLMs to scale.
English
Stories shouldn't die just because animation is hard. 🎬 @Anijam_ai completely flips the script. No complex software—just grab your phone, start a conversation, and watch your ideas come alive.
Anijam AI@Anijam_ai
Introducing Anijam AI — the first AI Animation Agent on your phone. Your story idea should not die before you open professional software. Just tell Anijam your idea. It does the rest. Available now on iOS & browser.
English