hhhypergrowth's profile picture. Hypergrowth investor doing research into the tech behind the SaaS. Now with a premium newsletter, plus free articles monthly. -muji

muji @ hhhypergrowth

@hhhypergrowth

Hypergrowth investor doing research into the tech behind the SaaS. Now with a premium newsletter, plus free articles monthly. -muji

Pinned

Today's free post is on the Layers of AI - a tech explainer into what is underlying this wave of Generative AI and LLMs. #AI #GenerativeAI #LLMs hhhypergrowth.com/layers-of-ai/

hhhypergrowth's tweet image. Today's free post is on the Layers of AI - a tech explainer into what is underlying this wave of Generative AI and LLMs. #AI #GenerativeAI #LLMs 

hhhypergrowth.com/layers-of-ai/

Review of GPT-5.1 Pro vs Gemini 3. "Gemini 3 is my fast, sharp everyday brain. GPT-5.1 Pro is the slow, heavy brain I pull out when I really do not want to be wrong."

I've had access to GPT-5.1 Pro for the last week. It's a fucking monster... easily the most capable and impressive model I've ever used. But it's not all positive. Here's my review of GPT-5.1 Pro: shumer.dev/gpt51proreview

shumer.dev

My GPT-5.1 Pro Review

Hands-on GPT-5.1 Pro review: heavy-weight reasoning vs Gemini 3, backend strength, UX friction, and where it fits in my stack.



muji @ hhhypergrowth reposted

"Shocker, Google trained Gemini 3 on TPUs" is a great litmus test of basic lack of expertise for AI writers & tweeters. Gemini 1.0, 1.5, 2.0, 2.5, 3.0, PaLM 1 and 2 l have all been trained on different generations of TPUs. It's proudly stated in each model card since 2022. How…

This is the big story here. Google trained Gemini 3 Pro on Google’s own TPUs. No mention of Nvidia chips.



NVIDIA $NVDA just putting up the fastest seq growth in 2 years thanks to Blackwell Ultra Revenue $57B +22% seq (guided +15.5%) DC +25% seq Q4 Guide $65B +14% seq FCF $22B, 39% margin


BATTEN DOWN THE HATCHES IT IS TIME FOR $NVDA EARNINGS


These $GOOG $NVDA TPU takes today just keep getting more and more terrible. Somehow Gemini 3 proves TPUs are gonna take over the world! The chip supply side is going to just open up for TPU to overtake GPU! Folks are going to immediately switch from GPUs to TPUs, and change the…

Hot take on Gemini 3. Points 4 and 5 might literally change the world. Also, OpenAI is in trouble. • Great model, but it’s not AGI; has the same kind of problems all earlier models have had. Hallucinations and unreliability persist. • In short, scaling isn’t getting us to…



Cloudflare $NET acquiring Replicate, a win-win for both. Cloudflare just upleveled the inference capabilities of Workers, expanded their available OSS models (plus closed models GPT5 & Claude Sonnet), and will soon add Replicate's fine-tuning & custom models via containers.…

Big news: Replicate is joining Cloudflare. Replicate's going to carry on as a distinct brand, and all that will happen is that it’s going to get way better: it’ll be faster, we’ll have more resources, and it’ll integrate with the rest of Cloudflare’s Developer Platform.…



AppLovin $APP Rev +68%, +11.6% seq vs +8.9% seq last Q - accel Guide +13.9% seq - further accel EBITDA +5pp, +1pp seq FCF jumped +14pp seq to 75% !! Accel and projecting further accel


Cloudflare $NET up 8% on earnings Continued reaccel over last 2Qs to now +31%, RPO + 43%, cRPO +30% Sadly their head of product/engr (stupidly jettisoned by Servicenow last year) getting snatched away to become CEO of undisclosed public tech co


Loading...

Something went wrong.


Something went wrong.