iiinova — where your digital self keeps evolving

Ashutosh Shrivastava @ashutoshai.com · 30/12/2024

4/4 ChatGPT’s audience is 54.48% male and 45.52% female. The largest age group of visitors are 25-34 years old.

010

Ashutosh Shrivastava @ashutoshai.com · 30/12/2024

3/4 YT drives the most traffic to ChatGPT from any other social media.

100

Ashutosh Shrivastava @ashutoshai.com · 30/12/2024

2/4 ChatGPT gets mostbweb traffic from USA and India.

100

Ashutosh Shrivastava @ashutoshai.com · 30/12/2024

In just two years, ChatGPT has become the 8th most visited site in the world, absolutely insane! The USA and India lead as the top users of ChatGPT. Here are 4 stats I found on similarweb that not everyone knows 👇 1/4

120

Ashutosh Shrivastava @ashutoshai.com · 18/12/2024

What Google did to 12 Days of Shipmas is brutal. Veo 2 took just few days to outdate OpenAI Sora. Google truly got revenge for OpenAI stealing the thunder from their 1M context token release by announcing Sora on the same day.

010

Ashutosh Shrivastava @ashutoshai.com · 14/12/2024

Messi and I enjoying a football match in latent space, because I don't think it’s ever going to happen in reality 😔 I’m a huge Messi fan. Created using Pika 2.0.

030

Ashutosh Shrivastava @ashutoshai.com · 06/12/2024

Okay this was my first prompt for Google new gemini-exp-1206 and it nailed this one.. I will do more testing but this definitely looks good to me. Prompt : Write a 3-paragraph essay on strawberries without using the letter 'e' Lucky or not I dunno but this was the response.

020

Ashutosh Shrivastava @ashutoshai.com · 06/12/2024

Shipmas 🎅 Google just released another Gemini model, gemini-exp-1206, available for eveyone in AI Studio for free. Looks like they are getting ready for Gemini 2.0

130

Ashutosh Shrivastava @ashutoshai.com · 06/12/2024

It took the Gamma app just 2 minute and 45 seconds to create this static website.

000

Ashutosh Shrivastava @ashutoshai.com · 06/12/2024

1. Log in to the Gamma app (gamma.app) 2. Go to Create with AI. 3. Provide your document - you can use a PDF, PPT, or just paste plain text. 4. Select Website (you can also create a PPT or document). 5. Choose your settings, and you’re done!

100

Ashutosh Shrivastava @ashutoshai.com · 06/12/2024

Incredible what AI can do! In less than 180 seconds, I turned a plain word document file into a perfect static website using AI. With 25M+ users and counting, this tool is unbelievable, no one even comes close. Plus, they’ve got a free tier! Just try it... Here’s a short tutorial 👇

120

Ashutosh Shrivastava @ashutoshai.com · 06/12/2024

o1-preview vs o1 vs o1 Pro benchmarks for Math and Phd Level Science.

110

Ashutosh Shrivastava @ashutoshai.com · 06/12/2024

12 Days of OpenAI: Day 1 - ChatGPT Pro Mode – $200 - ChatGPT o1 Pro Mode – Uses more compute and delivers greater intelligence than o1. - ChatGPT o1 with Multimodal – Now available. All the important updates from Day 1 are here 🧵👇

110

Ashutosh Shrivastava @ashutoshai.com · 05/12/2024

OpenAI is gonna release Sora. Rest of of the video generation tool

000

Ashutosh Shrivastava @ashutoshai.com · 04/12/2024

Google announced Genie 2, a foundation world model capable of generating an endless variety of action-controllable, playable 3D environments for training and evaluating embodied agents. Based on a single prompt image, it can be played by a human or AI agent using keyboard and mouse inputs.

010

Ashutosh Shrivastava @ashutoshai.com · 04/12/2024

Starting tomorrow at 10 am pacific, OpenAI are doing 12 days of openai. each weekday, they will have a livestream with a launch or demo.

020

Ashutosh Shrivastava @ashutoshai.com · 04/12/2024

I’ve created the best Hailuo video of all time, and I made it with my son, who loves dinosaurs alot. He absolutely loved this video!

040

Ashutosh Shrivastava @ashutoshai.com · 03/12/2024

Amazon launched Nova ( family of foundation model) and Pro model is already better than GPT-4o / Claude and Gemini Pro. More details below 👇

230

Ashutosh Shrivastava @ashutoshai.com · 03/12/2024

OpenAI - ❌ ClosedAI - ✅ ProfitAI - ✅ OpenAI is the only 'non-profit' company that seems entirely focused on making a profit and increasing revenue. 🙄

020

Ashutosh Shrivastava @ashutoshai.com · 03/12/2024

Use this trick to make ChatGPT say David Mayer 😉

000

Ashutosh Shrivastava @ashutoshai.com · 02/12/2024

So, AI is only behind humans in complex math and common sense reasoning. We’re so cooked. 🙄 Source : 2024 Stanford AI Index Report

161

Ashutosh Shrivastava @ashutoshai.com · 27/11/2024

I created this podcast using Play AI new PlayDialog beta model, and literally, it's way better than Google NotebookLM : - API access for building custom applications. - 16 different voice . - I also build a starter application to help you get started.

130

Ashutosh Shrivastava @ashutoshai.com · 26/11/2024

OpenAI has shut down Sora early access temporarily for all artists.

010

Ashutosh Shrivastava @ashutoshai.com · 26/11/2024

OpenAI Sora leaked on Huggingface by early testers. The artists involved are angry because they feel exploited. They claim they were invited as "testers" but ended up doing free work, giving feedback, and creating content that only benefits OpenAI marketing.

110

Ashutosh Shrivastava @ashutoshai.com · 26/11/2024

Once the 3D model is ready, you can animate it

110

Ashutosh Shrivastava @ashutoshai.com · 26/11/2024

check: www.csm.ai For this video , I used the image-to-3D model feature. You can choose from different types of models based on quality and texture. Once you have 3d model ready you can use AI retexture to improve it even more.

100

Ashutosh Shrivastava @ashutoshai.com · 26/11/2024

CSM is like OpenAI, but for AI 3D workflows, I created this image-to-3D model and even animated it. CSM can do alot : - Image to 3D / Sheets to 3D - AI Retexturing - Animate 3D Model - 3D Generative Agents (launched today) They also offer a free tier so try out. more details 👇

110

Ashutosh Shrivastava @ashutoshai.com · 25/11/2024

MCP follows a client-server architecture where multiple services can connect to any compatible client. Clients are applications like Claude Desktop, IDEs, or AI tools. Servers are light adapters that expose data sources.

100

Ashutosh Shrivastava @ashutoshai.com · 25/11/2024

Anthropic launched the Model Context Protocol (MCP) F**K man it looks like software developers have one more less job now. Claude desktop can now connect directly to GitHub, create a new repo, and make a PR. More details 👇

440

Ashutosh Shrivastava @ashutoshai.com · 23/11/2024

What the f*ck 🙄 This is probably the worst use case of AI, and a whole lot of people will lose jobs if this gets implemented. If employees are so stressed about monitoring system, performance will go down anyway.

010

Ashutosh Shrivastava @ashutoshai.com · 23/11/2024

🚨 Darin Fisher has joined OpenAI. Can we now confirm that OpenAI is building its own browser and going head-to-head with Google?

131

Ashutosh Shrivastava @ashutoshai.com · 23/11/2024

AI exploded in 2023, Nvidia's data center revenue skyrocketed from $5 billion to over $30 billion. Jensen is the real winner of AI, it's only going to increase.

020

Ashutosh Shrivastava @ashutoshai.com · 22/11/2024

Google Gemini team are cooking... They shipped the new model, "gemini-exp-1121," less than 12 hours ago, and now they’ve shipped another update in AI Studio "model selector experience". Love that they have added "Best for" and "Use case" too .

000

Ashutosh Shrivastava @ashutoshai.com · 21/11/2024

What the heck, this can't be true! The new Google Gemini experimental model 'Gemini-exp-1121' is number 1 for almost all in lmsys arena?? Surprisingly number 1 for coding too.

120

Ashutosh Shrivastava @ashutoshai.com · 21/11/2024

gemini-exp-1121! Google’s latest experimental gemini model, with: - significant gains on coding performance - stronger reasoning capabilities - improved visual understanding Available on Google AI Studio and the Gemini API right now:

110

Ashutosh Shrivastava @ashutoshai.com · 21/11/2024

This is F***ING crazy... Now your AI avatar can take Zoom calls in real-time, and you don’t even have to sit in front of the camera. Imagine giving an online interview now. What is real anymore?? Source : Pickle(X)

152

Ashutosh Shrivastava @ashutoshai.com · 21/11/2024

Full trends in what people think.

020

Ashutosh Shrivastava @ashutoshai.com · 21/11/2024

Over 29% of HR professionals from Gen Z think AI will replace them in the hiring process.

120

Ashutosh Shrivastava @ashutoshai.com · 20/11/2024

You can try this at 🌐 Try it now at chat.deepseek.com

010

Ashutosh Shrivastava @ashutoshai.com · 20/11/2024

What the heck! I wasn't expecting this for sure... DeepSeek-R1-Lite-Preview is as good as o1-preview and even outperforming it in AIME and Math. open-source models and APIs are coming soon!

130

Ashutosh Shrivastava @ashutoshai.com · 20/11/2024

Google is releasing a new model, LearnLM 1.5 Pro Experimental. It's available to me on Google AI Studio Could this be aimed at educational purposes or universities? Has anyone else seen this and got more info?

020

Ashutosh Shrivastava @ashutoshai.com · 20/11/2024

What the F**K? The o1-preview model got only 1 question wrong in the Korean SAT exam, placing it among the top 4%. What's crazy is that these were new questions, meaning they were outside the training data, and yet o1-preview nailed it. Bro, what are we supposed to do?

060