AshutoshShrivastava

@ai_for_success on X

Writes about AI. 81% of posts open with an observation. A typical post reaches 451 likes.

Remix their style

Posts: 21
Likes: 15K
Typical: 451
Best: 2.7K

What they write about

AI: 90%

How they open

observation: 81%
number: 10%
question: 10%

Every post of theirs in the library

AshutoshShrivastava@ai_for_success

Price comparison of OpenAI o1 and DeepSeek AI R1: R1 is significantly cheaper across all categories (96–98% savings). Now you know why big organizations don't want open-source to continue, If humanity is ever going to benefit from AI, it will be from open-source .

Opens with an observation

AshutoshShrivastava@ai_for_success

Somebody plz stop China 🔥🔥Hunyuan-T1 is here.. - Better than GPT-4.5 and DeepSeek R1. - ultra-large-scale reasoning. - Hybrid-Mamba-Transformer MoE Architecture – - Strong Logic & Concise Writing - 60-80 tokens/sec generation speed. More details below 👇

Opens with an observation

AshutoshShrivastava@ai_for_success

What the heck, DeepSeek's cost-profit margin is 545%. They don’t charge $200, and everything is open source. Just imagine what DeepSeek would do if they got $500B. Respect the 🐋.

Opens with a number

AshutoshShrivastava@ai_for_success

How the heck is Anthropic able to do this? Anthropic is projecting twice the revenue from selling AI models to businesses this year compared to OpenAI. Right now, OpenAI’s only moat is ChatGPT’s 800M+ monthly active users, and Google is coming for that too.

Opens with a question

AshutoshShrivastava@ai_for_success

Can someone ask China to take a break? 🤯 Qwen just dropped another model, Qwen2.5-Max, which outperforms DeepSeek V3 You can try it for free on Qwen Chat and also available on API. More details below 👇

Opens with a question

AshutoshShrivastava@ai_for_success

This is probably the best AI video of all time. At first, it looks simple , but then you realize how insanely good Veo-2 is at simulating the physical world. Damn Google how?? 🫡

Opens with an observation

AshutoshShrivastava@ai_for_success

So, Google really rolled out their best model to everyone even free users. That’s good competition 🤣 I can't wait for the final version of Gemini 2.5 Pro. And if they integrate it with the Gemini Canvas feature, it’s gonna be awesome!

Opens with an observation

AshutoshShrivastava@ai_for_success

Google Gemini 2.0 Flash Multimodal feature is next-level! I can’t believe Google rolled this out before OpenAI. It takes live feeds, answers in real-time with almost no latency, Pretty mind-blowing. Thanks, Google, for this early Christmas gift! Adding more details below 👇

Opens with an observation

AshutoshShrivastava@ai_for_success

Everyone should take some time to read this article from Satya Nadella. The future winners will not be the companies with the best models. They'll be the companies with the best learning loops between humans and AI.

Opens with an observation

AshutoshShrivastava@ai_for_success

GOOGLE IS SO BACK 🔥 Gemini 3.8 Flash is beating GPT-5.6 Sol and Claude Opus 5 on several benchmarks, while costing significantly less and being extremely fast. It’s already live for me in the Gemini app.

Opens with an observation

AshutoshShrivastava@ai_for_success

What Moonshot AI is doing is unbelievable. HLE and IMO both reached SOTA, and it’s SOTA worldwide, not only SOTA in China. This company’s valuation (Moonshot AI) is only 0.5% of OpenAI’s, 2% of Anthropic’s and Grok’s. And they manage to bring up some model like this.

Opens with an observation

AshutoshShrivastava@ai_for_success

Google is so back 🔥🔥 They just announced Veo 2, an AI video generation model, and it’s incredible. They’ve also updated the Imagen 3 image-generation model. 10 examples below 👇

Opens with an observation

AshutoshShrivastava@ai_for_success

🚨 Gemini 3 Deep Think got a major upgrade focused on solving hard problems in science, research, and engineering using advanced reasoning. - 84.6% on ARC-AGI-2 - 48.4% on Humanity’s Last Exam without tools - 3455 Elo score on Codeforces

Opens with an observation

AshutoshShrivastava@ai_for_success

Google processes over 1.3 quadrillion tokens every month, that’s equivalent to: > 500 million per second > 30 billion per minute > 1.8 trillion per hour > 43 trillion per day > 303 trillion per week

Opens with an observation

AshutoshShrivastava@ai_for_success

🚨 IT'S OVER!!! Meta’s $2 billion Manus acquisition blocked by China. On April 27, 2026, China’s National Development and Reform Commission blocked Meta AI’s acquisition of AI startup Manus, ordering all parties to cancel the deal.

Opens with an observation

AshutoshShrivastava@ai_for_success

⚡️Gemini 3.6 Flash is now available in Google AI Studio, and it's cheaper than Gemini 3.5 Flash. Pricing: - Input: $1.50 - Output: $7.50 (Gemini 3.5 Flash: $9.00) Knowledge cutoff: March 2026.

Opens with an observation

AshutoshShrivastava@ai_for_success

It's so over guys... GOOGLE has new model Gemini 2.0 Flash Thinking Experimental, thier reasoning model and I was correct with my prediction. It's available on AI Studio Thanks you @BalajiAkiri for sharing the info. This is real shipmas🎅

Opens with an observation

AshutoshShrivastava@ai_for_success

Google DeepMind’s new model, Gemini 3.1 Pro, is now available on AI Studio and the Gemini app and NotebookLM. They have cooked this model really well. You are all going to love it. Thanks to the Gemini team for early access. It was fun testing it. 77.1 in ARC-AGI-2 🔥

Opens with an observation

AshutoshShrivastava@ai_for_success

🚨 John Jumper, Nobel Prize winner is leaving Google DeepMind after 9 years to join Anthropic. Yesterday it was Noam Shazeer joining OpenAI. Today it's John Jumper joining Anthropic. What's happening 👀

Opens with an observation

AshutoshShrivastava@ai_for_success

Gemini API usage went up from 8.3 trillion to over 90 trillion monthly tokens in a year on Vertex AI from retail customers alone. That is more than 11x growth year over year. Imagine the number by the end of the year.

Opens with an observation

AshutoshShrivastava@ai_for_success

🚨 Atlassian is laying off 10% of its workforce. CEO Mike Cannon-Brookes explained in a video to employees that about 1,600 roles will be cut as the company restructures around AI. He said he is "deeply sorry for the disruption" the layoff creates in impacted people lives.

Opens with a number